Schedule at a glance

 

	Monday, August 19, 2024
	14:30～18:00	Registration
	18:00～20:00	Conference Cocktail Reception
	Tuesday, August 20, 2024
	9:00～9:30	Opening Ceremony
	9:30～10:30	Keynote Speech:
		Synergistic Evolution: The Convergence of Signal Processing,
		Communication Technologies and Computational Intelligence
		for Smart Energy Systems
		Prof. Henry Shu-hung Chung
		City Unversity of Hong Kong; IEEE Distinguish Lecturer
	10:30～11:00	Coffee Break
	11:00～12:00	Invited Speech
		AI for next-generation network security:
		A research case study
		Tram Truong-Huu
		Singarpore Institute of Technology
	12:00～12:30	Industry Sponsor Presentation
	12:30～14:00	Lunch
		Track 1               Cinnamon Ballroom	Track 2           Peppercorn 1	Track 3           Peppercorn 2
	14:00～16:00	BSPC 01	CPT 01	COM 01
	18:00～20:00	Dinner - Cinnamon Ballroom
	Wednesday, August 21, 2024
		Track 1               Cinnamon Ballroom	Track 2           Peppercorn 1	Track 3           Peppercorn 2
	9:00～10:30	SPG 01	CPT 02	COM 02
	10:30～10:50	Coffee Break
	10:50～12:20	SPG 02	CPT 03	COM 03
	12:30～14:00	Lunch
	14:00～15:30	SPG 03	SPG 07	COM 04
	15:30～16:00	Coffee Break
	16:00～17:30	SPG 04	SPG 08	COM 05
	18:30～20:00	Conference Banquet - Cinnamon Ballroom
	Thursday, August 22, 2024
		Track 1               Cinnamon Ballroom	Track 2           Peppercorn 1	Track 3           Peppercorn 2
	9:00～10:30	SPG 05	SPG 09	SPG 11
	10:30～10:50	Coffee Break
	10:50～12:20	SPG 06	SPG 10	SPG 12
	12:20～14:00	Lunch
	SPG: Oral Presentation for Signal Processing
	CPT: Oral Presentation for Computing
	COM: Oral Presentation for Communication
	BSPC: Best Student Paper Contest

	WELCOME FROM THE CONFERENCE CHAIRS
	On behalf of the organizing committee, we would like to extend a warm welcome to all the participants of the IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC) 2024 located in Bali, Indonesia. The goal of this conference stays the same regardless of whether it is held in-person or online, i.e., to bring together leading engineers, researchers, and scholars from all over the world to discuss new theories, technologies, and applications in the field of signal processing, communications, and computing. We hope that this year's conference can make every outstanding participant full of enthusiasm and motivation, and facilitate academic exchanges smoothly.
	ICSPCC 2024 is composed of many volunteers, and we thank all of them for their time, energy, and efforts in service to our conference. In particular, we would like to express our gratitude to the Technical Program Chairs Gongping Huang, Mahnaz Arvaneh, and Peter Tam for their leadership in handling efficiently the review of the submitted papers and putting together an exciting technical program, the Special Session Chair Hamed Ahmadi, Xiaolei Zhang, and Wentao Shi for putting together some high-quality special sessions, and the Student Best Paper Award Chair Xiaodong Cui, Jing Han, and Wing-hong LAU, and their team for identifying high-quality student papers and organizing the student paper award competition. We are very grateful to all the keynote speaker, invited speaker, session chairs, reviewers, student volunteers, and authors for their contributions to the technical success of this conference. We are most grateful for the sponsorship from IEEE Xi’an Section, Northwestern Polytechnical University, Lian Feng Acoustic Technologies Co., Ltd, and the technical sponsorship from IEEE Hong Kong CASCOM Joint Chapter, IEEE Xi’an SP Chapter.
	We look forward to welcoming you and hope that you will enjoy this fourteenth ICSPCC!
	Steering Committee Chairs:
	Prof. Jianguo Huang, Northwestern Polytechnical University, Xi’an
	Prof. Jingdong Chen, Northwestern Polytechnical University, Xi’an
	General Chairs:
	Prof. Susanto Rahardja, Singapore Institute of Technology, Singapore
	Prof. Jie Chen, Northwestern Polytechnical University, Xi’an
	Prof. Pasi FRÄNTI, University of Eastern Finland, Finland
	Prof. Ray Cheung, City University of HongKong
	WELCOME FROM THE TECHNICAL PROGRAM CHAIRS
	As co-chairs Technical Program Committee, it is our pleasure to welcome you in this face-to-face to an exciting technical program offered by our 14th IEEE ICSPCC.
	Now let me share with you some information about the technical program of this year. The technical program for this year’s conference includes keynote speech sessions, regular sessions, and a best student paper contest session. The keynote speaker, Henry Shu-Hung Chung, from the Hong Kong Polytechnic University, will give us his talk in the mornings of the 20th. The invited speaker, Tram Truong-Huu, from the Singapore Institute of Technology, will give us his talk in the mornings of the 20th. Our regular sessions will take place in parallel, and cover a variety of broad and interesting topics. We can see in this program how signal processing, communications, and computing continue to play an important role in the technological revolution in modern society. Here are some "classic" fields, including array signal processing, estimation and detection theory, wireless communication, and speech processing. Additionally, our sessions also encompass a range of works on new methods and applications of AI techniques. Following the tradition of ICSPCC, we also organized a student paper competition.
	In terms of manuscript processing, this year we received a total of 173 submissions. All papers submitted for the special and regular sessions underwent a rigorous peer-review process, resulting in the selection of 119 papers. The acceptance rate of the paper is approximately 68.8%. Selecting these papers required significant efforts from the reviewers, who wrote 346 review reports to ensure that each submission received at least two reports. We would especially like to thank them for their valuable work. As we mentioned, we will have a session on the best student paper contest. Based on the evaluations, Student Paper Award Co-Chairs Xiaodong Cui, Jing Han and Wing-hong Lau have shortlisted 6 candidates out of 65 initial requests for the award. The final winner will be selected from these 6 candidates during the contest session this afternoon. And I believe that you are also interested in participating in the award ceremony.
	It takes a lot of effort from many individuals to prepare the program of ICSPCC 2024. I would like to extend special thanks to the other Technical Program Chairs Prof. Gongping Huang, Mahnaz Arvaneh, and Peter Tam for their dedication and hard work which helped us complete an outstanding technical project.
	There are many scenic places to visit in Bali Island, and participants will have unforgettable experiences. I hope that this trip to Bali Island will provide every participant with not only a pleasant academic exchange but also a wonderful and unforgettable travel experience.
	Finally, we very much wish you an enjoyable and productive conference.
	Technical Program Chairs:
	Gongping Huang, Wuhan University
	Mahnaz Arvaneh, University of Sheffield
	Peter Tam, IEEE HongKong Section
	KEYNOTE SPEECH
	Time: 9:30~10:30, Tuesday, August 20, 2024
	Session Chair: Edward Cheung, Polytechnic University of HongKong
	Synergistic Evolution: The Convergence of Signal Processing, Communication Technologies and Computational Intelligence for Smart Energy Systems
	Professor Henry Shu-Hung Chung
	City University of Hong Kong, China; IEEE Distinguish  Lecturer
	Abstract: From microwatts to gigawatts, Power Electronics, operates behind the scenes as a critical technology that drives various daily applications, plays a central role in efficiently and flexibly converting, controlling, and transmitting energy. It is projected that by 2030, as much as 80 percent of global electric power will depend on various forms of power electronics systems. This expansion is fueled by notable progress in microelectronics, embedded systems, power semiconductor devices, and communication technologies. This keynote address will investigate the transformative potential of merging Signal Processing, Communication technologies, and Computational Intelligence (SPCC) with power electronics to revolutionize smart energy systems. The discussion will illustrate how this integration can enhance energy
	efficiency, fortify system stability, and establish a path towards a sustainable energy future. Furthermore, the discourse will underscore the interdisciplinary essence of this amalgamation, welcoming diverse fields to involve in reshaping the energy technology sphere. Through collaborative endeavours and innovative solutions, the fusion of SPCC and power electronics stands to unveil fresh opportunities in energy management, grid integration, and the utilization of renewable energy sources.
	BIOGRAPHY
	Henry Shu-Hung Chung received the B.Eng. and Ph.D. degrees in electrical engineering from Hong Kong Polytechnic University, Hong Kong, in 1991 and 1994, respectively. Since 1995, he has been with the City University of Hong Kong, where he is currently the Dean of Students, a Chair Professor with the Department of Electrical Engineering, and the Director of the Centre for Smart Energy Conversion and Utilization Research. He has published two books, authored ten research book chapters, and more than 500 technical papers including 250 refereed journal papers in his research areas, and holds 80 patents. His current research interests include renewable energy conversion technologies, lighting technologies, energy harvesting, smart grid technologies, and computational intelligence for power electronic systems. Dr. Chung is Fellow of IEEE. He was recipient of 2021 IEEE PELS R. David Middlebrook Achievement Award for his contribution to energy utilization technologies for smart cities. He received CityU Outstanding Research Award in 2020 and CityU Teaching Excellence Awards in 2018 and 2022, respectively. He is currently an Associate Editor for the IEEE Transactions on Power Electronics and the IEEE Journal of Emerging and Selected Topics in Power Electronics. He was the Editor-in-Chief of the IEEE Power Electronics Letters 2014–2018. He was also the Chair of the Technical Committee of the High-Performance and Emerging Technologies, IEEE Power Electronics Society in 2010–2014. He was the recipient of numerous industrial awards for his invented energy saving technologies.
	INVITED SPEECH
	Time: 11:00~12:00, Tuesday, August 20, 2024
	Session Chair: Prof. Susanto Rahardja, Singapore Institute of Technology, Singapore
	AI for next-generation network security: A research case study
	Dr. Tram Truong-Huu
	Singapore Institute of Technology, Singapore
	Abstract: Artificial intelligence has demonstrated its capability in various domains such as computer vision, natural language processing, and robotics. In this talk, I will introduce the application of AI to the area of next-generation network security. I will present the motivation and challenges when adopting AI to networking and how our team addressed such scientific challenges through our research works. I will demonstrate the usefulness of AI, especially the recent developments such as generative adversarial networks and deep graph convolution neural networks in two applications including network anomaly detection and classification of mobile applications using network traffic. Before concluding my talk, I will draw a few future research directions involving the use of large language models or foundation models in next-generation network security.
	BIOGRAPHY
	Dr. Tram Truong-Huu is an Associate Professor at the InfoComm Technology (ICT) Cluster, Singapore Institute of Technology (SIT). He also holds a joint appointment with the Agency for Science, Technology and Research (A*STAR) where he has worked as a senior computer scientist at the Institute for Infocomm Research (I2R) since May 2019. He received the Ph.D. degree in computer science from the University of Nice - Sophia Antipolis (now Côte d'Azur University), France in December 2010. From January 2011 to June 2012, he held a post-doctoral fellowship at the French National Center for Scientific Research (CNRS), France. He worked at the National University of Singapore as a research fellow from July 2012, then senior research fellow from January 2017. His research focuses on the application of artificial intelligence to cybersecurity and networking. He has been a member of the IEEE since 2012 and Senior Member of the IEEE since 2015.

	Oral Presentation Instructions
	IEEE Stipulated that one of the author shall present his paper in the conference in according to the IEEE requirement, or the paper may not be published in IEEE Explore Digital Library (and EI).  The authors should try best to join the conference and shared his research work and experience.
	The presentation is a sharing of experience and is not an examination.  Delegates shall take it easy and have a good practice is communicating his research result in the presentation, not with any pressure at all.   Delegates should not worry the English presentation but should treat it as an oral practice and learning process.
	Oral Presentation enable author to share their knowledge and experience with experts and researchers.   Authors should relax, just summarize the key points of the Paper and give 10-15 minutes talk. Just like talking to your colleagues will do.
	The allocated presentation time for each paper is 15 minutes which includes the time for questions and answer from the audiences. In general, each PPT slide presentation takes 30 seconds, so your PPT should not have more than 30 slides.
	Presenters are required to report to their Session Chairs at least 10 minutes prior to the start of their Session. All oral presentations must be loaded from USB into the note-book and tested before the session. PDF, Power Point are recommended. Movies or animations in MPEG, Windows Media, and etc., should be tested before the session.

	Session  SPG01
	Time - 9:00 ~ 10:30, Wednesday, August 21, 2024
	Session:	SPG01 - 1, Paper No.3531
	Title:	Study on Experimentation of LDVT Frequency Band Characteristics
	Author:	Zhenhao Wang, Pengda Ren and Bin Wang
	Abstract:	This paper presents a novel testing apparatus designed to evaluate the frequency characteristics of Linear Variable Differential Transformer (LVDT). The proposed system incorporates a rotary wheel displacement actuation mechanism to comprehensively investigate the dynamic behaviors of LVDTs. Key components include a servo motor, rotary wheel, sliding rail, LVDT fixture, the sensor under test, conditioning circuitry, acquisition equipment, and an upper computer. By employing controlled displacement simulation through the servo motor and sinusoidal signals, the system applies the linear sweep method to analyze the response of the LVDT across various frequency ranges. Experimental data is leveraged to derive the mathematical model of the LVDT, thus obtaining its transfer function. Notably, the amplitude bandwidth and phase bandwidth of the tested LVDT are essentially equal. The work demonstrates that it is feasible to study the frequency band characteristics of the sensor by inputting sinusoidal signals of the same magnitude but different frequencies into the displacement sensor and applying spectral analysis methods.
	Session:	SPG01 - 2, Paper No.3535
	Title:	SR-YOLOv8: Small Object Detection Method for Traffic Scenes
	Author:	Chenghong Zhang, Bo Yu, Wei Wang, Hanting Wei and Yuanze Meng
	Abstract:	Small object detection is one of the key challenges faced by object detection in traffic scenes. In this paper, we propose SR-YOLOv8, a small object detection algorithm for traffic scenes. Firstly, we introduce the SPD-Conv architecture to solve the problem of feature information loss in YOLOv8 downsampling process. Then, the ResBlock-CBAM module is designed to enhance the ability of YOLOv8 neck to extract important features. Finally, we construct the SR-YOLOv8 network structure for small object detection in traffic scenes. The effectiveness of the proposed algorithm is proved on the large-scale small object detection dataset SODA, the mAP50 and MAP50-95 are increased by 12.7% and 8.6% respectively, and the model complexity is much lower than other object detection models.
	Session:	SPG01 - 3, Paper No.3556
	Title:	Design and Implementation of a Lightweight Underwater Acoustic Modulation Classification
	Author:	Chaojin Ding, Xianbin Hu, Wei Su and Yifeng Zhao
	Abstract:	To develop an efficient and accurate lightweight underwater acoustic (UWA) modulation recognition algorithm, this study proposes an algorithm called Channel Expansion-Lightweight Recognition Network (CE-LRN). The CE-LRN significantly reduces computational complexity by adopting small convolutional kernels and quantized convolutional layer techniques, achieving deployment standards friendly to embedded devices.Additionally, a channel expansion module is introduced in the CE-LRN algorithm, which effectively improves recognition accuracy with only a slight increase in computational complexity. Compared with two recently proposed recognition algorithms, the CE-LRN greatly reduced computational complexity, with its floating-point operations (FLOPs) being only 16.9M, a reduction of over 83.4\%. Experimental results demonstrated that CE-LRN has good classification performance for Multi-Frequency Shift Keying (MFSK) and Multi-Phase Shift Keying (MPSK) signals.
	Session:	SPG01 - 4, Paper No.3576
	Title:	Digital Design Method of Passive Acoustic Zero-crossing Detection System
	Author:	Houkun Hu, Jian Luo and Yuyang Han
	Abstract:	Abstract—The traditional passive acoustic detection system has a long design cycle and a large cost of experimental improvement, which is difficult to meet the rapid development and iteration requirements of complex electronic systems. In this paper, the passive acoustic zero-crossing detection system is gradually decomposed into functional modules by using the top-down digital design method, and the functional modules are digitally modeled respectively. According to the requirements of the system design index, the digital prototype of the passive acoustic zero-crossing detection system is designed by using the established digital function module model. The function of each module is verified by single frequency signal. The parameters such as false alarm probability and anti-locking time are simulated iteratively by using the actual measured signal. The reasonable system parameters are optimized and the digital prototype is verified. The results show that the designed digital prototype can accurately detect the target, and the physical prototype design of the passive acoustic zero-crossing detection system can be carried out accordingly.
