Shriram Tallam Puranam Raghu, Dawn T. MacIsaac, Erik J. Scheme
In this study, we investigate the application of self-supervised learning via pre-trained Long Short-Term Memory (LSTM) networks for training surface electromyography pattern recognition models (sEMG-PR) using dynamic data with transitions. While labeling such data poses challenges due to the absence of ground-truth labels during transitions between classes, self-supervised pre-training offers a way to circumvent this issue. We compare the performance of LSTMs trained with either fully-supervised or self-supervised loss to a conventional non-temporal model (LDA) on two data types: segmented ramp data (lacking transition information) and continuous dynamic data inclusive of class transitions. Statistical analysis reveals that the temporal models outperform non-temporal models when trained with continuous dynamic data. Additionally, the proposed VICReg pre-trained temporal model with continuous dynamic data significantly outperformed all other models. Interestingly, when using only ramp data, the LSTM performed worse than the LDA, suggesting potential overfitting due to the absence of sufficient dynamics. This highlights the interplay between data type and model choice. Overall, this work highlights the importance of representative dynamics in training data and the potential for leveraging self-supervised approaches to enhance sEMG-PR models.
{"title":"Self-Supervised Learning via VICReg Enables Training of EMG Pattern Recognition Using Continuous Data with Unclear Labels","authors":"Shriram Tallam Puranam Raghu, Dawn T. MacIsaac, Erik J. Scheme","doi":"arxiv-2409.11632","DOIUrl":"https://doi.org/arxiv-2409.11632","url":null,"abstract":"In this study, we investigate the application of self-supervised learning via\u0000pre-trained Long Short-Term Memory (LSTM) networks for training surface\u0000electromyography pattern recognition models (sEMG-PR) using dynamic data with\u0000transitions. While labeling such data poses challenges due to the absence of\u0000ground-truth labels during transitions between classes, self-supervised\u0000pre-training offers a way to circumvent this issue. We compare the performance\u0000of LSTMs trained with either fully-supervised or self-supervised loss to a\u0000conventional non-temporal model (LDA) on two data types: segmented ramp data\u0000(lacking transition information) and continuous dynamic data inclusive of class\u0000transitions. Statistical analysis reveals that the temporal models outperform\u0000non-temporal models when trained with continuous dynamic data. Additionally,\u0000the proposed VICReg pre-trained temporal model with continuous dynamic data\u0000significantly outperformed all other models. Interestingly, when using only\u0000ramp data, the LSTM performed worse than the LDA, suggesting potential\u0000overfitting due to the absence of sufficient dynamics. This highlights the\u0000interplay between data type and model choice. Overall, this work highlights the\u0000importance of representative dynamics in training data and the potential for\u0000leveraging self-supervised approaches to enhance sEMG-PR models.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"54 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251311","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We propose a new covert communication scheme that operates without pre-sharing side information and channel estimation, utilizing a Gaussian-distributed Grassmann constellation for noncoherent detection. By designing constant-amplitude symbols on the Grassmann manifold and multiplying them by random variables, we generate signals that follow an arbitrary probability distribution, such as Gaussian or skew-normal distributions. The mathematical property of the manifold enables the transmitter's random variables to remain unshared with the receiver, and the elimination of pilot symbols that could compromise covertness. The proposed scheme achieved higher covertness and achievable rates compared to conventional coherent Gaussian signaling schemes, without any penalty in terms of complexity.