	Session:	SPG01 - 5, Paper No.3578
	Title:	A GNSS RTK/INS Tight Coupling Method Under Challenging Scenarios
	Author:	Ji Zikang, Zhao Yanyan, Zhao Hongwei and Li Xiaoang
	Abstract:	The Global Navigation Satellite System (GNSS) Real-Time Kinematic (RTK) positioning technology is a robust algorithm capable of providing centimeter-level accuracy in environments with favorable satellite visibility. However, in GNSS-Challenging environments such as urban canyons, where buildings and foliage may weaken GNSS signals, impede signal tracking by GNSS receivers, and exacerbate multi-path effects, RTK positioning can suffer performance degradation. The tight integration of Inertial Navigation Systems (INS) with RTK overcomes the limitations inherent when each system operates independently, delivering enhanced performance. This study realizes a tightly coupled RTK/INS algorithm. Performance tests and analyses were conducted on one public vehicular datasets and one empirically measured static data set, both of which presents challenges in RTK algorithms. The results indicate that the tight integration algorithm improved the positioning accuracy of the empirically measured static data and the open dynamic data set. Researchers may access our datasets via the provided link.
	Session:	SPG01 - 6, Paper No.3596
	Title:	Optimized Deployment Technology for Heterogeneous Multi-base Sonar System Based on The IHS-ABC Algorithm
	Author:	Xueli Sheng, Manxin Liu, Lei Tang, Guoliang Ji, Yan Wang and Hongbo Miao
	Abstract:	In addressing the optimization deployment issues of heterogeneous multi-base sonar systems, this paper establishes a detection performance evaluation model for heterogeneous multi-base sonar systems in complex underwater environments and proposes a hybrid algorithm combining improved harmony search (IHS) and artificial bee colony (ABC) algorithm called IHS-ABC. This algorithm draws from the concepts of the IHS algorithm, constructs a source memory, employs a full-dimensional probability selection strategy, and modifies the scouting mechanism. The precision and convergence speed of the algorithm are validated using classic benchmark functions. The algorithm is applied to optimize the deployment of heterogeneous multi-base sonar systems in various marine environments under the criterion of maximum coverage. Simulation results indicate that the optimized detection coverage significantly increased, achieving the goal of optimized deployment and enhancing detection performance.
	Session  SPG02
	Time - 11:00 ~ 12:30, Wednesday, August 21, 2024
	Session:	SPG02 - 1, Paper No.3604
	Title:	Mixed global and local attention alleviates domain shift between terahertz image datasets
	Author:	Rao Fu, Shaoxing Cui and Xiaoyi Feng
	Abstract:	Abstract—Terahertz imaging technology has a broad application prospect in the field of security inspection due to its strong penetrability and negligible radiation effect on people. With the development of machine learning in recent years, the application of machine learning methods to security inspection can save labor costs and ensure the accuracy of long-time security inspection. However, terahertz imagers produce different bottom noise at different temperatures, which leads to a significant domain offset problem between data sets collected at different temperatures. This domain shift makes the model trained on the source domain dataset to have a significant error when applied to the target domain dataset. Therefore, how to overcome the temperature-induced domain offset is an urgent problem for terahertz imaging techniques. In this paper, we innovatively propose to apply the attention mechanism to solve the problem of domain bias in object detection. We propose a lightweight attention mechanism Mixed global and local attention(MGLA) aimed at mitigating the domain shift between the target and source domains without affecting the source domain detection effect. MGLA can take into account the local information while fusing the channel information and spatial information and global information to improve the representation of the network. The experiment results show a 10% improvement in mAP0.5 and a 6.3% improvement in mAP50:95 compared to YOLOV8 baseline.
	Session:	SPG02 - 2, Paper No.3605
	Title:	Modulation Recognition of Underwater Acoustic Signals Based on Ghost-Former
	Author:	Jiwan Wang, Ke He, Hasqimeg Ordoqin, Haiyan Wang and Xiaohong Shen
	Abstract:	In order to address the issues of high computational complexity, low accuracy, and the cumbersome manual feature extraction steps involved in traditional algorithms for modulating and recognizing underwater acoustic signals, this paper proposes a modulation recognition method based on Ghost-former for hydroacoustic signals. This approach harnesses the advantages of Ghost Net in generating more feature maps through cost-effective operations, along with a two-way bridge for global interaction. Experimental results with simulated signals demonstrate that this method achieves excellent recognition performance even with relatively low FLOPS.
	Session:	SPG02 - 3, Paper No.3613
	Title:	MFCS-Depth: An Economical Self-supervised Monocular Depth Estimation Based on Multi-scale Fusion and Channel Separation Attention
	Author:	Zeyu Cheng, Yi Zhang, Xingxing Zhu, Yang Yu, Zhe Song and chengkai tang
	Abstract:	Self-supervised monocular depth estimation plays an extremely important role in fields such as autonomous driving and intelligent robot navigation. However, general monocular depth estimation models require massive computing resources, which seriously hinders their deployment on mobile devices, which is urgently needed in fields such as autonomous driving. To address this problem, we propose MFCS-Depth, an economical monocular depth estimation method based on multi-scale fusion and channel separation attention mechanism. We use the Transformer architecture with linear self-attention as its encoder to ensure its global modeling and economy. A high-performance and low-cost decoder has also been designed to improve the local and global reasoning of the network through multi-scale attention fusion and uses scale-wise channel separation to reduce parameters and computing costs significantly. Extensive experiments show that MFCS-Depth achieves competitive results with very few parameters on the KITTI and DDAD datasets and achieves state-of-the-art performance among methods of similar size.
	Session:	SPG02 - 4, Paper No.3614
	Title:	Toward Efficient Target Detection under Complex Clutter Backgrounds: An Unsupervised Energy-based Method
	Author:	Chenxi Zhang, Wenchao Chen, Chang Gao, Bo Chen and Hongwei Liu
	Abstract:	Traditional radar target detectors, each driven by a single mathematical model, often experience severe performance degradation in practice due to limitations in precisely modeling complex clutter. Existing data-driven deep learning methods significantly alleviate this problem but at the price of expensive and cumbersome annotations. Given this issue, we treat radar targets as anomalies within clutter data and develop an unsupervised Energy-based Anomaly Detector (EAD) for narrow-band radar target detection tasks. By modeling complex clutter flexibly with the energy-based model (EBM) and meticulously designing a radar target detection strategy, EAD can achieve superior detection performance in a straightforward manner. Experimental results on simulated data, finally shown in this paper, demonstrate the effectiveness of the proposed detector.
		Keywords—complex clutter background, narrow-band radar target detection, deep learning, unsupervised, energy-based model (EBM)
	Session:	SPG02 - 5, Paper No.3615
	Title:	Spatial Global Context Attention for Convolutional Neural Networks: An Efficient Method
	Author:	Yang Yu, Yi Zhang, Xingxing Zhu, Zeyu Cheng, Zhe Song and chengkai tang
	Abstract:	Capturing global contextual information within an image can significantly enhance visual understanding. However, current attention methods model long-range dependencies between features by aggregating query-specific global context to each query position. These methods are inefficient and consume a huge amount of memory and computational resources, making them less practical. To address this issue, we propose a simple, low-cost, and high-performance Spatial Global Context Attention (SGCA) module. This module aggregates query-independent global context to update features at each query position, capturing spatial global contextual information in an efficient and effective manner, significantly improving feature representations, which contributes to more precise classification results. The proposed SGCA is lightweight and flexible, making it suitable as an independent add-on component that can be applied to various convolutional neural networks (CNNs) to create a family of new architectures named SGCANet. Without bells and whistles, extensive experimental results on CIFAR-100 and ImageNet-1K for image recognition tasks demonstrate that our method significantly outperforms other counterparts in classification performance at a cheaper cost, achieving leading results.
	Session:	SPG02 - 6, Paper No.3621
	Title:	A high-precision direction finding method assisted by multipoint ranging sequences
	Author:	Yangyang Liu, Tianyu Li, Yuan Zhao, chengkai tang, Zesheng Dan and Yankang Bai
	Abstract:	The positioning accuracy of radio aviation search and rescue is composed of two parts: ranging error and direction finding error. Direction finding error is easily affected by complex environmental interference, leading to a rapid increase in positioning error. To solve this problem, this work proposes a novelly method of using multi-point ranging information to improve direction finding accuracy. Firstly, virtual target positions are generated using direction finding estimation information. Then, a set of ranging sequences is constructed, and direction finding information is estimated using multi-point ranging information and airborne position information in the set. Subsequently, a method using GDOP metric to update the set of ranging sequences is proposed, which significantly reduces the computational complexity and storage unit of the algorithm while ensuring its accuracy. Finally, simulation results show that the proposed method can achieve high-precision direction finding performance in complex environments.
	Session  SPG03
	Time - 14:00 ~ 15:30, Wednesday, August 21, 2024
	Session:	SPG03 - 1, Paper No.3626
	Title:	The Doppler Positioning Technology of Iridium-next and Oneweb Fusion Constellation
	Author:	Guanbo Chen, Yangyang Liu, Zhe Fan, Xiaoting Zhang, Deyan Li and Chengkai Tang
	Abstract:	With the advancement of technology, the Global Navigation Satellite System (GNSS) faces increasing challenges. In response to various disruptions or weakening of GNSS navigation satellite signals, and when the visibility of navigation satellites does not meet the requirements for pseudorange positioning in complex positioning environments, there is growing attention towards the use of Low Earth Orbit Satellites (LEO) for positioning. This study focuses on the establishment of a fused LEO satellite constellation using the Iridium-next constellation and some Oneweb satellites as examples. Considering the noncooperative nature of LEO satellites, an instantaneous doppler positioning model is developed. Two-Line Element (TLE) parameters are utilized as the orbit prediction model to obtain doppler shift information and receiver clock drift, enabling doppler positioning of LEO satellites.
	Session:	SPG03 - 2, Paper No.3629
	Title:	K-Means Clustering Algorithm Based on GEO Optimization
	Author:	Xu Zhang, Zesheng Dan, Yangyang Liu, Deyan Li, Xiaoting Zhang and Chengkai Tang
	Abstract:	This paper proposes a k-means algorithm enhanced with GEO (Golden Eagle Optimizer) optimization to address the challenges encountered by traditional k-means algorithm, such as being highly sensitive to initial values and prone to getting stuck in local optima during clustering. By incorporating GEO's attack and cruising vectors into the original loss function and iterative process of k-means, our algorithm enhances its exploratory capability while retaining its original clustering prowess. Theoretical analysis and simulation results demonstrate that our method can further minimize the loss function and exhibit superior clustering performance.
	Session:	SPG03 - 3, Paper No.3647
	Title:	Modified CA-CFAR Detector for Oversampled Sensors
	Author:	FengDeng Gu, Chang Gao, Qingfu Zhang, Rongrong Wang and Hongwei Liu
	Abstract:	To ensure less quantization error and sampling loss, oversampling is very common in radar, sonar, communications and other applications. The cell averaging constant false alarm rate (CA-CFAR) is widely used in target detection due to its ability to adaptively adjust the detection threshold based on varying background noise levels and maintain a constant false alarm rate. However, a high oversampling ratio (OSR) will increase the correlation degree of the sample data, which will distort the maximum likelihood estimation (MLE) of the background noise power obtained by the traditional CA-CFAR detector and consequently weaken its detection performance. To solve this problem, a modified CA-CFAR detector is proposed in this paper. Utilizing the colored filter matrix, the MLE of the covariance matrix of the data in the reference window is obtained, based on which the modified test statistics are derived via the generalized likelihood ratio test. The simulation results show that the modified CA-CFAR detector can achieve the upper performance bound for different OSRs.
	Session:	SPG03 - 4, Paper No.3660
	Title:	Classification of Lung Diseases through Artificial Intelligence Models: A Multi-Dataset Evaluation
	Author:	Masabah Bint E Islam and Muhammad Salman Khan
	Abstract:	Respiratory ailments like chronic obstructive pulmonary disease(COPD), asthma, and pulmonary fibrosis pose substantial diagnostic challenges that impact global health. This study combines artificial intelligence with non-invasive diagnostics, employing models from machine learning and deep learning model to evaluate on three open source lung sounds datasets. Mel-frequency cepstral coefficients (MFCCs) was employed for feature extraction and a range of models were evaluated including  Support Vector Machines (SVMs), Random Forests (RFs), Gradient Boosting, Naive Bayes (NB), Artificial Neural Network (ANN), Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM) networks and a transformer model for classifying respiratory sounds into nuanced categories, with defined performance metrics. Results illustrate the criticality of selecting models tailored to dataset characteristics; where prominent models include SVM, LSTM, and CNNs, to detect disease-specific acoustic patterns. This offers a path to understand toward accessible and efficient diagnostic methods in relation to open source datasets.
	Session:	SPG03 - 5, Paper No.3662
	Title:	A method for recognizing complex communication navigation signals based on neural network
	Author:	LIU Wenjing, Wang Beibei and Liu Xiaoya
	Abstract:	Electromagnetic spectrum resources have great economic, society and military value and are one of the important strategic resources of the country. With the rapid development of wireless technologies such as communication and Internet of Things, and the gradual deepening informatization, spectrum space is becoming more and more crowded. How to monitor and manage the electromagnetic spectrum intelligently and efficiently has become a major challenge in the development of radio technology. Existing signal recognition methods has a series of problems, such as small application range, poor generalization ability and few recognition types, which make it difficult to meet the needs of application scenarios. To address the above problems, this paper proposes a signal feature extraction recognition algorithm based on Dropout-DNN fused with wavelet and higher-order cumulant   (WH-D-DNN), which firstly extracts the spectral features of the signal by using the short-time Fourier transform, and then introduces the wavelet transform and higher-order cumulant algorithms to extract the smooth and higher-order information of the signal, and finally adopts the Dropout strategy to train the DNN neural network for signal classification and recognition. Experimental results show that the algorithm can recognize 40 communication navigation signals such as BDS, GPS, 4G, WIFI, FM and radar, with 99.4% correct rate when the SNR is higher than 10dB.
	Session:	SPG03 - 6, Paper No.3674
	Title:	Classification of Sleep Stages Using Single-Channel ECG Signal: A Comparative Analysis of Machine Learning and Deep Learning Methods
	Author:	Ahmad Ullah and Muhammad Salman Khan
	Abstract:	Abstract—In this paper, we provide an extensive evaluation of machine learning (ML) and deep learning (DL) methods for automatic sleep stage classification using a single-channel electrocardiogram (ECG) signal. To explore ML methods, we extracted 21 heart rate variability (HRV) features using the R-peak detection. These features were used to train and compare the performance of various models such as support vector machine (SVM), decision tree (DT), random forest (RF), k-nearest neighbor (K-NN), and Naïve Bayes (NB). Furthermore, we analyzed the DL methods to overcome the manual feature extraction. This method involved feeding raw time-series ECG to a 1-dimensional convolutional neural network (CNN) and short-time Fourier transform (STFT) based spectrogram to a 2D CNN. The overall evaluation was performed using 15-subject data each from two publicly available datasets. To process the data for analysis, resampling, bandpass filtering, baseline correction, 30-second interval extraction from whole-night ECG recordings, and z-score normalization techniques were applied. We conducted experiments on one dataset and then replicated the experiments on the other dataset to assess the robustness and generalizability of the methods. Through a comparative analysis, the spectrogram-based 2D CNN emerged as the effective method with the highest test accuracy and k-score of 87.65% and 0.82 respectively. This study contributes to ECG-based sleep classification research, highlighting the effectiveness of deep learning for effective classification.
	Session  SPG04
	Time - 16:00 ~ 17:30, Wednesday, August 22, 2024
	Session:	SPG04 - 1, Paper No.3680
	Title:	Simulation of Echoes from Surface Ship Based on Highlight Model
	Author:	Xin Fa, Haiyan Wang, Xiaohong Shen and Feifei Pang
	Abstract:	Abstract—The simulation of the amplitude factor of the echo for traditional highlight models varies inaccurately with azimuth angle, and the established echo highlight model only considers the scenario where the target and torpedo are at the same horizontal plane, without considering the situation where the torpedo is positioned below the target at a large elevation angle. In this paper, the boundary element method is adopted to calculate the azimuthal variation of the target strength for different elevation angles of the surface ship's overall underwater portion and three highlight parts (bow, hull, and stern). Based on this, the echo simulation of the surface ship at all azimuth angles is realized based on the highlight model, constructing a three-highlight surface ship echo model. Compared with the traditional highlight model, it establishes a highlight model of the surface ship echo at large elevation angles, which is closer to the actual scenario of torpedo attacking surface ships, and more accurately simulates the fluctuations of the amplitude factors of the target echo highlights. It better solves the problem of echo model's distortion caused by occlusion effects between highlights. From the simulation results, this model can more accurately simulate the basic characteristics of active sonar echoes, and has important theoretical significance and practical application value in the field of surface ship echo analysis.