{"title":"Covert Communications Without Pre-Sharing of Side Information and Channel Estimation Over Quasi-Static Fading Channels","authors":"Hiroki Fukada, Hiroki Iimori, Chandan Pradhan, Szabolcs Malomsoky, Naoki Ishikawa","doi":"arxiv-2409.11755","DOIUrl":"https://doi.org/arxiv-2409.11755","url":null,"abstract":"We propose a new covert communication scheme that operates without\u0000pre-sharing side information and channel estimation, utilizing a\u0000Gaussian-distributed Grassmann constellation for noncoherent detection. By\u0000designing constant-amplitude symbols on the Grassmann manifold and multiplying\u0000them by random variables, we generate signals that follow an arbitrary\u0000probability distribution, such as Gaussian or skew-normal distributions. The\u0000mathematical property of the manifold enables the transmitter's random\u0000variables to remain unshared with the receiver, and the elimination of pilot\u0000symbols that could compromise covertness. The proposed scheme achieved higher\u0000covertness and achievable rates compared to conventional coherent Gaussian\u0000signaling schemes, without any penalty in terms of complexity.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"116 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251307","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Søren Føns Nielsen, Francesco Da Ros, Mikkel N. Schmidt, Darko Zibar
This paper investigates the application of end-to-end (E2E) learning for joint optimization of pulse-shaper and receiver filter to reduce intersymbol interference (ISI) in bandwidth-limited communication systems. We investigate this in two numerical simulation models: 1) an additive white Gaussian noise (AWGN) channel with bandwidth limitation and 2) an intensity modulated direct detection (IM/DD) link employing an electro-absorption modulator. For both simulation models, we implement a wavelength division multiplexing (WDM) scheme to ensure that the learned filters adhere to the bandwidth constraints of the WDM channels. Our findings reveal that E2E learning greatly surpasses traditional single-sided transmitter pulse-shaper or receiver filter optimization methods, achieving significant performance gains in terms of symbol error rate with shorter filter lengths. These results suggest that E2E learning can decrease the complexity and enhance the performance of future high-speed optical communication systems.
{"title":"End-to-End Learning of Transmitter and Receiver Filters in Bandwidth Limited Fiber Optic Communication Systems","authors":"Søren Føns Nielsen, Francesco Da Ros, Mikkel N. Schmidt, Darko Zibar","doi":"arxiv-2409.11980","DOIUrl":"https://doi.org/arxiv-2409.11980","url":null,"abstract":"This paper investigates the application of end-to-end (E2E) learning for\u0000joint optimization of pulse-shaper and receiver filter to reduce intersymbol\u0000interference (ISI) in bandwidth-limited communication systems. We investigate\u0000this in two numerical simulation models: 1) an additive white Gaussian noise\u0000(AWGN) channel with bandwidth limitation and 2) an intensity modulated direct\u0000detection (IM/DD) link employing an electro-absorption modulator. For both\u0000simulation models, we implement a wavelength division multiplexing (WDM) scheme\u0000to ensure that the learned filters adhere to the bandwidth constraints of the\u0000WDM channels. Our findings reveal that E2E learning greatly surpasses\u0000traditional single-sided transmitter pulse-shaper or receiver filter\u0000optimization methods, achieving significant performance gains in terms of\u0000symbol error rate with shorter filter lengths. These results suggest that E2E\u0000learning can decrease the complexity and enhance the performance of future\u0000high-speed optical communication systems.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hongyu Huang, Zhenming Yu, Yi Lei, Wei Zhang, Yongli Zhao, Shanguo Huang, Kun Xu
To effectively mitigate the influence of atmospheric turbulence, a novel discrete-time analog transmission free-space optical (DTAT-FSO) communication scheme is proposed. It directly maps information sources to discrete-time analog symbols via joint source-channel coding and modulation. Differently from traditional digital free space optical (TD-FSO) schemes, the proposed DTAT-FSO approach can automatically adapt to the variation of the channel state, with no need to adjust the specific modulation and coding scheme. The performance of the DTAT-FSO system was evaluated in both intensity modulation/direct detection (IM/DD) and coherent FSO systems for high-resolution image transmission. The results show that the DTAT-FSO reliably transmits images at low received optical powers (ROPs) and automatically enhances quality at high ROPs, while the TD-FSO experiences cliff and leveling effects when the channel state varies. With respect to the TD-FSO scheme, the DTAT-FSO scheme improved receiver sensitivity by 2.5 dB in the IM/DD FSO system and 0.8 dB in the coherent FSO system, and it achieved superior image fidelity under the same ROP. The automatic adaptation feature and improved performance of the DTAT-FSO suggest its potential for terrestrial, airborne, and satellite optical networks, addressing challenges posed by atmospheric turbulence.