	Session:	SPG04 - 2, Paper No.3681
	Title:	RTK-UAV Laser Surveying Applied to Monitoring of Mine Residual Walls
	Author:	Jieling Wu and Mitsugu Saito
	Abstract:	The challenges of conducting mining operations in open-pit mines, especially while attempting to maintain a stable residual wall, crucially relate to ensuring the safety of miners and preserving the landscape. Currently, many mines monitor the displacement of the pointwise residual wall using APS and GPS with light wave rangefinders, but it is difficult to capture the areal behavior of the residual wall. However, issues persist such as equipment failure and the need for timely monitoring; thus, UAVs, which allow for highly accurate, safe, and efficient surveying, are being used. This study aims to develop a regular monitoring system, using RTK-UAV laser surveying, designed to ensure the safety of the residual wall slope.
	Session:	SPG04 - 3, Paper No.3696
	Title:	Decompose, Attend and Detect: A Robust Framework for Time Series Anomaly Detection
	Author:	Junqi Chen, Xu Tan and Susanto Rahardja
	Abstract:	Deep learning-based reconstruction frameworks are widely employed in Time Series Anomaly Detection (TSAD) tasks. However, these detectors encounter challenges in generalizing to non-stationary time series data, leading to inaccurate reconstructed outputs and limiting their detection performance. To address this issue, we adopt Time Series Decomposition (TSD) methods to separately reconstruct the seasonal and trend components, as they exhibit clearer patterns for generalization. However, existing TSD methods are sensitive to anomalies, resulting in unreliable decomposed results in TSAD tasks. To bridge this gap, we introduce a novel robust auto-correlation attention mechanism into the detector. Extensive experiments conducted on both simulated and real datasets demonstrate that the proposed method outperforms all 14 competing baseline methods.
	Session:	SPG04 - 4, Paper No.3697
	Title:	Research on high-precision acoustic positioning algorithm for small-scale confined underwater space
	Author:	KaiDi Xu, AiPing Huang, YuSong Bai, JiaLe Wang and LinWei Tao
	Abstract:	In this paper, a underwater acoustic high-precision positioning system is investigated for the accurate positioning of the UUV in small confined space. The least squares Newton (LSN) algorithm based on Modified Chirp-z Transform (MCZT) with generalised quadratic correlation delay estimation is designed. The algorithm is modelled on an eight element cubic array. The MCZT algorithm produces the correlation function with more sharp peaks and higher peaks, resulting in more accurate delay estimates. The LSN algorithm performs a more accurate time-delay localization. Positioning simulations at different SNR and different locations are performed and experimental verification is conducted in the anechoic pool. Simulation results show that the LSN algorithm can achieve high accuracy positioning. Experimental results show that the UUV positioning error is controlled within 20mm, achieving high-precision positioning requirements.
	Session:	SPG04 - 5, Paper No.3707
	Title:	3S : A New Data Quality Control Algorithm Based on RTK
	Author:	Yanyan Zhao, Zikang Ji, Xiaoang Li and Hongwei Zhao
	Abstract:	Hydraulic engineering dams are affected by surrounding environmental and natural factors, leading to deformations of the dam body. These deformations are usually minor. When GNSS RTK technology is used for deformation monitoring of dams, the presence of numerous outliers and noise in the monitoring data significantly affects data quality. To solve this problem, this paper proposes the 3S algorithm, which improves the quality of deformation monitoring data.Firstly, it employs a 3σ criterion method to eliminate outliers from the data. Subsequently, the Savitzky-Golay filtering method is used to filter out the noise within the data, resulting in higher quality deformation monitoring data. Experiments show that the 3S algorithm outperforms wavelet and Kalman filtering in terms of filtering effects, making it the best option for improving the accuracy of deformation monitoring. The use of the 3S algorithm significantly enhances the precision of deformation monitoring.
	Session:	SPG04 - 6, Paper No.3712
	Title:	An Improved Probability Hypothesis Density Filter Based on Variational Bayes for Closely Spaced Multi-target Tracking
	Author:	Xingchen Jiang, Jingjie Gao, Fei Hua, Yuhai Gao, Xinqi Du and Wenqian Xu
	Abstract:	An improved probability hypothesis density filter based on variational Bayes is proposed in closely spaced multi-target tracking scenario with unknown measurement noise in a complex environment. The proposed algorithm combines adjacent Gaussian Inverse Wishart components representing the same target as much as possible, so as to minimize the number of components in the subsequent filtering iteration process, ensure the efficient operation of the multi-target tracking algorithm, and better adapt to the tracking scenarios of neighboring targets. The simulation results show that the improved algorithm can not only avoid the wrong merging of different target components, but also effectively improve the precision of target state and number estimation when the measurement noise is unknown and the target is in close proximity.
	Session  SPG05
	Time - 09:00 ~ 10:30, Thursday, August 22, 2024
	Session:	SPG05 - 1, Paper No.3713
	Title:	Wireless Tracking Algorithm with Direct Signal in Strong Doppler Environment
	Author:	Yuhai Gao, Jingjie Gao, Fei Hua, Chun Zhang, Xingchen Jiang and Peizhe Liu
	Abstract:	With the development of wireless communication technology, orthogonal frequency division multiplexing (OFDM) technology faces challenges in high mobility and high frequency communication scenarios. To deal with this problem, orthogonal time frequency space (OTFS) modulation technology arises at the historic moment. It can perform signal processing in the delayDoppler (DD) domain, convert time-varying channel into slow fading channels, and enhance the robustness of the system. In this study, a direct wireless signal in Doppler environment tracking (DWSD-Tra) method based on OTFS signal is adopted, which makes use of its stability against high speed in DD domain, reduce the intermediate steps and improve the positioning accuracy. The results show that the algorithm can reduce the tracking error in the strong Doppler environment, has excellent performance and strong robustness.
		the strong Doppler environment, has excellent performance and strong robustness.
	Session:	SPG05 - 2, Paper No.3717
	Title:	Fault Detection and Classification of DC-DC converters Based on Permutation Entropy
	Author:	Songpei Zhang, Shuqiao Zhou, Changyu Mo, Guilian Shi, Fan Chen and Xiaojin Huang
	Abstract:	In DCS systems of nuclear power plants, IO boards are widely used, but they are also the most susceptible to failure. The operational stability of the nuclear power station would be affected if an IO board fails. The SABG01X board researched in this paper is a typical type of IO board, and its core component, the DC-DC converter, is particularly susceptible to malfunctions. Therefore, figuring out a effective fault detection and classification method is necessary. In this paper, various fault diagnosis and classification methods for DC-DC converters are discussed. The permutation entropy method is adopted for trend feature extraction, with the consideration of using outputting current and voltage as the main variables and their variation tendency. Furthermore, the permutation entropy matrix distance method is introduced for fault identification and classification. The effectiveness of the proposed method is demonstrated through simulation. The results show that the permutation entropy method can be applied to not only online fault detection but also fault classification of DC-DC converters.
	Session:	SPG05 - 3, Paper No.3730
	Title:	Classification of Typical Tursiops aduncus Whistle Signals Using Convolutional Neural Networks
	Author:	Ming Xiang, Yankun Chen, Zhanwei Li, Kangrong Li, Zhuo Liu, Zhengqiao Zhao and Jie Chen
	Abstract:	Dolphins, recognized as highly intelligent marine mammals, exhibit sophisticated communication and echolocation systems. Precise classification of dolphin whistles is pivotal for comprehending their communicative behaviors and monitoring their population dynamics, including size, structure, and distribution. This study presents the development of an extensive and high-quality dataset of dolphin whistle signals, sourced from the Chimelong Ocean Kingdom. This dataset includes unique whistle types that were previously unavailable to the research community. We investigate the application of Convolutional Neural Network (CNN) models to classify the whistle signals of the Indo-Pacific bottlenose dolphin (Tursiops aduncus). Multiple CNN architectures are employed to analyze and categorize these whistle signals. The performance of these models is evaluated using mean Average Precision (mAP), demonstrating that CNN-based methodologies can effectively distinguish between different dolphin whistle signals. This work provides valuable tools for marine biologists and researchers specializing in animal acoustics, enhancing the understanding of dolphin communication. It also contributes to the conservation and management efforts of dolphin populations, offering significant insights into their behavior and ecological needs.
	Session:	SPG05 - 4, Paper No.3742
	Title:	Multi-modal Signal Prediction Method Based on Trajectory Time Series
	Author:	Qian Wei, Xinlong Ding, Yake Xuan, Zanru Chen, Zheng Zhao and Yongjin Huo
	Abstract:	As technology advances, target positioning technology is widely applied in automated systems. However, the multi-modal nature of trajectory time series makes accurately predicting dynamic target behaviors challenging. Modal decomposition can be used to process and analyze multi-modal signals, but the decomposed signals still face challenges in classification accuracy and modeling prediction. This paper innovatively proposes a multi-modal signal analysis and fusion prediction algorithm (MSAFPA) based on trajectory time series to address these issues.First, the algorithm performs a detailed analysis of the decomposed signals to accurately reveal their intrinsic characteristics and patterns. Based on these characteristics, the signals are classified accordingly. The algorithm then skillfully integrates the unique advantages of polynomial models, autoregressive moving average (ARMA) models, and support vector regression (SVR) models in time series prediction. It conducts targeted and detailed predictions for each type of modality, achieving deep fusion of multi-modal prediction results. This ensures the comprehensiveness and accuracy of the prediction outcomes.This innovative solution provides an effective approach to the challenges of predicting and fusing multi-modal signals. Comparative analyses with existing methods demonstrate the superior performance of MSAFPA in trajectory prediction.
	Session:	SPG05 - 5, Paper No.3750
	Title:	Long Distance Pipeline Leak Detection and Location Based on Center Frequency Method
	Author:	zhang guang yao, Zhou Xing yue, Zheng Hong yuan and Meng Fan tai
	Abstract:	In recent years, the condition monitoring of underground transportation pipelines has garnered significant attention from various sectors, as their condition directly impacts daily operational efficiency. This paper proposes an acoustic detection method and introduces a long-distance pipeline leak detection and localization strategy based on the center frequency method. Hydrophones are placed upstream and downstream of external valves along long-distance transportation pipelines to receive leak signals. The coherence function and cross-spectral density between upstream and downstream signals are calculated to identify similar frequency bands where the leak signals are concentrated. The signals are then pre-filtered to remove most environmental noise, and the long signals are segmented. Using the Variational Mode Decomposition (VMD) method, the signals are decomposed to a certain level where the center frequencies of each mode are close. Kurtosis is calculated for the mode components to construct high Signal-to-Noise Ratio (SNR), robust signals. Finally, cross-correlation analysis of the reconstructed signal is performed to obtain the time delay and calculate the leak location. Simulation experiments with multiple sets of measured data demonstrate that the proposed method has a minimal error percentage compared to the actual leak location（2.5%）, showcasing its practical value.
	Session:	SPG05 - 6, Paper No.3760
	Title:	Spatio-Temporal Reverberation Suppression Method Based on Itakura Distance Segmental Pre-whitening and Progressive Adaptive Binary SVD
	Author:	Yuan Cao and Qunfei Zhang
	Abstract:	In underwater acoustic signal processing, the overlapping presence of reverberation and target echoes within the time-frequency domain poses formidable challenges for target detection in active detection scenarios conducted from moving platforms. Responding to the demand for low Signal to Reverberation Ratio(SRR) in echo detection, this paper proposes a novel approach based on Itakura distance segmental pre-whitening technique and progressive adaptive binary Singular Value Decomposition (SVD) strategy, which effectively suppresses reverberation and significantly improves echo detection performance. Leveraging Itakura distance as a metric, the method evaluates and adjusts the pre-whitening process of autoregressive models, rendering it suitable for diverse reverberation environments. Furthermore, the progressive adaptive binary SVD refines signal characteristics by adapting decomposition paths to match signal properties, thus achieving efficient signal separation and target detection. Simulation experiments, incorporating factors such as multipath propagation and platform motion, validate the efficacy of the proposed method, demonstrating a 10dB increase in SRR.
	Session:	SPG05 - 7, Paper No.3761
	Title:	Enhanced Gesture Recognition through Graph-Based Multimodal Fusion
	Author:	Mobeen Ur Rehman, Talha Ilyas, Lakmal Seneviratne and Irfan Hussain
	Abstract:	This study introduces an advanced framework for recognizing hand gestures from a first-person view, leveraging the integration of multimodal data including optical flow, pose, depth, and RGB video recordings. Adeptly navigating the challenges and opportunities presented by the integration of multimodal data. At its core, the framework employs two pivotal components: a cross-attention based adaptive graph convolutional network and relational graph interactions for modality fusion. The former is designed to extract features from skeleton-based gesture data, ensuring a nuanced capture of hand movements by emphasizing the interconnections within the hand's skeletal structure. The  latter component innovatively models each output modality feature as a node in a fully connected relational graph, facilitating the fusion of heterogeneous data types through dynamic interactions between modalities. This approach allows for the leveraging of each data type's strengths and the mitigation of their weaknesses, significantly enhancing the system's classification accuracy and robustness. Tested on a public benchmark dataset, the framework achieved a remarkable accuracy of 98.48%, demonstrating its efficacy. Moreover, it proves resilient, maintaining strong performance (93.48% accuracy) even in scenarios where only one modality is available, highlighting its potential for real-world applications. This advancement sets a new benchmark in hand gesture recognition, promising future developments in multimodal data fusion.
	Session  SPG06
	Time - 11:00 ~ 12:30, Thursday, August 22, 2024
	Session:	SPG05 - 6, Paper No.3760
	Title:	A Method for Distinguishing Conical-Like Precession Targets Based on Range Profiles
	Author:	Donglin Tan and Junfeng Wang
	Abstract:	With advances in research on ballistic missile penetration, ballistic missiles have many ways to break through defenses. One of these ways is to release not only warheads but also decoys in ballistic missile midcourse. Like real warheads, some highly simulated decoys spin to control their attitude and have the same shape as the real warheads, which are all conical-like. Therefore, it is difficult to distinguish such highly simulated decoys simply by their appearance. In this paper, we propose a method to distinguish conical-like warheads and decoys of the same shape by using the different changes of range profile. Firstly, one can recognize the conical-like target by template matching. Then, the conical-like warheads and decoys can be distinguished by comparing the change in the range profiles over a period of time. Due to the mass of the real warheads is greater than the mass of the decoys, there will be differences in the precession speed of the warheads and decoys, as well as in the speed of change in the angle of sight relative to the radar. The range profile of the real warheads will change faster than the one of the decoys. The range profile changing rate of the warheads will bigger. Simulation experiments verify the effectiveness of this method.
	Session:	SPG06 - 2, Paper No.3571
	Title:	Optimization of Search Step Size for Three-Channel MEO SAR Adaptive Clutter Suppression
	Author:	Yingcong Wang, Yongkang Li and Xincheng Liang
	Abstract:	The synthetic aperture radar (SAR) adaptive clutter suppression method based on the post-Doppler space-time adaptive processing technique is distinguished by its excellent performance in improving the signal-to-clutter-noise ratio (SCNR) of a moving target. However, since the target's steering vector depends on its unknown motion parameters, it is necessary to search parameters in clutter suppression. When the search step size is small, good performance can be obtained, but the computation load would be large. In this paper, the optimization of search step size for three-channel medium-earth-orbit (MEO) SAR adaptive clutter suppression is studied. First, the target's steering vector is developed. Then, an optimization method of the parameter search step size for adaptive clutter suppression is proposed, with the allowable SCNR loss being 3 dB. Finally, experimental results validate the proposed method.
	Session:	SPG06 - 3, Paper No.3599
	Title:	A Two-Stage Magnitude-Complex Spectrum Attention Model for Self-Noise Suppression in Shipborne Sonar Platforms
	Author:	li li, Chenxi Li, Yongzhi Li and He Wu
	Abstract:	Self-noise suppression technology for shipborne sonar platforms is designed to mitigate the inherent challenge of significant self-noise interference during sonar detection. This is achieved by suppressing the self-noise generated by the sonar platform itself. Traditional underwater signal suppression algorithms often fall short in effectively suppressing platform self-noise, particularly in scenarios with low signal-to-noise ratios. To address this problem, we propose a two-stage magnitude-complex spectrum attention network model (MP-CPNet) for self-noise suppression in shipborne sonar platforms. The model comprises a magnitude processing network (MPNet) and a complex processing network (CPNet). The former part inputs the signal's magnitude and performs a coarse recovery, then coupling the output magnitude with the phase of the noise signal to obtain a coarse estimated signal. The latter part accepts the coarse estimated signal and noise spectrum as input to further estimate the spectral details, then performing a fine recovery of the signal phase and magnitude. The result estimated by the entire network is then output. Each sub-network includes multiple improved self-attention modules, which can learn and model the long-term dependencies of features in time and frequency, thereby promoting overall spectrum recovery. Experimental comparisons demonstrate that MP-CPNet outperforms traditional underwater noise suppression algorithms and other deep learning-based algorithms in terms of suppression effects.
	Session:	SPG06 - 4, Paper No.3600
	Title:	Analysis of Transient Ice Events in Underwater Ambient Noise
	Author:	Xueli Sheng, Yao Bi, Cong Liu, Mengfei Mu and Xian Xiu
	Abstract:	The increasing human activities in the Arctic region make it crucial to utilize under-ice acoustic devices for the detection and discrimination of transient events caused by these activities. This paper utilizes the noise data collected under the ice of Songhua Lake to analyze the under-ice environmental noise, confirming that it possesses acoustic characteristics similar to those found under the Arctic ice. The detection of a large number of peak events in the noise data was carried out, and an analysis was conducted on the number of peaks, noise kurtosis, peak intervals, and the frequency correlation of the power spectrum. Subsequently, feature analysis was performed on the three main types of transient events present in the data. The results indicate that the correlation between the number of low-frequency peaks and the power spectrum increases with the increase of frequency, reaching the maximum at 150Hz. Moreover, the three types of transient events exhibit distinct time-frequency and energy characteristics, which can be used to differentiate these events based on their primary features.
	Session:	SPG06 - 5, Paper No.3625
	Title:	Short-range Propagation Characteristics in an Ice-covered Lake
	Author:	Chaoran Yang, Qing Ling, Xueli Sheng, Xian Xiu, Zixiao Liu and Andreas Jakobsson
	Abstract:	This paper examines the short-range propagation characteristics of sound wave propagation in an ice covered lake experiment. The studied measurements were made in February 2024 in the shallow Song-Hua lake in northern China. The experiment illustrates how the wave propagation notably varies as a function of depth and due to changes in the water temperature, but also how the ice cover influences the sound propagation.
	Session:	SPG06 - 6, Paper No.3646
	Title:	Improved Track Initiation for Bistatic Radar on Moving Platforms Using Coupled Position and Doppler Measurements
	Author:	Zhihao Lin, Chang Gao, Qingfu Zhang, Junkun Yan, Tianyi Jia and Hongwei Liu
	Abstract:	Traditional algorithms often fail to consider or inadequately utilize Doppler information during track initiation, leading to limited performance in this regard. Moreover, existing methods mainly focus on monostatic radar, neglecting the influence of platform motion and bistatic radar configurations on track initiation algorithms. In this paper, we propose a Doppler-aided track initiation algorithm for the bistatic radar on moving platforms. The proposed algorithm considers the coupling between target position measurements and Doppler measurements, making full use of Doppler information. Additionally, to avoid the significant computational burden brought by batch processing, we employ a sequential processing strategy for track initiation. Experimental results demonstrate that compared to the traditional track initiation algorithm, the proposed algorithm achieves a higher probability of correct track initiation and effectively suppresses the initiation of false tracks.
	Session:	SPG06 - 7, Paper No.3683
	Title:	Experiments on Through Wall Radar by using an Ultra-Wide Band-FMCW Radar Trainer
	Author:	Rizka Maulina Fitri and Andriyan Bayu Suksmono
	Abstract:	A Through Wall Radar (TWR) is an important device for detecting the presence of targets hidden behind a wall, which is useful in a defense activity, or victims buried under soil or rubble after a disaster happens. In this paper, we present experiments of Ultra-Wide Band (UWB) Frequency Modulated Continuous Wave (FMCW) TWR using an FMCW Radar Trainer. The system operates in the frequency range of 2-3 GHz to achieve a desired range resolution, and which is obtained after processing beat signals and performs Fast Fourier Transform (FFT). The processing is performed in system modules that have been integrated with a microcontroller, which in turn is presented in a PPI (Plan Position Indicator). Experiments results show that a UWB-FMCW radar employed as a TWR can gives a good detecting and ranging performance.
	Session  SPG07
	Time - 14:00 ~ 15:30, Wednesday, August 21, 2024
	Session:	SPG07 - 1, Paper No.3765
	Title:	Progressive Sub-Graph Clustering Algorithm for Semi-Supervised Domain Adaptation Speaker Verification
	Author:	Zhuo Li, Jingze Lu, Zhenduo Zhao, Wenchao Wang, Mou Wang, Ziteng Wang and Xin Liu
	Abstract:	Utilizing the large-scale unlabeled data from the target domain via pseudo-label clustering algorithms is an important approach for addressing domain adaptation problems in speaker verification tasks. In this paper, we propose a novel progressive subgraph clustering algorithm based on multi-model voting and double-Gaussian based assessment (MVGPG clustering). To fully exploit the relationships among utterances and the complementarity among multiple models, our method constructs multiple knearest neighbors graphs based on diverse models and generates high-confidence edges using a voting mechanism. Further, to maximize the intra-class diversity, the connected subgraph is utilized to obtain the initial pseudo-labels. Finally, to prevent disastrous clustering results, we adopt an iterative approach that progressively increases k and employs a double-Gaussian based assessment algorithm to decide whether merging sub-classes.
	Session:	SPG07 - 2, Paper No.3676
	Title:	MTBV: Multi-Trigger Backdoor Attacks on Speaker Verification
	Author:	Mengyao Liu, Xueqing Li, Mou Wang, Xiao-Lei Zhang and Susanto Rahardja
	Abstract:	Speaker verification, a commonly used approach for authenticating speakers' identities based on their voices, was recently found to be vulnerable to backdoor attacks. Most of existing backdoor attack research is limited to injecting only a single type of trigger, which may lead to poor performance. To improve the performance of backdoor attacks, this paper proposes a backdoor attack method named Multi-Trigger Backdoor Attacks on Speaker Verification (MTBV). Different from existing single-trigger injection methods, we inject multiple types of triggers into the training data simultaneously to establish connections between multi-triggers and target attack objects. Experiments were conducted on the TIMIT and VoxCeleb2 datasets. Results show that MTBV's attack success rate increased 5.5% and 19.4% respectively over the single-trigger attack baseline. Meanwhile, we validated the impact of combination with different trigger types and different portions of poisoning rate on attack performance.
	Session:	SPG07 - 3, Paper No.3721
	Title:	Improved Adaptive Feedback Cancellation in Two Microphone Hearing Aids with a Novel Bias Reduction Technique
	Author:	Parishmita Deka and Bijit Kumar Das
	Abstract:	In hearing aids (HAs), adaptive feedback cancellation (AFC) is used widely to reduce unwanted feedback. The prediction error method-based adaptive feedback cancellation (PEM-AFC2) in a two-microphone system is a recently proposed approach in which it is considered that there is negligible presence of the second microphone feedback signal due to its distant position from the speaker. In real-life scenarios, this assumption is impractical, especially with the compact structure of hearing aid devices placing the speaker and microphones in close proximity. To address this issue which results in undesired bias in the feedback path estimate, a novel adaptive method, named improved PEM-AFC2, is introduced in this paper.
	Session:	SPG07 - 4, Paper No.3723
	Title:	A Real-time Algorithm for Selection of Reference Microphone Pair in BLCMV Beamformers in Hearing Aids
	Author:	Deep M. Baruah, Sauravjyoti Senchowa and Bijit Kumar Das
	Abstract:	The binaural linearly constrained beamformers are frequently employed in digital multi-microphone hearing aids. In this paper, a real-time algorithm for selecting the best reference microphone pair is proposed based on a study on the impact of reference microphone pair selection on binaural cue preservation for directional sources. The preservation of the binaural cues can be determined by the preservation of the Interaural Transfer Function (ITF) of the interfering signals for different reference
		microphone pair combinations. Based on realistic simulation based experiments, the performance of the proposed algorithm is assessed.
	Session:	SPG07 - 5, Paper No.3731
	Title:	Development of a High-Quality Dolphin Whistle Signal Dataset and Clustering Analysis of Sousa Chinensis Calls
	Author:	Kangrong Li, Yankun Chen, Zhuo Liu, Ming Xiang, Zhengqiao Zhao, Zhanwei Li and Jie Chen
	Abstract:	Dolphin communication signals facilitate social interactions and behaviors such as parenting and courtship. Understanding these signals is crucial for environmental protection and ecosystem monitoring including the estimation of the species abundance. This study develops a high-quality dolphin whistle signal dataset using passive acoustic monitoring techniques and employs the pYIN-based algorithm to extract pitch features from the whistle signals of the endangered \textit{Sousa chinensis} species. Clustering analysis is subsequently performed to identify the primary call types and investigate the characteristics of various call types. The findings demonstrate the quality of the collected whistle signals as well as the efficacy of the proposed algorithm. Moreover, the clustering analysis successfully detects and categorizes different whistle types without human intervention. The clustering analysis achieves an average silhouette score of 0.42, indicating that the clusters effectively group similar whistle signals while distinguishing between different call types. This study lays the groundwork for future research on dolphin communication and supports conservation strategies by facilitating more efficient and effective monitoring of dolphin populations through PAM techniques.
	Session:	SPG07 - 6, Paper No.3740
	Title:	Plug-and-play WPE guided by deep spectrum estimation for speech dereverberation
	Author:	Ziye Yang, Jie Chen, Cédric Richard and Junjie Li
	Abstract:	Speech dereverberation aims to attenuate the effects of late-reverberant components. While the plug-and-play weighted prediction error (PnPWPE) method represents an innovative approach to dereverberation with exceptional performance, it can be further enhanced from several aspects. These include improving the accuracy of power spectral density (PSD) initialization to enhance energy normalization and optimizing the process of parameter selection. To address these areas for improvement, this paper introduces an enhanced PnPWPE framework. Within this framework, PSD initialization is supported by a deep neural network, and a dynamic strategy is implemented for parameter adjustment, eliminating the need for manual selection. Experimental findings validate the efficacy of the proposed method.
	Session  SPG08
	Time - 16:00 ~ 17:30, Wednesday, August 22, 2024
	Session:	SPG08 - 1, Paper No.3601
	Title:	Fast Low-Rank Approximation of Matrices via Randomization with Application to Tensor Completion
	Author:	Maboud F. Kaloorazi, Salman Ahmadi-Asl, Jie Chen and Susanto Rahardja
	Abstract:	The approximation of voluminous datasets, which admit a low-rank structure, by ones of considerably lower ranks have recently found many practical applications in science and engineering. Randomized algorithms have emerged as an powerful choice, due to their efficacy and efficiency, particularly in exploiting parallelism in modern architectures. In this paper, we present a fast randomized rank-revealing algorithm tailored for low-rank matrix approximation and decomposition. However, unlike the previous works, which have applied deterministic decompositional algorithms such as the singular value decomposition (SVD), pivoted QR and QLP, we make use of a randomized algorithm to factorize the compressed matrix. We furnish bounds for the rank-revealing property of the proposed algorithm. In addition, we utilize our proposed algorithm to develop an efficient algorithm for the low-rank tensor decomposition, namely the tensor-SVD. We apply our proposed algorithms to various classes of multidimensional synthetic and real-world datasets.
	Session:	SPG08 - 2, Paper No.3622
	Title:	A Fast and Effective Golomb Code with Asymmetric Numerical System
	Author:	Liu Shumin, Jie Chen, Keng Pang Lim and Susanto Rahardja
	Abstract:	This paper presents the Golomb-ANS code, an advanced coding scheme that combines Golomb coding with the asymmetric numerical system (ANS). Golomb code is known for its computational simplicity, while ANS is recognized for its coding efficiency. The Golomb-ANS code capitalizes on the advantages of both methods to create an innovative encoding solution. The main breakthrough lies in utilizing Golomb code’s distinctive property, where the probability of encountering a ”0” in the suffix remains stable under a specific distribution. This characteristic simplifies the ANS component of Golomb-ANS into a finite state machine, significantly reducing coding complexity. To illustrate the practical benefits of Golomb-ANS, experiments were conducted by replacing Golomb-Rice code with GolombANS in the JPEG-LS and CCSDS 123.0.B image compression standards. The results demonstrated substantial improvements in average compression ratios, with incr
	Session:	SPG08 - 3, Paper No.3631
	Title:	Low-Light Optical Flow Estimation Method Based on Implicit Modeling of Mixed Noises
	Author:	Chuanzi Xu, Xinyu Xiang, Hong Li, Ang Li, Yi Chen, Tao Feng, Qing Luo, Yuchuan Han and Jiawen Wang
	Abstract:	This paper presents a data-driven solution to address the challenge of decreased accuracy in optical flow estimation under low-light conditions. It proposes synthesizing a dataset with multiple mixed noises to train a neural network for optical flow estimation. The neural network includes a noise removal module to enhance its performance. Experimental results demonstrate the superiority of this approach, particularly in datasets with higher levels of noise. By incorporating a noise removal module into the optical flow estimation network, the proposed method effectively mitigates the impact of noise on accuracy. Additionally, the synthesis of a dataset with mixed noises allows for robust training of the neural network, improving its performance in real-world low-light scenarios.
	Session:	SPG08 - 4, Paper No.3693
	Title:	Multi-Exposure Fusion Light Field Image Quality Assessment Based on Object Detection
	Author:	Keke Yao, Guanglong Liao, Gangyi Jiang, Yeyao Chen and Mei Yu
	Abstract:	The multi-exposure fusion algorithms can effectively improve the dynamic range of light field imaging, but distortion is inevitably produced during the process. Therefore, it is crucial to construct an effective and accurate quality assessment method for multi-exposure fusion light field images (MEFLFIs). Aiming at the significant artifact distortion phenomenon of MEFLFIs and the limitations of the traditional dynamic region extraction methods, the MEFLFI quality assessment method combined with object detection is proposed. Firstly, an object detection model is utilized to extract dynamic region masks of the original LFI sequences, and the artifact feature fusion module is further constructed to outstand artifact distortion. Subsequently, the MEFLFIs are converted from the RGB color space to the HSV color space to measure color distortion. Then, the multi-stream feature extraction network with attention mechanism is constructed to extract artifact, spatial and angular distortion features, respectively. Finally, a regression model is constructed to measure the visual quality of MEFLFIs. The experimental results show that the proposed method outperforms current advanced image quality assessment methods and maintains better consistency with human subjective perception.
	Session:	SPG08 - 5, Paper No.3727
	Title:	HYPERSPECTRAL ANOMALY DETECTION WITH CNN-BASED VAE AND RX ALGORITHM
	Author:	Linruize Tang
	Abstract:	Hyperspectral imaging (HSI) offers rich spectral and spatial information, making anomaly detection in hyperspectral images a critical and widely applied task. While combining Variational Autoencoders (VAE) with the Reed-Xiaoli (RX) algorithm has shown advantages, existing methods often rely on individual pixels without fully exploiting spatial information. In this work, we propose a CNN-based VAE framework that extracts latent representations from multi-scale data cubes, which are then analyzed by the RX algorithm to detect anomalies. This method effectively incorporates both spectral and spatial information, addressing the limitations of existing detectors. Experiments on multiple datasets demonstrate the superior accuracy and efficiency of our approach in detecting anomalies in hyperspectral images
	Session  SPG09
	Time - 09:00 ~ 10:30, Thursday, August 22, 2024
	Session:	SPG09 - 1, Paper No.3558
	Title:	Geomagnetic Matching aided Unmanned Surface Vehicle Cooperative Positioning Method
	Author:	Hanzhang Shi, chengkai tang, Lingling Zhang, yue zhe, yangyang liu and zesheng dan
	Abstract:	Due to the complex maritime environment, satellite navigation is easily interfered by weather, radio, etc., and the error of inertial navigation will accumulate rapidly with the increase of sailing time, the existing cooperative positioning methods cannot meet the accuracy and stability of Unmanned Surface Vehicle (USV) clusters to realize rescue at sea. Aiming at the above problems, this paper proposes a Geomagnetic Matching Cooperative Positioning Method for Unmanned Surface Vehicle Based on Geometric Information (GMCP-GI) for USV clusters based on geometric information. Simulation results show that the proposed g GMCP-GI for USVs can realize the cooperative positioning of USV clusters and can effectively improve the positioning accuracy and stability.
	Session:	SPG09 - 2, Paper No.3586
	Title:	Research on Balloon-based Optical Imaging System with Attitude Controller
	Author:	Yan Sun, Lu lu Qian, Min Huang, Zhan chao Wang, Guang ming Wang, Yi xin Zhao, Wen hao Zhao, Zi xuan Zhang and Xiang ning Lu
	Abstract:	In this paper, we propose a near-space balloon-based optical imaging system with attitude controller, mainly used for imaging weak emission targets such as airglow. The system includes a optical imager, a primary attitude controller system, and a secondary attitude controller system. By comparing the images obtained from two scientific flight experiments, it was found that the balloon-based optical imaging system with attitude controller can better meet the requirement of long exposure time for weak emission targets, thus providing a new application for future similar balloon-based payloads.