{"title":"Atmospheric Turbulence-Immune Free Space Optical Communication System based on Discrete-Time Analog Transmission","authors":"Hongyu Huang, Zhenming Yu, Yi Lei, Wei Zhang, Yongli Zhao, Shanguo Huang, Kun Xu","doi":"arxiv-2409.11928","DOIUrl":"https://doi.org/arxiv-2409.11928","url":null,"abstract":"To effectively mitigate the influence of atmospheric turbulence, a novel\u0000discrete-time analog transmission free-space optical (DTAT-FSO) communication\u0000scheme is proposed. It directly maps information sources to discrete-time\u0000analog symbols via joint source-channel coding and modulation. Differently from\u0000traditional digital free space optical (TD-FSO) schemes, the proposed DTAT-FSO\u0000approach can automatically adapt to the variation of the channel state, with no\u0000need to adjust the specific modulation and coding scheme. The performance of\u0000the DTAT-FSO system was evaluated in both intensity modulation/direct detection\u0000(IM/DD) and coherent FSO systems for high-resolution image transmission. The\u0000results show that the DTAT-FSO reliably transmits images at low received\u0000optical powers (ROPs) and automatically enhances quality at high ROPs, while\u0000the TD-FSO experiences cliff and leveling effects when the channel state\u0000varies. With respect to the TD-FSO scheme, the DTAT-FSO scheme improved\u0000receiver sensitivity by 2.5 dB in the IM/DD FSO system and 0.8 dB in the\u0000coherent FSO system, and it achieved superior image fidelity under the same\u0000ROP. The automatic adaptation feature and improved performance of the DTAT-FSO\u0000suggest its potential for terrestrial, airborne, and satellite optical\u0000networks, addressing challenges posed by atmospheric turbulence.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"20 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251266","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Alejandro de la Fuente, Guillem Femenias, Felip Riera-Palou, Giovanni Interdonato
Cell-free massive multiple-input multiple-output (CF-mMIMO) is a breakthrough technology for beyond-5G systems, designed to significantly boost the energy and spectral efficiencies of future mobile networks while ensuring a consistent quality of service for all users. Additionally, multicasting has gained considerable attention recently because physical-layer multicasting offers an efficient method for simultaneously serving multiple users with identical service demands by sharing radio resources. Typically, multicast services are delivered either via unicast transmissions or a single multicast transmission. This work, however, introduces a novel subgroup-centric multicast CF-mMIMO framework that divides users into several multicast subgroups based on the similarities in their spatial channel characteristics. This approach allows for efficient sharing of the pilot sequences used for channel estimation and the precoding filters used for data transmission. The proposed framework employs two scalable precoding strategies: centralized improved partial MMSE (IP-MMSE) and distributed conjugate beam-forming (CB). Numerical results show that for scenarios where users are uniformly distributed across the service area, unicast transmissions using centralized IP-MMSE precoding are optimal. However, in cases where users are spatially clustered, multicast subgrouping significantly improves the sum spectral efficiency (SE) of the multicast service compared to both unicast and single multicast transmission. Notably, in clustered scenarios, distributed CB precoding outperforms IP-MMSE in terms of per-user SE, making it the best solution for delivering multicast content.
{"title":"User Subgrouping in Scalable Cell-Free Massive MIMO Multicasting Systems","authors":"Alejandro de la Fuente, Guillem Femenias, Felip Riera-Palou, Giovanni Interdonato","doi":"arxiv-2409.11871","DOIUrl":"https://doi.org/arxiv-2409.11871","url":null,"abstract":"Cell-free massive multiple-input multiple-output (CF-mMIMO) is a breakthrough\u0000technology for beyond-5G systems, designed to significantly boost the energy\u0000and spectral efficiencies of future mobile networks while ensuring a consistent\u0000quality of service for all users. Additionally, multicasting has gained\u0000considerable attention recently because physical-layer multicasting offers an\u0000efficient method for simultaneously serving multiple users with identical\u0000service demands by sharing radio resources. Typically, multicast services are\u0000delivered either via unicast transmissions or a single multicast transmission.\u0000This work, however, introduces a novel subgroup-centric multicast CF-mMIMO\u0000framework that divides users into several multicast subgroups based on the\u0000similarities in their spatial channel characteristics. This approach allows for\u0000efficient sharing of the pilot sequences used for channel estimation and the\u0000precoding filters used for data transmission. The proposed framework employs\u0000two scalable precoding strategies: centralized improved partial MMSE (IP-MMSE)\u0000and distributed conjugate beam-forming (CB). Numerical results show that for\u0000scenarios where users are uniformly distributed across the service area,\u0000unicast transmissions using centralized IP-MMSE precoding are optimal. However,\u0000in cases where users are spatially clustered, multicast subgrouping\u0000significantly improves the sum spectral efficiency (SE) of the multicast\u0000service compared to both unicast and single multicast transmission. Notably, in\u0000clustered scenarios, distributed CB precoding outperforms IP-MMSE in terms of\u0000per-user SE, making it the best solution for delivering multicast content.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"47 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251267","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper illustrates the development of two efficient source localization algorithms for electroencephalography (EEG) data, aimed at enhancing real-time brain signal reconstruction while addressing the computational challenges of traditional methods. Accurate EEG source localization is crucial for applications in cognitive neuroscience, neurorehabilitation, and brain-computer interfaces (BCIs). To make significant progress toward precise source orientation detection and improved signal reconstruction, we introduce the Accelerated Linear Constrained Minimum Variance (ALCMV) beamforming toolbox and the Accelerated Brain Source Orientation Detection (AORI) toolbox. The ALCMV algorithm speeds up EEG source reconstruction by utilizing recursive covariance matrix calculations, while AORI simplifies source orientation detection from three dimensions to one, reducing computational load by 66% compared to conventional methods. Using both simulated and real EEG data, we demonstrate that these algorithms maintain high accuracy, with orientation errors below 0.2% and signal reconstruction accuracy within 2%. These findings suggest that the proposed toolboxes represent a substantial advancement in the efficiency and speed of EEG source localization, making them well-suited for real-time neurotechnological applications.
{"title":"Accelerated Algorithms for Source Orientation Detection (AORI) and Spatiotemporal LCMV (ALCMV) Beamforming in EEG Source Localization","authors":"Ava Yektaeian Vaziri, Bahador Makkiabadi","doi":"arxiv-2409.11751","DOIUrl":"https://doi.org/arxiv-2409.11751","url":null,"abstract":"This paper illustrates the development of two efficient source localization\u0000algorithms for electroencephalography (EEG) data, aimed at enhancing real-time\u0000brain signal reconstruction while addressing the computational challenges of\u0000traditional methods. Accurate EEG source localization is crucial for\u0000applications in cognitive neuroscience, neurorehabilitation, and brain-computer\u0000interfaces (BCIs). To make significant progress toward precise source\u0000orientation detection and improved signal reconstruction, we introduce the\u0000Accelerated Linear Constrained Minimum Variance (ALCMV) beamforming toolbox and\u0000the Accelerated Brain Source Orientation Detection (AORI) toolbox. The ALCMV\u0000algorithm speeds up EEG source reconstruction by utilizing recursive covariance\u0000matrix calculations, while AORI simplifies source orientation detection from\u0000three dimensions to one, reducing computational load by 66% compared to\u0000conventional methods. Using both simulated and real EEG data, we demonstrate\u0000that these algorithms maintain high accuracy, with orientation errors below\u00000.2% and signal reconstruction accuracy within 2%. These findings suggest that\u0000the proposed toolboxes represent a substantial advancement in the efficiency\u0000and speed of EEG source localization, making them well-suited for real-time\u0000neurotechnological applications.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"178 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251309","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jun Wei Yeow, Ee-Leng Tan, Jisheng Bai, Santi Peksi, Woon-Seng Gan