	Session:	SPG09 - 3, Paper No.3597
	Title:	Modeling and suppression of wind turbine clutter based on space-air bistatic radar configuration
	Author:	Shuo zhang, Shuangxi Zhang, Ning Qiao, Yongliang Wang and Qinglei Du
	Abstract:	The space-air bistatic radar system is affected by wind turbine clutter (WTC), which broadens its clutter spectrum and reduces the performance of space-time adaptive processing estimated by the covariance matrix. A clutter suppression method has been proposed to address this issue based on the space-time multiple-beam combined with the OPTICS (Ordering points to identify the clustering structure) clustering idea. This method effectively solves the problem of WTC suppression by screening WTC-containing samples and clustering data after clutter suppression. Simulation experiments have confirmed the effectiveness of this algorithm.
	Session:	SPG09 - 4, Paper No.3603
	Title:	A Lower Bound on the Second-Order RDP Function
	Author:	Ryo Nomura
	Abstract:	The rate-distortion-perception (RDP) problem is investigated in a general context. In recent years, the RDP problem has attracted significant attention from both theoretical and practical standpoints. Various researchers have explored this problem, seeking to better understand its implications and applications. Recently, Nomura [1] explored the self-random number generation (SRNG) problem using f-divergence, which serves as a subproblem within the larger RDP framework. This paper builds upon the work of [1], extending the analysis to the second-order setting and deriving the optimal rate. Furthermore, we apply our findings to the RDP problem, obtaining a lower bound for the second-order RDP function.
	Session:	SPG09 - 5, Paper No.3616
	Title:	UAV Networks Geometry Configuration Aided Cooperative Positioning Algorithm
	Author:	Zhe Song, Yi Zhang, Xingxing Zhu, Yang Yu, Zeyu Cheng and chengkai tang
	Abstract:	Due to their flexible and lightweight nature, unmanned aerial vehicles (UAVs) find extensive application in cooperative navigation and positioning technology. Additionally, wireless sensor networks (WSNs) serve as a conduit for exchanging and merging information among cooperative UAV nodes. Addressing the need for lightweight design and real-time adaptability in UAV networks, this study introduces a cooperative positioning algorithm employing manifold gradient filtering assisted by Geometric Dilution of Precision (GDOP). The algorithm leverages the measurement model among cooperative sensor nodes within the UAV network to construct a Riemannian manifold. By computing the gradient on this manifold, the algorithm identifies the steepest descent direction during iteration. Furthermore, it derives the GDOP corresponding to the geometric configuration of cooperative UAV nodes, thereby adjusting the iterative descent rate to expedite convergence. Simulation results demonstrate the algorithm's rapid convergence and high accuracy.
	Session  SPG10
	Time - 11:00 ~ 12:30, Thursday, August 22, 2024
	Session:	SPG10 - 1, Paper No.3620
	Title:	Collaborative positioning technology based on Iridium/UWB
	Author:	Yuan Zhao, Yangyang Liu, Lei Lei, Guanbo Chen, Zesheng Dan and chengkai tang
	Abstract:	Low earth orbit (LEO) satellites have low orbit altitude, strong landing signal power, and strong anti-interference ability, which is an attractive way to provide available navigation and positioning services for GNSS-challenged environments. This work studies Iridium signals and establishes a mathematical model for Iridium Doppler positioning by extracting instantaneous Doppler information from iridium opportunity signals. Then, UWB is used to perform peer-to-peer ranging measurement and communication between agent nodes. Finally, a collaborative positioning technology based on Iridium/UWB is proposed. Simulation results show that compared with iridium satellite opportunity signal positioning, the proposed method can significantly improve the convergence speed and positioning accuracy of the algorithm.
	Session:	SPG10 - 2, Paper No.3637
	Title:	An information fusion emergency positioning algorithm for UAVs based on LEO constellation
	Author:	Jiaqi Liu, Yi Zhang, chengkai tang, Yang Yu, Zhe Song and Xingxing Zhu
	Abstract:	Low-earth-orbit(LEO) satellite can ensure the acquisition of external timing and positioning for UAVs in Global Navigation Satellite System(GNSS) -denied environments. This paper proposes a multi-source information fusion positioning algorithm within the framework of Riemannian Information Geometry, which based on LEO Instantaneous Doppler positioning and the inertial navigation sensors(INS) carried by drones. The algorithm derives the covariance propagation process of LEO Doppler positioning based on the Newton-LS method, completes the modeling of probability distribution functions, and maps the probability distribution functions of heterogeneous sensors to Riemannian space to unify the navigation source information format. Utilizing the more precise information granularity of Riemannian space, the proposed algorithm solved the multi-sensor information fusion positioning problem through convex optimization. The accuracy of LEO positioning in GNSS-denied environments has been effectively improved. Compared with single LEO Instantaneous Doppler positioning and INS, the positioning accuracy has been improved by 41.9% and 32.34%, respectively.
	Session:	SPG10 - 3, Paper No.3654
	Title:	Analysis of Noise Characteristics of Pump Injection Propeller
	Author:	Shuji Zhou
	Abstract:	Abstract—Based on the feasibility of a new pump injection propeller as the research object, the optimal condition of the pump injection propeller is the initial value, using the finite element / infinite element Lighthill sound analogy theory and FW-H time domain equation. The results show that from the results of the sound pressure frequency spectrum, the stator components of the pump injection propeller show obvious line spectrum characteristics at its leaf frequency and harmonic frequency; from the nine sets of monitoring field points, the closer the stator noise, the greater the sound pressure level, the distance, the noise is the pump injection noise, which is composed of low frequency narrow band line spectrum and high frequency broadband continuous spectrum. The results have some reference effect on the noise characteristics and noise reduction design of pump injection propeller.
	Session:	SPG10 - 4, Paper No.3661
	Title:	Realizing Ultra-Low Phase Noise 100MHz Crystal Oscillator Utilizing Modified Leeson Model
	Author:	Lin Xu, Peng Ye, Feng Tan, Cheng Chen and Shuang Liao
	Abstract:	This paper presents the design and implementation of an ultra-low phase noise 100MHz crystal oscillator, along with the development and simplification of its noise theory. Specific circuit designs and noise measurements are provided. The noise theory encompasses the oscillator's primary noise sources, offering a treatment approach ranging from oscillator nonlinearity to local linearization. The design incorporates a low-noise sustaining and buffer amplifier, electrical tuning circuitry, and resonator selection. Phase noise measurements demonstrate a performance of -173dBc/Hz at 1kHz offsets, matching the highest phase noise index currently attainable (NEL's 100 MHz Reference OCXO). Furthermore, the analysis reveals that reducing dominant noise in various frequency bands away from the carrier is an effective strategy for achieving ultra-low phase noise.
	Session:	SPG10 - 5, Paper No.3703
	Title:	Research on the influence of breast sound velocity inversion based on Ultrasound Computed Tomography
	Author:	Changsheng Yang, Yan Wei, Shijiao Wu, Feiyu Gao and Ziyang Lou
	Abstract:	With the passage of time, breast cancer has overtaken lung cancer as the most common malignant tumor in the world. Its morbidity and mortality are extremely high, and more than 90% of breast cancers can be cured if detected and treated early. Nevertheless, the most common diagnostic methods for breast cancer are basically flawed, with relatively long examination time, high cost, and low diagnosis rate. To address the above problems, this paper proposes a breast sound velocity inversion algorithm based on ultrasound computed tomography, which uses the K-Wave toolbox in MATALB to model the adipose tissues, soft tissues and lesions (tumors) of the breast, and selects the SIRT or reconstructing the breast sound velocity image, and performs an error analysis on its reconstructed image. In the experimental tests, the effect of the grid size of the model and the number of distributed transducers (array elements) on the speed and size of sound in the inverse-performing internal breast tissue was analyzed.
	Session:	SPG10 - 6, Paper No.3746
	Title:	The Application of Knowledge Graph in Experimental Teaching of Electronics Majors
	Author:	Zesheng Dan, chengkai tang, Yangyang Liu, Lingling Zhang and Yuan Zhao
	Abstract:	The rapid development of modern science and technology such as big data technology and information technology is constantly driving changes and progress in the field of education, especially the application of artificial intelligence technology, which has brought new changes to traditional educational concepts, models, and systems. In the research boom of smart education, experimental education is gradually moving towards the path of intelligent development. To improve the teaching quality of online experiments and improve the guidance system of online experiments, this paper is based on electronic circuit basic course experiments. Artificial intelligence technology is applied to the guidance process of online experiments, and the platform's user data and teaching resources are used to provide accurate, standardized, and personalized online experiment guidance for experimental users. A knowledge graph based experimental knowledge recommendation model is designed, which uses collaborative filtering recommendation algorithm to analyze the historical operation records of all users on the platform, calculate the similarity between various experimental operation errors, and then fuse the semantic connotation of experimental faults in the knowledge graph. Knowledge graph based experimental fault similarity is introduced, providing users with accurate and personalized experimental guidance. This model can better predict user experimental behavior and greatly improve the accuracy of recommendations while considering both user operations and intrinsic knowledge.
	Session  SPG11
	Time - 09:00～10:30,Wednesday, August 21, 2024
	Session:	SPG11 - 1, Paper No.3546
	Title:	Developing a fast, energy-efficient carry skip adder with configurable delay extension utilizing the Han Carlson adder.
	Author:	RASUPALLI VENKATESWARA, BEERE SHYAM SUNDAR and HASANAPURAM PAVAN KUMAR
	Abstract:	This paper introduces a novel Developing a fast, energy-efficient carry skip adder with configurable delay extension utilizing the Han Carlson adder, leveraging Han Carlson adder algorithm. Traditional carry skip adders are renowned for their ability to mitigate propagation delays associated with carry signals in conventional adders [1]. The Han Carlson adder further refines this approach, improving efficiency. Moreover, the inclusion of a variable latency extension empowers the adder to dynamically adjust its operating speed to suit varying computational requirements. The primary objective of this design is to strike a balance between speed and energy efficiency, rendering it applicable across a broad spectrum of applications [2]. The paper meticulously presents and analyzes the design and implementation details of this advanced architecture. Experimental results underscore its superior performance compared to existing adder architectures, validating its efficacy and potential impact on diverse computational domains.
	Session:	SPG11 - 2, Paper No.3595
	Title:	Sampling Error Analysis of FTIR and Design of Low Noise Sampling System
	Author:	Xiangning Lu, Min Huang, Wei Han, Lulu Qian, Zhanchao Wang and Yan Sun
	Abstract:	In this paper, the sampling errors that can cause spectral noise of Fourier infrared spectrometer are introduced. The main reasons for the sampling errors are the position change of the interferometer moving mirror, the jitter in the motion process, the circuit delay and other random noise, etc. The Matlab is used to model and simulate, and the influence of these sampling errors on spectral noise is intuitively displayed. In order to reduce the sampling noise, an isochronous acquisition system through digital filter and oversampling is proposed. The sampling system does not require high precision for the driving structure of the moving mirror, and can reduce the time delay error caused by the analog filter circuit, simplify the structure and circuit design of the spectrometer, and improve the restoration quality of the infrared spectrum imaging.
	Session:	SPG11 - 3, Paper No.3630
	Title:	Open framework design for electronic virtual simulation experiments with Unity3D
	Author:	chengkai tang, songnian zhang, lingling zhang, Zesheng dan, Xunbin Zhou and Chenyu Wang
	Abstract:	With the continuous development of virtual simulation technology, the role of electronic virtual simulation experiments in the teaching of electronic disciplines and scientific research work is also increasingly important.It has become an important issue to synthesize various technologies to develop a scalable and highly reusable framework for the development of virtual simulation experiments in electronics. In this paper, we propose a Unity3D-based virtual simulation open framework experimental design for electronic major categories, and put forward a framework hierarchical architecture model that contains three layers: the foundation layer, the communication layer, and the application layer, to realize the modularity, hierarchicality, and scalability of the virtual simulation experimental framework.Finally, this paper carries out the development of virtual simulation experiments in electronics based on the framework, which verifies the feasibility and scalability of the framework.
	Session:	SPG11 - 4, Paper No.3682
	Title:	Design and Implementation of Digital Signal Processing Platform for Ultra-short Baseline Positioning System Based on PSOC
	Author:	Peiyuan Wang, Zhaotong Zhu, Yansheng Ke, Shilong Xu and Jing Wang
	Abstract:	With the gradual increase in human exploitation of the ocean, more and more underwater operations require underwater positioning to assist developers in managing underwater equipment and developing resources. Ultra-short Baseline positioning system, with its small size and ease of installation, boasts a wide range of application scenarios. However, the majority of the existing digital processing platforms for USBL positioning systems use a structure that combines discrete MCU with discrete FPGA. This structure increases the number of components, enlarges the occupied area, and introduces potential failure points. Therefore, this article proposes the design and implementation of a digital signal processing platform based on the PSOC chip. The PSOC chip integrates the CPU and FPGA into a single chip, reducing the number of components and the communication costs between different devices, as well as shortening the development cycle.This article analyzes the design of the main control module and the storage circuit module of the platform, and focuses on the design of the power supply module. It also introduces the chip selection for other modules and implements these designs on a PCB. The final tests demonstrate that the digital signal processing platform designed and implemented in this article meets the design requirements and can operate normally.
	Session:	SPG11 - 5, Paper No.3685
	Title:	Design and Implementation of Hardware Circuits for Single-Beam Imaging Sonar
	Author:	Shilong Xu, Zili Chen, Zhaotong Zhu, Peiyuan Wang and Jing Wang
	Abstract:	The ocean, as the earliest origin of life, harbors abundant water resources, mineral resources, biological resources, as well as fuel resources such as methane hydrates, oil, and natural gas. General Secretary Xi Jinping attaches great importance to the construction of a maritime power, and has delivered significant speeches and made important directives on multiple occasions regarding maritime affairs. Against this backdrop, relevant units commissioned the joint development by the author's research group of a single-beam imaging sonar device capable of domestically replacing foreign sonar products, aiming to meet the demands of underwater detection in both military and civilian fields. This paper, starting from the perspective of circuit design, establishes the overall design framework of the sonar product, introduces the circuit designs of the receiving circuit module, power supply module, stepper motor drive module, and conducts performance tests on the circuit to verify its functionality. Finally, underwater tests are conducted on the entire sonar system to validate its imaging effectiveness.
	Session  SPG12
	Time - 11:00～12:30,Wednesday, August 21, 2024
	Session:	SPG12 - 1, Paper No.3688
	Title:	Waveform Optimization of CS-MIMO Radar Based on Gradient Descent Algorithm
	Author:	Muhammad Moin Akhtar, Yong Li, Wei Cheng and Yumei Tan
	Abstract:	In the past few years, the adaptive waveform design algorithm has emerged; it improves multiple input and multiple output (MIMO) radar performance by designing the transmitting radar waveforms to adapt to the changes of targets, noise, and clutter in the environment. A MIMO radar waveform design method based on compressed sensing is proposed by using the gradient descent method. Considering the prior information of noise and interference in the environment, the compressed sensing (CS) MIMO model is designed, and the objective function is to maximize the signal-to-interference-plus-noise-ratio (SINR) of the radar. Then the method of joint optimization of transmit signal and filter is used to obtain the transmit waveform with better detection performance. Compared with the conventional MIMO radar, the optimized waveform of CS-MIMO can greatly reduce the amount of data. Computer simulation shows that the detection performance of the optimized waveform is better than that of the traditional linear frequency modulated (LFM) waveform.
	Session:	SPG12 - 2, Paper No.3741
	Title:	A NOVEL CLUTTER COVARIANCE MATRIX ESTIMATION METHOD BASED ON SPACE-TIME LOCAL SLIDING WINDOW
	Author:	Ning Qiao, Shuangxi Zhang, Shuo Zhang, Yongliang Wang and Qinglei Du
	Abstract:	The clutter suppression performance of airborne radar depends on the estimation accuracy of clutter covariance matrix. The Brennan criterion requires that the number of training samples should be greater than twice the system degrees of freedom in space time adaptive processing. In this paper, a clutter covariance matrix estimation method based on space-time local sliding window is proposed. For multi-beam locations data, joint domain localized method is used to obtain local clutter data in beam-Doppler domain. Then, the local data is interpolated to improve the signal dimension by adopting bilinear interpolation, and the training samples are obtained by the space-time local sliding window of the interpolated data along the clutter ridge. Different from traditional methods, the proposed method can make full use of the clutter information of a single range ring in beam-Doppler domain and the problem of insufficient training samples of range domain can be ameliorated. The simulation results show the effectiveness of the proposed method.