Sound event localization and detection (SELD) is critical for various real-world applications, including smart monitoring and Internet of Things (IoT) systems. Although deep neural networks (DNNs) represent the state-of-the-art approach for SELD, their significant computational complexity and model sizes present challenges for deployment on resource-constrained edge devices, especially under real-time conditions. Despite the growing need for real-time SELD, research in this area remains limited. In this paper, we investigate the unique challenges of deploying SELD systems for real-world, real-time applications by performing extensive experiments on a commercially available Raspberry Pi 3 edge device. Our findings reveal two critical, often overlooked considerations: the high computational cost of feature extraction and the performance degradation associated with low-latency, real-time inference. This paper provides valuable insights and considerations for future work toward developing more efficient and robust real-time SELD systems
{"title":"Real-Time Sound Event Localization and Detection: Deployment Challenges on Edge Devices","authors":"Jun Wei Yeow, Ee-Leng Tan, Jisheng Bai, Santi Peksi, Woon-Seng Gan","doi":"arxiv-2409.11700","DOIUrl":"https://doi.org/arxiv-2409.11700","url":null,"abstract":"Sound event localization and detection (SELD) is critical for various\u0000real-world applications, including smart monitoring and Internet of Things\u0000(IoT) systems. Although deep neural networks (DNNs) represent the\u0000state-of-the-art approach for SELD, their significant computational complexity\u0000and model sizes present challenges for deployment on resource-constrained edge\u0000devices, especially under real-time conditions. Despite the growing need for\u0000real-time SELD, research in this area remains limited. In this paper, we\u0000investigate the unique challenges of deploying SELD systems for real-world,\u0000real-time applications by performing extensive experiments on a commercially\u0000available Raspberry Pi 3 edge device. Our findings reveal two critical, often\u0000overlooked considerations: the high computational cost of feature extraction\u0000and the performance degradation associated with low-latency, real-time\u0000inference. This paper provides valuable insights and considerations for future\u0000work toward developing more efficient and robust real-time SELD systems","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"20 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251316","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Alejandro de la Fuente, Giovanni Interdonato, Giuseppe Araniti
Massive multiple-input-multiple-output (MIMO) is unquestionably a key enabler of the fifth-generation (5G) technology for mobile systems, enabling to meet the high requirements of upcoming mobile broadband services. Physical-layer multicasting refers to a technique for simultaneously serving multiple users, demanding for the same service and sharing the same radio resources, with a single transmission. Massive MIMO systems with multicast communications have been so far studied under the ideal assumption of uncorrelated Rayleigh fading channels. In this work, we consider a practical multicast massive MIMO system over spatially correlated Rayleigh fading channels, investigating the impact of the spatial channel correlation on the favorable propagation, hence on the performance. We propose a subgrouping strategy for the multicast users based on their channel correlation matrices' similarities. The proposed subgrouping approach capitalizes on the spatial correlation to enhance the quality of the channel estimation, and thereby the effectiveness of the precoding. Moreover, we devise a max-min fairness (MMF) power allocation strategy that makes the spectral efficiency (SE) among different multicast subgroups uniform. Lastly, we propose a novel power allocation for uplink (UL) pilot transmission to maximize the SE among the users within the same multicast subgroup. Simulation results show a significant SE gain provided by our user subgrouping and power allocation strategies. Importantly, we show how spatial channel correlation can be exploited to enhance multicast massive MIMO communications.