	Session:	SPG12 - 3, Paper No.3768
	Title:	Simulation of Bistatic Detection Echoes and Performance
	Author:	Kaixuan Liu, Tong Wang, Dengwen Chen and Qunfei Zhang
	Abstract:	With the improvement of submarine stealth performance, traditional single-base sonar has limited detection range and can no longer meet the needs of submarine detection. Nowadays, more submarine detections are carried out using multiple platforms in a coordinated manner, and the bistatic (multistatic) base sonar, where one platform transmits and the others receive, is a key technology. This article is based on the modeling and simulation of the output signals of the bistatic receiving array to verify the performance of the bistatic detection system. First, a simulation receiver signal model framework for bistatic detection waveform level simulation is constructed, and simulated noise is obtained by extending measured hydrophone noise of finite length. The channel impulse response is obtained through the calculation of sound rays using HJRAYD software, and the waveform of underwater target echo at the output of the bistatic base receiver array is simulated. Then, a target detection method is used on the received signal array in the time domain followed by the spatial domain to obtain the distance and bearing of the target.
	Session:	SPG12 - 4, Paper No.3549
	Title:	Linear driving system of phased array ultrasound transducer: a study on improving the accuracy and effectiveness of transcranial ultrasound stimulation therapy
	Author:	Xu Liu, Xiaoqing Zhou, Zhipeng Liu and Tao Yin
	Abstract:	This research proposes a linear driving system for phased array ultrasound transducers based on AB class power amplification structure, which is used to improve the therapeutic effect and accuracy of transcranial ultrasound stimulation (TUS). Transcranial ultrasound stimulation, as an emerging non-invasive neuromodulation technique, its effectiveness largely depends on whether ultrasound can accurately focus on specific brain regions. Although nonlinear driving systems are widely used due to their high power output, these systems often suffer from signal distortion and thermal effects, which may affect the safety and effectiveness of treatment.
		To overcome these challenges, we have designed a linear driving system based on class AB power amplifiers. Compared to nonlinear systems, this system combines the high signal fidelity of Class A amplifiers and the high efficiency of Class B amplifiers, providing more stable and clear ultrasound signals while reducing system heating and improving energy utilization efficiency. In addition, by precisely controlling the excitation timing of each component in the phased array ultrasound transducer, this system can achieve more precise focusing control, thereby improving the target accuracy and efficacy of TUS treatment.
	Session:	SPG12 - 5, Paper No.3607
	Title:	TCN-based Virtual Array Element Beamforming for Restricted Aperture Underwater Sensing Arrey
	Author:	Yaohui Lyu, Yang Jia, Yuan Xu, Yihua Liu, Yongfeng Wang and Shengyu Tang
	Abstract:	Virtual element beamforming technology is a means to enhance the array resolution of conventional beamforming algorithms. However, the linear predictor-based virtual element beamforming algorithm therein is equivalent to the early autoregressive models in time series prediction tasks, exhibiting limited capability in learning the patterns of data variations, particularly for small-scale arrays. Considering the emergence of deep learning techniques, time series prediction methods based on deep learning have gradually evolved. Therefore, this paper proposes a virtual element beamforming algorithm based on temporal convolutional network (TCN) neural network. This algorithm utilizes TCN neural network to learn the temporal characteristics between real element received signals and virtual element received signals, and enhances computational efficiency through GPU, thereby achieving efficient target detection. Simulation results demonstrate that, under small-scale arrays, compared to the linear predictor-based virtual element beamforming algorithm, the TCN-based virtual element beamforming algorithm not only improves the direction finding accuracy of target azimuth estimation but also enhances angular resolution.
	Session:	SPG12 - 6, Paper No.3673
	Title:	Beamforming under Correlated Noise Background Based on Generalized Sidelobe Cancellation Algorithm
	Author:	Yuyang Han, Yan Ma, Jian Luo and Houkun Hu
	Abstract:	Aiming at the problem that the correlation of noise affects the beamforming performance when the array element spacing is less than half wavelength, the beamforming based on generalized sidelobe cancellation algorithm is studied. Taking the uniform linear array as an example, a method of correlated noise generated by two uncorrelated noises and satisfying a certain correlation coefficient is derived. Then the noise of other array elements satisfying the correlation coefficient is recursively generated. The Conventional BeamForming ( CBF ) algorithm and Generalized Sidelobe Cancellation ( GSC ) algorithm under noise uncorrelated and noise correlated conditions in uniform linear array model is numerically simulated. The performance of beamforming under correlated noise and uncorrelated noise is compared from the aspects of main lobe width, side lobe height and array gain. The results show that the array processing gain decreases with the increase of the correlation coefficient of the correlated noise. Finally, the performance of the two algorithms is compared. The simulation results show that the GSC algorithm has better beamforming performance than the CBF algorithm under noise-related conditions.
	Session:	SPG12 - 7, Paper No.3726
	Title:	Exploring Aroma Type and Alcohol Content Classification of Chinese Liquor Using Temperature-Modulated Electronic Nose
	Author:	Danlei Chen, Yun Wang, Lihua Guo, Zhengqiao Zhao, Aowen Luo and Jingdong Chen
	Abstract:	Chinese liquor, also known as baijiu, is one of the most widely consumed and top-selling spirits globally. Researchers have developed electronic noses (e-noses) to capture digital odor character profiles of different types of Chinese liquors. However, in many baijiu brands, two critical attributes, the aroma type and alcohol content, are entangled, posing challenges in the study of nuances between aroma types. In this study, we survey e-nose signals of various baijiu aroma types at different alcohol content using temperature modulated metal oxide semiconductor gas sensors. By employing principal component analysis (PCA) and support vector machines (SVM) for dimension reduction and classification, we demonstrate that it is more challenging to distinguish between different aroma types than to differentiate common alcohol contents in terms of classification accuracy and silhouette scores. This emphasizes the importance of creating an atlas of baijiu odor character profiles to better understand aroma related features and enhance the specificity of e-nose-based brand classification. Our finding provides a novel perspective on aroma type classification that has not been thoroughly explored before, opening up new possibilities for the development of practical e-nose-based Chinese liquor classification models.
	Session  CPT 01
	Time - 14:30～16:30 Tuesday, August 20, 2024
	Session:	CPT 01 - 1, Paper No.3655
	Title:	Communicate and Control System for Planar Single-axis Photovoltaic Array Clusters in Complex Terrain
	Author:	Yue Zhao, Wenxuan Yang, Yun Wang and Luoxiao Yang
	Abstract:	In the current discourse on renewable energy, photovoltaic (PV) technology surfaces as a keystone in the sustainable energy portfolio. However, the efficiency of PV systems is often compromised by geographical and temporal challenges, specifically in undulating terrains where shadowing can markedly degrade output. Addressing this, we introduce the Dynamic Adaptive Posture Adjustment Control for Photovoltaic Cluster Systems (DAPAC-PVCS), a novel system engineered to dynamically modulate the orientation of PV panels in real-time. Through an adept fusion of hardware innovation and evolutionary computation algorithms, DAPAC-PVCS transcends traditional fixed installations by counteracting shadowing effects and optimizing energy capture. Computational analyses authenticate that this avant-garde control system significantly surmounts the efficiency benchmarks of extant models, heralding a 1.5% amplification in annual power generation. Our contribution marks a quantum leap in PV technology, poised to enhance the viability of solar farms in topographically complex regions and bolster the shift towards renewable energy paradigms.
	Session:	CPT 01 - 2, Paper No.3671
	Title:	Acoustic and Visual Information Fusion of Unmanned Surface Vehicles
	Author:	Ruoxi Hou, Xingyue Zhou and Wutao Yin
	Abstract:	In many scenarios, sensors installed in USVs can only detect either the above-water or underwater part of obstacles, which leads to errors in obstacle avoidance. In this paper, the obstacle information detected by the sonar and visual camera is used to perform cross-domain data fusion. In low-speed operation mode, the prior binding information (surface and underwater) such as coordinate and direction will be used to match the collected image information (surface and underwater) until the desired matching level is achieved, directly determining the target orientation. When the coordinate matching level falls below the threshold or when the target ID information is missing, data processing module including cluster analysis, confidence filtering, and information entropy assessment will be employed to allocate weights to heterogeneous information, such as the distance and direction of the obstacle. In the fusion model, the Monte Carlo algorithm is applied to calculate the nonsynchronous error between the obstacles detected by the sonar and the visual sensor. Corresponding angle calculations are performed, and the information such as the distance and direction is converted into latitude and longitude coordinates. Compared to the traditional approach of simply merging surface or underwater obstacle information, we have integrated the information of the same obstacle from both above and below the water surface and enhance the reliability and accuracy of decision-making for unmanned surface vehicles.
	Session:	CPT 01 - 3, Paper No.3698
	Title:	A Study on Multistatic Sonar Localization Method with Consideration of AUV Mobility
	Author:	Yi Zhang and Xiaomin Zhang
	Abstract:	In this article, a mathematical model has been developed for underwater target positioning using a dynamic multistatic sonar system based on Autonomous Underwater Vehicle (AUV) clusters. A simulation calculation model for the dynamic multistatic sonar system positioning process has also been constructed. The study focuses on the arrival time positioning algorithm of the multistatic sonar system in a moving state. Additionally, simulation calculations of dynamic multistatic sonar target positioning of three receiving stations have been carried out. The research results can provide a new technical way for underwater detection and positioning of AUV clusters.
	Session:	CPT 01 - 4, Paper No.3699
	Title:	Discussion on design method of mine fuze system
	Author:	Liu Qin, Han Peng and Chen Jiangning
	Abstract:	In order to improve the overall forward design capability of mine fuze, the digital design method of ship power system engineering is studied. The design process of mine fuze system based on Harmony MBSE is proposed, and the model system supporting the operation of the process is proposed, and the requirement analysis modeling, function modeling, system architecture analysis and synthesis of mine fuze system design are completed. It embodies the advantages of model-based design method in early verification of requirements, rapid functional simulation and data traceability in the development process of mine fuze, so as to effectively improve the efficiency of R&D design.
	Session:	CPT 01 - 5, Paper No.3743
	Title:	Optimizing Computational Load and Energy Efficiency in UAV-Based Port Surveillance System
	Author:	Hariharan Sureshkumar, Shardul Jitendra Gharat, Dhumravarna Sharad Ambre, Lavanya Prashanth Shetty, Aniruddha Dipak Kadam, Danish Anis Ansari and Gajanan Birajdar
	Abstract:	The unauthorized entry of vessels into restricted port areas poses significant security risks and regulatory challenges. Traditional surveillance methods often fall short of providing timely and comprehensive monitoring, leading to potential security breaches and operational inefficiencies. We propose a system that utilizes image stitching technology onboard the drones for enhanced mapping and object detection applications. This research addresses the issue of identifying unauthorized watercraft and vessels entering authorized port regions and notifying the appropriate authorities by transmitting real-time coordinates of the intruding vessel. Unauthorized vessel detection is carried out by using a Deep Learning Algorithm on the Micro-processor onboard each drone. Pub-Sub Model is used for fast and secure communication between the Drones and the MIS. Authorities use drones to scan the designated area simultaneously, processing images by image stitching and storing them in the drones. It will process on the drone's platform and verify the authorization of the vessel, if the unauthorized vessel is detected, then The Image along with the coordinates of the vessel is sent to the MIS via the Pub-Sub Model. Drones process this by balancing the workload over a certain period.
	Session:	CPT 01 - 6, Paper No.3759
	Title:	Development and performance analysis of a miniature plasma sound source system
	Author:	Liu Shuxun, Xiancheng Wang, Kaizhuo Lei, Qunfei Zhang, Lingling Zhang and Kaixuan Liu
	Abstract:	At present, the device of plasma sound source is huge, which is not favorable for the outfield experiments, so there is an urgent need for a plasma sound source with a small size, lightweight, and good portability. This paper presents a design scheme of a miniature plasma sound source. With the MSP430 series processor as the core, a new system is designed, which combines a control module, power supply module, high-voltage generating module, energy storage module, discharge module, trigger module, etc. Some functions, such as switch control, cycle control, and pulse coding, are realized in this system. The new system enhances the controllability of the plasma sound source while meeting the practical needs. The prototype is developed and tested, and the results show that this system performs well.
	Session  CPT 02
	Time - 9:00 ~ 10:30, Wednesday, August 21, 2024
	Session:	CPT 02 - 1, Paper No.3584
	Title:	Leveraging Deep Learning and Multimodal Signals from Social Media to Enhance Credit Risk Prediction
	Author:	Tian Gao and Raymond Yiu Keung Lau
	Abstract:	With the rise of Internet-based finance, microlending (m-lending) firms such as Prosper, Funding Circle, Welab, and so on have increasingly tapped into online social media to extract vital signals to enhance credit risk prediction. On one hand, microlending firms may not have comprehensive financial records of their Internet-based clients. On the other hand, these m-lending firms also want to significantly expand their customer bases by evaluating the credit risk of their clients out of purely traditional quantitative features and signals. However, systematic studies about the effectiveness of leveraging multimodal social media signals for online credit scoring in the context of microlending are rare. Our study just tries to fill such a research gap by proposing a deep learning-based credit scoring model which utilizes multimodal signals extracted from online social media to enhance the credit scoring processes. Based on the real-world client data provided by a listed microlending firm, our experimental results show that the proposed deep learning-based credit scoring model that leverages multimodal social media signals can significantly improve credit risk prediction by 26.07% in terms of accuracy when compared to the same model that utilizes traditional quantitative features alone.  Our research opens the door to apply deep learning and multimodal social media signals to enhance an array of Internet-based financial applications.
	Session:	CPT 02 - 2, Paper No.3618
	Title:	Intelligent Workload Partitioning of CPU-GPU Heterogeneous system
	Author:	Usama Daniyal, Partha Acharya and Kompella Ram Pranav Kasyap
	Abstract:	Computer Science is used everywhere and development in this field creates a huge impact on the world. Any type of arithmetic or non- arithmetic calculation that obeys a well-defined model is in short called computation. There are many ways to develop the field of computing, but one efficient way is increasing computational speed. Normally CPU (Central Processing Unit) is commonly used Processors responsible for execution of the computation process. The performance of CPU is bounded by some limitations such as limited cores. To overcome such limitations GPU are integrated with CPUs to obtain productive performance. Thus, heterogeneous platforms were introduced to achieve high computational performances but scheduling of task to appropriate processor is crucial factor to be defined. This paper automatically predicts the percentage of optimal partitioning of a task to be distributed between CPU and GPU using a Multi-Layered Perceptron (MLP) and based on the complexity of the blocks, it identifies and partitions the entire workload of a system as serial and parallel blocks.
	Session:	CPT 02 - 3, Paper No.3667
	Title:	Research of RSSI Indoor Ranging Algorithm Based on Sparrow Search Algorithm and BP Neural Network
	Author:	Zuo Yin, Kun Ye and Haixin Sun
	Abstract:	Traditional methods for location estimation using Received Signal Strength Indication (RSSI) rely on the log-normal shadow model to formulate the range measurement model. However, the parameter selection in this method is usually based on empirical data, which is easily affected by environmental factors, resulting in a decrease in the accuracy of ranging. In order to improve the ranging accuracy and reduce the impact of RSSI fluctuation, we proposed a new ranging method to develop a powerful ranging model using sparrow search algorithm and back propagation neural network (SSA-BP). In this method, the RSSI value of the target node is first initially normalized and then input into the SSA-BP ranging model to output the distance between the target node and the anchor node. Experimental results show that compared with the traditional BP algorithm and genetic algorithm (GA), the SSA-BP algorithm has faster convergence speed and higher ranging accuracy.
	Session:	CPT 02 - 4, Paper No.3675
	Title:	Fast Adaptation of ABR Algorithm in Meta Learning Approach
	Author:	Wangyu Choi and Jongwon Yoon
	Abstract:	Over the past years, the use of video streaming applications has surged significantly. There have been developments in adaptive bitrate (ABR) algorithms that use machine learning (ML) to enhance the quality of experience (QoE) for users. However, it remains uncertain if these algorithms maintain their effectiveness in today's intricate settings. Several meta-learning approaches have emerged, but the models still need a lot of updating to adapt to the environment. In this paper, we introduce a novel ABR algorithm designed to adapt to different environments while consistently delivering high QoE. By treating various environments as separate challenges, we manage to isolate ABR algorithm from direct environmental influences. In addition, we introduce online trainer and environment collector to further improve the adaptation ability in the online phase. We evaluation the system in a range of settings and confirmed its ability to adapt effectively to new and unforeseen environments.