大规模多输入多输出（MIMO）无疑是第五代（5G）移动系统技术的关键推动因素，能够满足即将到来的移动宽带服务的高要求。物理层多播指的是通过一次传输同时为多个用户提供服务的技术，这些用户需要相同的服务并共享相同的无线电资源。带组播通信的大规模多输入多输出系统迄今一直是在不相关的瑞利衰减信道的理想假设下进行研究的。在这项工作中，我们考虑了在空间相关的瑞利衰减信道上的实用多播大规模 MIMO 系统，研究了空间信道相关性对有利传播的影响，从而对性能的影响。我们提出了一种基于信道相关矩阵相似性的多播用户分组策略。建议的分组方法利用空间相关性来提高信道估计的质量，从而提高预编码的有效性。此外，我们还设计了一种最大最小公平（MMF）功率分配策略，使不同组播子组之间的光谱效率（SE）保持一致。最后，我们为上行链路（UL）先导传输提出了一种新的功率分配方案，以最大限度地提高同一组播子组内用户之间的光谱效率（SE）。仿真结果表明，我们的用户分组和功率分配策略带来了显著的 SE 增益。重要的是，我们展示了如何利用空间信道相关性来增强组播大规模 MIMO 通信。
{"title":"User Subgrouping and Power Control for Multicast Massive MIMO over Spatially Correlated Channels","authors":"Alejandro de la Fuente, Giovanni Interdonato, Giuseppe Araniti","doi":"arxiv-2409.11891","DOIUrl":"https://doi.org/arxiv-2409.11891","url":null,"abstract":"Massive multiple-input-multiple-output (MIMO) is unquestionably a key enabler\u0000of the fifth-generation (5G) technology for mobile systems, enabling to meet\u0000the high requirements of upcoming mobile broadband services. Physical-layer\u0000multicasting refers to a technique for simultaneously serving multiple users,\u0000demanding for the same service and sharing the same radio resources, with a\u0000single transmission. Massive MIMO systems with multicast communications have\u0000been so far studied under the ideal assumption of uncorrelated Rayleigh fading\u0000channels. In this work, we consider a practical multicast massive MIMO system\u0000over spatially correlated Rayleigh fading channels, investigating the impact of\u0000the spatial channel correlation on the favorable propagation, hence on the\u0000performance. We propose a subgrouping strategy for the multicast users based on\u0000their channel correlation matrices' similarities. The proposed subgrouping\u0000approach capitalizes on the spatial correlation to enhance the quality of the\u0000channel estimation, and thereby the effectiveness of the precoding. Moreover,\u0000we devise a max-min fairness (MMF) power allocation strategy that makes the\u0000spectral efficiency (SE) among different multicast subgroups uniform. Lastly,\u0000we propose a novel power allocation for uplink (UL) pilot transmission to\u0000maximize the SE among the users within the same multicast subgroup. Simulation\u0000results show a significant SE gain provided by our user subgrouping and power\u0000allocation strategies. Importantly, we show how spatial channel correlation can\u0000be exploited to enhance multicast massive MIMO communications.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"25 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251308","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Integrated sensing and communications (ISAC) has emerged as a transformative paradigm for next-generation wireless systems. In this paper, we present a novel ISAC scheme that leverages the diffusion Schrodinger bridge (DSB) to realize the sensing of electromagnetic (EM) property of a target as well as the reconstruction of the wireless channel. The DSB framework connects EM property sensing and channel reconstruction by establishing a bidirectional process: the forward process transforms the distribution of EM property into the channel distribution, while the reverse process reconstructs the EM property from the channel. To handle the difference in dimensionality between the high-dimensional sensing channel and the lower-dimensional EM property, we generate latent representations using an autoencoder network. The autoencoder compresses the sensing channel into a latent space that retains essential features, which incorporates positional embeddings to process spatial context. The simulation results demonstrate the effectiveness of the proposed DSB framework, which achieves superior reconstruction of the targets shape, relative permittivity, and conductivity. Moreover, the proposed method can also realize high-fidelity channel reconstruction given the EM property of the target. The dual capability of accurately sensing the EM property and reconstructing the channel across various positions within the sensing area underscores the versatility and potential of the proposed approach for broad application in future ISAC systems.