	Session:	CPT 02 - 5, Paper No.3677
	Title:	Weakly-supervised anomaly detection for multimodal data distributions
	Author:	Xu Tan, Junqi Chen, Sylwan Rahardja, Jiawei Yang and Susanto Rahardja
	Abstract:	Weakly-supervised anomaly detection can outperform existing unsupervised methods with the assistance of a very small number of labeled anomalies, which attracts increasing attention from researchers. However, existing weakly-supervised anomaly detection methods are limited as these methods do not factor in the multimodel nature of the real-world data distribution. To mitigate this, we propose the Weakly-supervised Variational-mixture-model-based Anomaly Detector (WVAD). WVAD excels in multimodal datasets. It consists of two components: a deep variational mixture model, and an anomaly score estimator. The deep variational mixture model captures various features of the data from different clusters, then these features are delivered to the anomaly score estimator to assess the anomaly levels. Experimental results on three real-world datasets demonstrate WVAD's superiority.
	Session  CPT 03
	Time - 11:00 ~ 12:30, Wednesday, August 21, 2024
	Session:	CPT 03 - 1, Paper No.3530
	Title:	Attention-Guided Underground Coal Mine Pedestrian Re-identification Network
	Author:	Tian Ma, Yu D. dan, Jiayi Yang, Weilu Shi and Jiehui Zhang
	Abstract:	Pedestrian re-identification algorithms are crucial in personnel localization tasks in underground coal mines. The high similarity in attire among personnel in this environment renders general pedestrian re-identification algorithms unsuitable for personnel identification tasks in underground coal mines. This paper introduces an attention-guided feature fusion network to address the issue of poor identification accuracy arising from high similarity among personnel in coal mine scenarios. Initially, a ResNet network is utilized to extract detailed information about the target personnel. Subsequently, an attention-induced crosslevel fusion module establishes a new feature fusion branch, enhancing cross-level learning and the representation of interclass comparable subjects. Finally, contrastive global features are used to generate a powerful feature representation, reducing the difficulty of distinguishing between similar personnel. Experiments conducted on the proposed method in-house MineData dataset and the public Market-1501 dataset show that the proposed method outperforms current advanced methods, achieving mAP scores of 88.32 and 65.63, respectively.
	Session:	CPT 03 - 2, Paper No.3636
	Title:	A Novel Deep Learning Approach for Heliostat Dirt Segmentation and Severity Assessment
	Author:	Wei Zhao, Zhongtao Pan, Tao Sheng, Junfeng Zhang, Yiming Xue and Luoxiao Yang
	Abstract:	Photothermal power generation is a promising technique for converting solar radiation into electricity with high efficiency and stability. However, the performance and maintenance of photothermal power plants depend on the cleanliness and reflectivity of the heliostats. This paper introduces an innovative approach to addressing the challenges of dirt detection and segmentation on heliostats. Leveraging the capabilities of deep learning, we propose the Multi-Scale Heliostat Dirt Segmentation and Classification (MSHDSC) framework, integrating a novel multi-scale feature fusion module (MSFFM) with an enhanced DeepLabV3+ network. This framework effectively segments small dirt areas on heliostat images, facilitating precise cleaning strategies. A unique aspect of our work is the introduction of an unsupervised clustering algorithm post-segmentation, which categorizes dirt based on color and texture, assigning a severity score to each category. This categorization assists in determining the cleaning complexity and prioritizing maintenance efforts. Experimental results show that our method outperforms several state-of-the-art image segmentation models in terms of accuracy and efficiency and provides useful information for targeted and prioritized cleaning of heliostats by robots or drones.
	Session:	CPT 03 - 3, Paper No.3692
	Title:	Mapping EEG Signals to Visual Stimuli: A Deep Learning Approach to Match vs Mismatch Classification
	Author:	Yiqian Yang, Zhengqiao Zhao, Qian Wang, Yan Yang and Jingdong Chen
	Abstract:	Existing approaches to modeling associations between visual stimuli and brain responses are facing difficulties in handling between-subject variance and model generalization. Inspired by the recent progress in modeling speech-brain response, we propose a "match-vs-mismatch" deep learning model in this study to classify whether a video clip elicits neural responses in recorded EEG signals. Our model employs dilated convolutional neural networks and gated recurrent units to extract features from both EEG and video signals, enabling the learning of associations between visual content and corresponding neural recordings. We demonstrate that our proposed model achieves the highest accuracy on unseen subjects compared to other baseline models. Additionally, we assess inter-subject noise using a subject-level silhouette score in the embedding space, revealing that our model effectively mitigates inter-subject noise and significantly reduces the silhouette score. Furthermore, we investigate Grad-CAM activation scores, revealing that brain regions linked to language processing contribute most to model predictions, followed by regions associated with visual processing. These findings hold promise for advancing neural recording-based video reconstruction and related applications.
	Session:	CPT 03 - 4, Paper No.3737
	Title:	Underwater Small Object Detection Based on FEA-YOLO
	Author:	Jiale Zhang, Zhuofan He, Yuhang Han, Qunfei Zhang and Xiaodong Cui
	Abstract:	Although side-scan sonar can provide wide and accurate views of submarine terrain and objects, it severely suffers from complex environmental noise, uneven illumination conditions, and small scale of the targets, resulting in a high misdetection rate. Here, this paper presents a Fine-grained feature Enhancement module together with an improved Attentional YOLOv9 model (FEA-YOLO) to tackle the challenges. FEA-YOLO first utilizes SwinIR to design a high resolution image enhancement process to extract fine-grained features for tiny objects. Next, it improves the recent released YOLOv9 model by a specifically constructed convolutional block attention module to instruct the model to focus on region of interests. Finally, the model is verified on a public side-scan sonar image dataset. Experiment results indicate that FEA-YOLO achieves precision and recall at 88.9% and 85.7%, respectively, and obtains mAP at 87.4%, surpassing the current state-of-art object detection methods as well as the original YOLOv9. Therefore, the FEA-YOLO model provides a new and effective approach for small underwater object detection in side-scan sonar images.
	Session:	CPT 03 - 5, Paper No.3756
	Title:	Wave prediction based on long short-term memory neural network
	Author:	Yuan Zhang, Fantai Meng, Hongyuan Zheng, Bingyong Guo and Siya Jin
	Abstract:	Wave prediction is of great importance in the field of wave energy. Machine learning provides an efficient new idea for wave prediction Due to its strong adaptability and excellent ability to deal with complex data. In this work, the long short-term memory (LSTM) model is investigated under four typical wave cases, including two extreme waves hitting the 50-year contour line of European Marine Energy Centre (EMEC) site, UK with the largest significant wave height and two different wave spectral shapes; another two wave conditions at the targeted resonance condition of a designed wave energy converter (WEC) with two different wave spectral shapes. Physical wave data recorded from wave tank are used to train a LSTM model. Validated with the physical data, the results show that the developed LSTM model has great performance for the prediction of the different waves studied in this work.
	Session:	CPT 03 - 6, Paper No.3757
	Title:	An Improved YOLOv8 Network for Underwater Fisheye Image Object Detection
	Author:	Hongyu Wei, Jianfeng Chen, Ju He, Jisheng Bai and Han Yin
	Abstract:	Abstract—Accurately detecting underwater objects is of great significance in fields such as marine exploration. The unique imaging principle of fisheye lenses allows for a wider field of view without compromising resolution. Although object detection technology has become relatively mature in terre-strial environments, the complex underwater environment, combined with the severe distortion of image edge regions caused by fisheye lenses, presents significant challenges for underwater object detection. In response to these issues, this paper proposed an improved YOLOv8 network architecture. A CA (Channel Attention) mechanism is added to the Bac-kbone part of YOLOv8, and a deformable convolution DCNv3 is introduced into the Bottleneck module of the C2f layer in YOLOv8. Experimental results on publicly available datasets demonstrate that the proposed method achieves higher accu-racy in fisheye image object detection compared to other methods, indicating that the improvements made can enhance the feature extraction capabilities of the network structure for regions with severe fisheye image distortion. Additionally, ablation experiments are conducted to further validate the effectiveness of the proposed method.
	Session  COM 01
	Time - 14:30～16:30 Tuesday, August 20, 2024
	Session:	COM 01 - 1, Paper No.3583
	Title:	Transmission scheme based on collaborative AUV communication
	Author:	jia hao sun, Chao Qi, Zeng Li Liu and Xuan Zhi Zhao
	Abstract:	Underwater wireless sensor networks can monitor ocean information, which provides a new approach to marine environmental monitoring, disaster warning and resource exploration. However, the development of underwater wireless sensing networks is limited by factors such as ocean currents and limited energy of the sensors themselves. Based on this situation, this paper proposes a transmission scheme based on collaborative communication with autonomous underwater vehicles (AUVs), in which autonomous underwater vehicles (AUVs) are used as a tool for distributed data collection and forwarding, which can solve the problem of imbalanced energy consumption of sensor nodes in the traditional multi-hop underwater communication network. At the same time, we consider the influence of ocean currents on the node positions to maximize the imitation of the real underwater environment, and the simulation results show that the proposed transmission scheme for collaborative communication with AUVs can significantly prolong the service life of underwater sensor networks.
	Session:	COM 01 - 2, Paper No.3606
	Title:	Field Experiments of OTFS Based Underwater Acoustic Communication in Shallow Water
	Author:	Guanhui Li, Fangjiong Chen, Xing Zhang, Hua Yu and Lijun Xu
	Abstract:	The rate-distortion-perception (RDP) problem is investigated in a general context. In recent years, the RDP problem has attracted significant attention from both theoretical and practical standpoints. Various researchers have explored this problem, seeking to better understand its implications and applications. Recently, Nomura [1] explored the self-random number generation (SRNG) problem using f-divergence, which serves as a subproblem within the larger RDP framework. This paper builds upon the work of [1], extending the analysis to the second-order setting and deriving the optimal rate. Furthermore, we apply our findings to the RDP problem, obtaining a lower bound for the second-order RDP function. Index Terms—general source, f-divergence, R
	Session:	COM 01 - 3, Paper No.3642
	Title:	Digital Design Method of Target Judgement Model for Underwater Active Acoustic Detection System
	Author:	Lu Kang, Xiaomin Zhang and Luo Jian
	Abstract:	The judgment model plays a crucial role in the digital design of underwater active acoustic detection systems. This paper analyzes the signal characteristics processed by hydrophones and analog signals, and proposes a method for determining the presence of a target. Utilizing the Matlab-Simulink platform, we adopt a digital model design approach. Virtual verification of the model demonstrates its effectiveness in accurately identifying target occurrences.
	Session:	COM 01 - 4, Paper No.3648
	Title:	Impulse-Mitigation Kalman Filter for Underwater Acoustic Communication
	Author:	Xinran Cao, Lijun Xu and Qingqing Zhao
	Abstract:	Due to the complex underwater environment, underwater acoustic (UWA) communication system is one of the most challenging systems in wireless communication. The existence of interference is the major detrimental factor which constrains the performance of UWA communication. Impulse noise, introduced by nature sources and human activities, is the common interference in UWA communication. In order to mitigate the impulse noise, in this paper, we investigate an impulse-mitigation (IM) Kalman filter for UWA system. Different from the conventional Kalman filter, IM Kalman filter considered the effect of impulse noise, reduced the proportion of measurement affected by impulse noise to the final estimated state, and calculated the suitable weight for Kalman gain, which can mitigate the impact of impulse noise. Simulation results showed that the proposed method can effectively reduce the bit error rate and improve the reliability of the communication system compared with the existing impulse noise mitigation algorithms.
	Session:	COM 01 - 5, Paper No.3702
	Title:	Underwater Unmanned Vehicle Clusters Cooperative positioning method
	Author:	Shijiao Wu, Lingling Zhang, chengkai tang and Yimeng Gao
	Abstract:	Nowadays, UUV clusters have become the main carriers for ocean environment monitoring. In this application, the positioning of UUV nodes during voyages is of great importance. A cooperative positioning method for UUV clusters is presented in this paper. It utilizes factor graphs to fuse ranging information from each UUV node, and predicts movement path based on multiple confocal positioning models. The comparison results demonstrate that the proposed method achieves higher positioning accuracy, faster convergence speed, and good suppression capability for mutation errors.
	Session  COM 02
	Time - 9:00～10:30,Wednesday, August 21, 2024
	Session:	COM 02 - 1, Paper No.3715
	Title:	Constellation Optimization Based on DDPG for Cayley Differential Unitary Space-Time Code
	Author:	Yuewen Diao, Lei Wan, Yougan Chen and En Cheng
	Abstract:	In multiple-input multiple-output (MIMO) systems, differential space-time coding is an effective transmission scheme in which channel knowledge is not required at the receiver. Cayley differential unitary space-time code has the advantages of high transmission rate, suitable for arbitrary antenna sizes as well as easy to encode and decode. To find the optimal constellation for Cayley differential unitary space-time code, the general exhaustive search method is computationally intensive and slow to converge. In this paper, a method based on deep determined policy gradient (DDPG) is proposed to search for the optimal constellation arrangement of the mapping parameters efficiently for Cayley differential unitary space-time code. Simulation results demonstrate the effectiveness of the proposed scheme based on DDPG.
	Session:	COM 02 - 2, Paper No.3718
	Title:	Bit-Loading for Non-Coherent MFSK Underwater Acoustic Communication Systems with Uneven Transmission Source Level
	Author:	Xiao Lin, Lei Wan, Yougan Chen and En Cheng
	Abstract:	Non-coherent modulation is a prevalent technique in underwater acoustic (UWA) communications due to its robust performance and low complexity. However, the transmission source level of UWA communication systems is often uneven due to transducer characteristics, and the difficulty of matching between power amplifiers and transducers, which can influence the efficacy of non-coherent modulation. In this paper, a bit-loading and power allocation scheme is proposed for the non-coherent M-ary frequency shift keying (MFSK) UWA communication systems with uneven transmission source level. The algorithm adjusts the bit and power allocation on frequency groups according to the corresponding frequency gain, with the objective of improving the spectral efficiency and reliability. Simulation results show that the proposed algorithm outperforms the traditional even bit and power allocation, with a significant improvement in the effective throughput and spectral efficiency of the system.
	Session:	COM 02 - 3, Paper No.3736
	Title:	A WIDE BAND REFLECTIVE LINEAR TO CIRCULAR POLARIZATION CONVERTING METASURFACE
	Author:	Babar Kamal, Babar Khan, Mujahid Ali Shah, Jingdong Chen, Sadiq Ullah, Yin Yingzeng and Jian Ren
	Abstract:	This paper presents a novel, compact single-layer reflective metasurface designed for wideband linear-to-circular (LTC) polarization conversion. Simulations are carried and the results demonstrate that the proposed metasurface can achieve LTC polarization conversion spanning from 12 GHz to 26 GHz, with a bandwidth (BW) of 14 GHz and a fractional bandwidth (FBW, which is defined as the ratio between the maximum, minimum and central frequency within the operating band) of 74%. The axial ratio (AR) of the metasurface remains below 3 dB within the FBW. Furthermore, the polarization conversion ratio (PCR) of the LTC polarization converter reaches 99% within the specified operating band.
	Session:	COM 02 - 4, Paper No.3752
	Title:	Adaptive Decision Feedback Equalization for Underwater Acoustic OTFS System with Superposition Pilot
	Author:	Xuewei Zhang, Lianyou Jing, Wentao Shi, Xingliang Zhang and Chengbing He
	Abstract:	Orthogonal time-frequency-space (OTFS) modulation transforms the time-varying channel into the delay-Doppler domain, and the fading coefficient of the channel in the delay-Doppler domain is near constant. OTFS modulation realizes the quantized representation of the channel in the delay dimension and the Doppler dimension, and the system has a good performance of anti-multipath effect and anti-Doppler expansion. Therefore, we extend the two-dimensional decision feedback equalization (2D-DFE) to the OTFS system and propose a 2D adaptive multichannel DFE for the OTFS system with a superposition pilot (SP). Based on the superior performance of the algorithm, this paper applies the method to underwater acoustic (UWA) communications to deal with UWA channels with large delay and large Doppler. At the same time, the sparse nature of the delay-Doppler domain channel is exploited to achieve fast convergence and fast-tracking using the normalized least mean square (NLMS) algorithm. Simulation results show that the proposed equalizer could achieve more satisfactory BER performance and improve band utilization.