{"title":"Electromagnetic Property Sensing and Channel Reconstruction Based on Diffusion Schrödinger Bridge in ISAC","authors":"Yuhua Jiang, Feifei Gao, Shi Jin","doi":"arxiv-2409.11651","DOIUrl":"https://doi.org/arxiv-2409.11651","url":null,"abstract":"Integrated sensing and communications (ISAC) has emerged as a transformative\u0000paradigm for next-generation wireless systems. In this paper, we present a\u0000novel ISAC scheme that leverages the diffusion Schrodinger bridge (DSB) to\u0000realize the sensing of electromagnetic (EM) property of a target as well as the\u0000reconstruction of the wireless channel. The DSB framework connects EM property\u0000sensing and channel reconstruction by establishing a bidirectional process: the\u0000forward process transforms the distribution of EM property into the channel\u0000distribution, while the reverse process reconstructs the EM property from the\u0000channel. To handle the difference in dimensionality between the\u0000high-dimensional sensing channel and the lower-dimensional EM property, we\u0000generate latent representations using an autoencoder network. The autoencoder\u0000compresses the sensing channel into a latent space that retains essential\u0000features, which incorporates positional embeddings to process spatial context.\u0000The simulation results demonstrate the effectiveness of the proposed DSB\u0000framework, which achieves superior reconstruction of the targets shape,\u0000relative permittivity, and conductivity. Moreover, the proposed method can also\u0000realize high-fidelity channel reconstruction given the EM property of the\u0000target. The dual capability of accurately sensing the EM property and\u0000reconstructing the channel across various positions within the sensing area\u0000underscores the versatility and potential of the proposed approach for broad\u0000application in future ISAC systems.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"116 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We study a blind deconvolution problem on graphs, which arises in the context of localizing a few sources that diffuse over networks. While the observations are bilinear functions of the unknown graph filter coefficients and sparse input signals, a mild requirement on invertibility of the diffusion filter enables an efficient convex relaxation leading to a linear programming formulation that can be tackled with off-the-shelf solvers. Under the Bernoulli-Gaussian model for the inputs, we derive sufficient exact recovery conditions in the noise-free setting. A stable recovery result is then established, ensuring the estimation error remains manageable even when the observations are corrupted by a small amount of noise. Numerical tests with synthetic and real-world network data illustrate the merits of the proposed algorithm, its robustness to noise as well as the benefits of leveraging multiple signals to aid the (blind) localization of sources of diffusion. At a fundamental level, the results presented here broaden the scope of classical blind deconvolution of (spatio-)temporal signals to irregular graph domains.
{"title":"Blind Deconvolution on Graphs: Exact and Stable Recovery","authors":"Chang Ye, Gonzalo Mateos","doi":"arxiv-2409.12164","DOIUrl":"https://doi.org/arxiv-2409.12164","url":null,"abstract":"We study a blind deconvolution problem on graphs, which arises in the context\u0000of localizing a few sources that diffuse over networks. While the observations\u0000are bilinear functions of the unknown graph filter coefficients and sparse\u0000input signals, a mild requirement on invertibility of the diffusion filter\u0000enables an efficient convex relaxation leading to a linear programming\u0000formulation that can be tackled with off-the-shelf solvers. Under the\u0000Bernoulli-Gaussian model for the inputs, we derive sufficient exact recovery\u0000conditions in the noise-free setting. A stable recovery result is then\u0000established, ensuring the estimation error remains manageable even when the\u0000observations are corrupted by a small amount of noise. Numerical tests with\u0000synthetic and real-world network data illustrate the merits of the proposed\u0000algorithm, its robustness to noise as well as the benefits of leveraging\u0000multiple signals to aid the (blind) localization of sources of diffusion. At a\u0000fundamental level, the results presented here broaden the scope of classical\u0000blind deconvolution of (spatio-)temporal signals to irregular graph domains.","PeriodicalId":501034,"journal":{"name":"arXiv - EE - Signal Processing","volume":"54 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-09-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142251262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：