	Session:	COM 02 - 5, Paper No.3758
	Title:	ADAPTIVE RANDOM FOURIER FEATURES GAUSSIAN KERNEL NORMALIZED LMS ALGORITHM
	Author:	Wentao Shi, Mingqi Jin, Yuhao Qiu, Wei Gao, Lihan Zheng and Lianyou Jing
	Abstract:	In this paper,we propose an adaptive stochastic Fourier feature Gaussian kernel normalized LMS (ARFF-GKNLMS). Similar to many kernel adaptive filtering algorithms that use stochastic gradient descent, the ARFF-GKNLMS algorithm uses more flexible stochastic Fourier features to reduce computation intensity. The difference is that the algorithm can adjust the inherent core bandwidth online, thereby solving the problem of selecting the core bandwidth in advance to a certain extent. In addition, the ARFF-GKNLMS algorithm has fast convergence performance, low steady-state error and good tracking ability, especially in non-stationary environment with low signal-to-noise ratio or strong noise, which has good robustness and tracking performance. The simulation results show that compared with other kernel adaptive filters with preset core bandwidths, the performance of this method is significantly improved in terms of convergence speed, steady-state error and tracking ability in both transient and steady state.
	Session  COM 03
	Time - 11:00～12:30,Wednesday, August 21, 2024
	Session:	COM 03 - 1, Paper No.3554
	Title:	Adaptive accurate localization based on ranging for unknown indoor environment
	Author:	Bo Gao, Baowang Lian and chengkai tang
	Abstract:	The positioning accuracy and range of ultra-wideband (UWB) are restricted by the number and placement of base stations, and we need to install base stations in advance and measure their precise locations. How to obtain high positioning accuracy without the limitation of traditional usage is a hot issue in research. To solve this problem, a novel range/visual/inertial fusion localization algorithm for unknown indoor environment has been proposed. First, we extract point and line features from images and fuse them with inertial measurement unit (IMU) measurements. Second, we arbitrarily place multiple static base stations in the environment so that the base station can estimate its relative position when visible, and at the same time, the base station can form a ranging constraint on visual inertial odometry (VIO) to correct its cumulative error. We set filters and synchronization mechanisms for UWB range measurements and fuse them with VIO in a tightly coupled manner. The proposed algorithm does not have the UWB base station position as a prior value, and the local configuration can adapt to any number of base stations. Finally, we conduct experiments in different indoor environments, and the results show that the proposed algorithm has better performance.
	Session:	COM 03 - 2, Paper No.3668
	Title:	Energy Efficient Multi-User Computation Offloading Based on Throughput Prediction
	Author:	Zhengtao Liao, Jianxian Lu, Xuewei Huang, Xiaocong Zou, Zhiwei He, Meng Zhang and Fangjiong Chen
	Abstract:	Mobile Edge Computing(MEC) pushes computation and data processing capabilities toward the network edge to meet the demands of fast and low-latency applications. With the increasing interest in mobile applications such as vehicular networks, channel prediction-based edge computing become a crucial task due to the time-varying channels. This paper proposes a multi-user, multi-task computation offloading model with uplink throughput prediction. First, the proposed scheme uses an exponential smoothing model to predict future Reference Signal Received Power (RSRP) accurately. Based on the collected RSRP and uplink throughput datasets in real network environments, machine learning is employed to train prediction models for RSRP and uplink throughput. Second, based on the predicted uplink throughput, we then formulate the offloading slot selection as an integer linear programming problem based on minimizing energy consumption. Since this problem is considered NP-hard, the paper proposes a polynomial time complexity computation offloading algorithm. Simulation results show that the algorithm can approximately achieve the minimum energy consumption and has a more efficient solution rate than branch and bound (B&B) algorithm.
	Session:	COM 03 - 3, Paper No.3691
	Title:	Statistical Studies of Fading in Underwater Wireless Optical Channels in the Presence of Bubbles
	Author:	Zhe Jiang, Junbo Zhang and Haiyan Wang
	Abstract:	Bubbles, as one of the significant factors influencing turbulence distribution, profoundly affect the performance of underwater wireless optical communication. This paper aims to investigate the impact of bubbles on turbulence distribution through experimental means and to fit them using turbulence distribution models. To further investigate the mechanism of bubble influence, the experiment specifically considers bubble size and air flow rate as variables, revealing the impact of different bubble sizes and airflow rates on turbulence distribution. Where large-sized bubbles exhibit a two-lobe distribution, small-sized bubbles display a simple single-lobe distribution at low air flow rates and a two-lobe distribution at high airflow rates. It reveals the fitting performance of turbulence models. The fitting performance of the Weibull-Generalized Gamma Distribution model is superior, making it more suitable for general scenarios. Provide essential insights for analyzing and improving performance in subsequent underwater optical communication systems.
	Session:	COM 03 - 4, Paper No.3770
	Title:	Inter-Vector Interference Cancellation for Orthogonal Signal Division Multiplexing in Doubly-Selective Channels
	Author:	Shengqian Ma, Yujie Wang, Xinyu Wu, Lingling Zhang and Jing Han
	Abstract:	Orthogonal Signal-Division Multiplexing (OSDM) is a generalized scheme that bridges the gap between orthogonal frequency-division multiplexing and single-carrier modulation with frequency-domain equalization. However, OSDM systems suffer significant performance degradeation due to inter-vector interference (IVI) in doubly-selective channels. To mitigate this issue, this paper proposes two IVI cancellation techniques tailored for OSDM systems. The one-off IVI cancellation only targets the current demodulated vector, whereas the combined IVI cancellation considers all demodulated vectors associated with the current symbol vector. Importantly, both IVI cancellation methods possess low-complexity equalization realization. Simulation results validate the effectiveness of these IVI cancellation schemes in enhancing system performance.
	Session  COM 04
	Time - 14:00～15:30,Wednesday, August 21, 2024
	Session:	COM 04 - 1, Paper No.3666
	Title:	A Low-Complexity Orthogonal Matching Pursuit Algorithm based on Multi-Scale Multi-Lag Underwater Acoustic Channels
	Author:	Du jia qi, Yongsheng Yan and Li Xiang xiang
	Abstract:	This paper studies the estimation of parameter for wideband underwater acoustic (UWA) channel with multi-scale and multi-lag (MSML) characteristics, within the framework of the OFDM system. Based on the Orthogonal Matching Pursuit (OMP) reconstruction model, the high computational complexity issue caused by over-parameterized dictionaries during channel estimation is addressed. A low-complexity OMP algorithm based on independent stepwise search is proposed. This algorithm leverages the MSML characteristics of UWA channels to concentrates the computation cost waste problem of inner in critical regions, thereby solving the computational waste problem of the OMP algorithm, which requires repeated inner product calculations in each iteration. This provides a low-complexity, high-performance channel estimation solution for underwater OFDM communication.
	Session:	COM 04 - 2, Paper No.3695
	Title:	Modulation technology in cross-domain magnetic induction communication
	Author:	RX C, xin zhang, xiaoming qi, xiaoji zhang and yizhou ge
	Abstract:	Abstract—The research on technology of cross-domain communication, which spans air and sea domains, has been widely concerned. As an emerging cross-domain communication technology, magnetic induction (MI) communication boasts unique advantages such as stable channels, minimal susceptibility to the dynamic ocean environment, and low propagation delay. However, it also has limitations in terms of transmission distance×data rate. In this paper, the modulation technology in cross-domain magnetic induction communication is studied, and the very minimum shift keying (VMSK/2) coding modulation scheme based on direct sequence spread spectrum (DSSS) technology is proposed. VMSK modulation is used to adapt to the severely bandwidth-limited underwater magnetic induction channel, and DSSS technology is used to enhance the detection ability of weak magnetic induction signal, thereby increasing the water entry depth of cross-domain communication. In this paper, the scheme of VMSK/2-DSSS code modulation is given, and the simulation analysis and field trial are carried out. The simulation and lake test results show that compared to magnetic induction communication without DSSS, the VMSK/2-DSSS coding modulation scheme can achieve a greater water entry depth, enabling error-free cross-domain magnetic induction communication at a depth of 40 meters and a data rate of 100bps.
	Session:	COM 04 - 3, Paper No.3722
	Title:	Adaptive Equalization with Interference Reconstruction and Cancellation based on MSER  Criterion for OTFS System
	Author:	Tonghui Zheng, Chengbing He, Lianyou Jing, Xin Dong, Xinyu Cao and Qiankun Yan
	Abstract:	Orthogonal time-frequency space (OTFS) modulation has obtain much attentions due to its capability for providing reliable transmission compared to orthogonal frequency division multiplexing (OFDM), especially in dynamic scenarios. In this paper, we propose an adaptive decision feedback equalizer with interference reconstruction and cancellation (ADFE-IRC) based on minimal symbol error rate (MSER) criterion for OTFS systems. Firstly, the proposed method performs adaptive filtering in the delay dimension to eliminate inter-symbol interference (ISI). Then the performance loss caused by cyclic convolution and phase flipping is improved by interference cancellation and reconfiguration. Finally, the Doppler dimension filtered signal is used as the final output of the equalizer. Doppler dimensional interference elimination is finally performed. The simulation results demonstrate the error performance of the proposed method.
	Session:	COM 04 - 4, Paper No.3769
	Title:	Strategic Improvement of Radar Accuracy by Second Order Differential MTI Methods
	Author:	Chenyu Wang, chengkai tang and Lingling Zhang
	Abstract:	To achieve better measurement of the range distance and velocity in radar system, this paper presents a range measurement method utilizing the second-order differential Moving Target Indication (MTI) method, based on Linear FM pulse compression technique. By exploring the phase or timing differences between three consecutive echoes, the background noise of the low speed moving target is filtered to enhance the radar's precision. A new echo processing model has been developed, alongside refined data processing algorithms, effectively distinguishing target objects from static backgrounds and improving overall measurement accuracy. Simulation results demonstrate the robust performance in resisting signal interference and clutter.
	Session:	COM 04 - 5, Paper No.3733
	Title:	Uplink Performance of Scalable Cell-Free Massive MIMO With Low-Resolution ADCs
	Author:	Ruonan Wang, Hui Li and Heng Zhang
	Abstract:	This paper investigates the impact of low-resolution analog-to-digital converters (ADCs) on scalable cell-free massive MIMO systems. The ADC quantization distortion is modeled using the additive quantization noise model (AQNM) to derive the minimum mean square error (MMSE) channel estimate. The results show that low-resolution ADCs and pilot contamination lead to error floor of NMSE. Then, to achieve system scalability, a scalable local partial MMSE (LP-MMSE) combining and scalable near-optimal LSFD (no-LSFD) strategies are obtained. The simulation results show that the spectral efficiency (SE) performance of these two schemes is essentially the same as that of the unscalable optimal combining vectors and the optimal LSFD. In addition, we investigate the trade-off between spectral efficiency (SE) and energy efficiency (EE), as well as the effects of the number of quantization bits and the number of antennas at the APs on EE. The resulting observations show that a good tradeoff between SE and EE can be achieved by choosing an appropriate quantization bit.
	Session  COM 05
	Time - 16:00 ~ 17:30, Thursday, August 22, 2024
	Session:	COM 05 - 1, Paper No.3559
	Title:	UAV Communication and Navigation Signals Jamming Methods
	Author:	jiawei ding, Chengkai Tang, Lingling Zhang, zhe yue, yangyang liu and zesheng dan
	Abstract:	The "black flight" problem is seriously threatening the security of critical areas, and developing effective countermeasures is an urgent need in both military and civilian domains. However, the anti-jamming of unmanned aerial vehicles (UAVs) continues to evolve, and traditional countermeasure systems face challenges such as low jamming success rates and poor flexibility, making it difficult to meet the demands of defense and confrontation. To address these issues, this paper proposes a jamming method for UAV communication and navigation signals. Based on the YunSDR-Y550 software-defined radio platform, we developed the UAV_ATK upper computer interface software to communicate in real-time with the hardware platform, adjusting power, frequency bands, and bandwidth. We generate three modes of signal disruption—broadband jamming, comb jamming, and frequency sweeping jamming—to disrupt the UAV's telemetry link. We also disrupt the UAV's navigation link by jamming GPS and BDS navigation signals. By mimicking real GPS signals and generating false navigation messages, we gradually lead the receiver tracking loop to lock onto the deceptive signal, achieving intrusive deception of the UAV. Experimental results validate the effectiveness and correctness of the proposed method.
	Session:	COM 05 - 2, Paper No.3634
	Title:	Simulation and Analysis of the PMF-FFT Pseudocode Acquisition Algorithm
	Author:	Donghang Chai, Hong Liu, chengkai tang, Xiaochen Bai, Han Xie and Xianling Wang Wang
	Abstract:	In the process of low orbit satellite navigation search and rescue, there is significant relative motion between the satellite and the ground, resulting in substantial Doppler frequency shifts. Spread spectrum communication technology is introduced to address this type of scenario. The PMF-FFT (Partial Match Filtering Fast Fourier Transform) algorithm is a well-established signal acquisition algorithm in spread spectrum systems. To mitigate the correlation loss in the PMF part of this algorithm, we introduce a new windowing function. Additionally, for the sector loss generated in the FFT part, we introduce another window to improve the situation. Simulation results show that the introduction of these two window functions has a significant effect on mitigating losses and compensating for attenuation in this algorithm.
	Session:	COM 05 - 3, Paper No.3652
	Title:	Multi-dimensional Fusion Quality Evaluation Method of Radio Signal Based on Aviation Search and Rescue
	Author:	Yunna Luo, Chengkai Tang, Lingling Zhang, Zhe Yue, Cunle Zhang and Zesheng Dan
	Abstract:	When improving the reliability and accuracy of signals in aviation search and rescue system, the signal quality evaluation has a great significance of the service quality of the navigation search and rescue system. This paper integrates various evaluation indicators to evaluate radio signal's characteristics including the time-domain, frequency-domain, modulation-domain, correlation-domain, consistent-domain and signal accuracy. Further through the correlation between dimensions, the 2nd level radio signal multi-dimensional fusion quality evaluation model (MFEM) was established, and the corresponding rapid evaluation method was proposed to realize the independent assessment of each domain and overall signal quality assessment. After several sets of data tested, it indicates that the MFEM can show fleetly the dynamic change of the signal quality, and the evaluation value performs higher reliability and accuracy. Moreover, it also provides a good reference for the establishment of the signal quality assessment system for others.
	Session:	COM 05 - 4, Paper No.3739
	Title:	A Distributed Interference Source Forwarding Deception Jamming Method Based on Spatial Domain
	Author:	Guibin Wan, Wang Yao, Zhi Zeng, Haobo Li, Yan Gao and Xiao Ma
	Abstract:	In recent years, due to the rapid development of UAV, numerous UAV black flying incidents have occurred, bringing many threats such as security and privacy infringement. Therefore, it is imperative to develop anti-UAV  measures. This paper proposes a distributed interference source forwarding deception jamming method (DISJM) based on spatial domain to address the problems of strict limitations on the interference location of traditional single interference sources, easy detection by signal arrival angle detection modules, and poor interference concealment. This method eliminates strict limitations on the location of interference sources by placing them on concentric circles with inverted cones as the vertices in important places such as airports and large-scale events, and suspending them in the stratosphere, allowing for flexible station deployment. And by identifying and locating the target UAV, utilizing the location information of distributed interference sources, the airspace is divided to achieve the forwarding of complete constellation deception signals. The simulation results show that the distributed forwarding deception interference method based on spatial domain can accurately achieve delay control, achieve multi-domain fusion of forwarding deception signals in time-frequency domain, power domain, and spatial domain, successfully deceive the target UAV to the preset position, protect important areas, and effectively improve the concealment and interference success rate of deception interference signals.
	Session:	COM 05 - 5, Paper No.3751
	Title:	Multi-hop Routing based on LEACH in WSN
	Author:	Qian Wei and Zanru Chen
	Abstract:	In wireless sensor networks, if the cluster head node is too far away from the sink node, it will generate large transmission energy consumption, leading to premature death of the node and reducing the overall network life. In this paper,based on the LEACH cluster routing algorithm, an inter-cluster multi-hop network energy optimization algorithm is proposed-MRLW (Multi-hop Routing based on LEACH in WSN). After dividing the clusters, the node performance evaluation function is constructed through the energy consumption and residual energy of the cluster head nodes; then, the optimal path finding system is constructed; finally, the optimal transmission path of each cluster head node is calculated using convex optimization, which achieves the effect of decreasing the energy consumption of data transmission and prolonging the life of the network. Simulation results show that the proposed MRLW algorithm reduces energy consumption and extends the sensor network's lifetime.

 

Conference Secretary

Lingling Zhang, Northwestern Polytechnical Univ.

Si Zhao, Northwestern Polytechnical Univ.

Contact email: icspcc@163.com

Corresponding Address:

School of Marine Science and Technology
Northwestern Polytechnical University
Xi'an, Shaanxi, P.R.China, 710072