Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019074
Rolan Khalifa, Minar El-Aasser
Software Defined Networking (SDN) is a revolutionary networking architecture where it has a centralized controller since it separates the control plane and data plane of forwarding elements. In this way, SDN creates a flexible architecture that allows network devices to be configured quickly and easily. Openflow is now the most popular solution for implementing the SDN concept and providing significant flexibility in network flow routing. SDN is exposed to many security threats that will affect the performance of the network Network simulation is a simple and cost-effective technique to see how the network will perform under various operational conditions. The results of the simulation can be used to evaluate and analyze network performance under security threats. In this paper, the SDN scenario model will be developed in OMNeT++ using the INET framework and Openflow protocol. The developed SDN simulation model will be used to create a simulation setup to model security threats in SDN, where a Denial of Service attack (DoS) will be simulated on the Openflow switch and the Openflow controller.
{"title":"Network Security Challenges in SDN Environments","authors":"Rolan Khalifa, Minar El-Aasser","doi":"10.1109/ICCSPA55860.2022.10019074","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019074","url":null,"abstract":"Software Defined Networking (SDN) is a revolutionary networking architecture where it has a centralized controller since it separates the control plane and data plane of forwarding elements. In this way, SDN creates a flexible architecture that allows network devices to be configured quickly and easily. Openflow is now the most popular solution for implementing the SDN concept and providing significant flexibility in network flow routing. SDN is exposed to many security threats that will affect the performance of the network Network simulation is a simple and cost-effective technique to see how the network will perform under various operational conditions. The results of the simulation can be used to evaluate and analyze network performance under security threats. In this paper, the SDN scenario model will be developed in OMNeT++ using the INET framework and Openflow protocol. The developed SDN simulation model will be used to create a simulation setup to model security threats in SDN, where a Denial of Service attack (DoS) will be simulated on the Openflow switch and the Openflow controller.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114870631","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019071
Nader G. Rihan, M. Abdelaziz, Samy S. Soliman
Many different technologies were proposed in the past few years for enhancing indoor positioning: WiFi, Radio Frequency Identification (RFID), Ultra Wide Band (UWB), and Bluetooth to mention some. This study followed the recent IEEE positioning standard (P802.11 az). The standard was developed to enhance indoor navigation by minimizing the consumption power with low hardware complexity. Therefore, this standard enables the usage of artificial intelligence algorithms with relatively high complexity. Also, the usage of this standard will enhance indoor localization and positioning for different commercial purposes. We proposed two methods: Time Of Arrival (TOA) and fingerprinting-deep learning, considering a simple Single Input-Single Input (SISO) system at five Gigahertz with the highest standard allowable bandwidth. The behavior of TOA had very low performance considering a realistic multi-path case. On the other hand, the deep learning algorithm achieved ultra-high indoor positioning resolution (around twelve centimeters). Although TOA is a technique that relies on a simple hardware algorithm relative to deep learning, this paper proved the failure of TOA in a simple indoor environment even using the latest IEEE positioning standard compared with the deep learning method.
{"title":"A Hybrid Deep-learning/Fingerprinting for Indoor Positioning Based on IEEE P802.11az","authors":"Nader G. Rihan, M. Abdelaziz, Samy S. Soliman","doi":"10.1109/ICCSPA55860.2022.10019071","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019071","url":null,"abstract":"Many different technologies were proposed in the past few years for enhancing indoor positioning: WiFi, Radio Frequency Identification (RFID), Ultra Wide Band (UWB), and Bluetooth to mention some. This study followed the recent IEEE positioning standard (P802.11 az). The standard was developed to enhance indoor navigation by minimizing the consumption power with low hardware complexity. Therefore, this standard enables the usage of artificial intelligence algorithms with relatively high complexity. Also, the usage of this standard will enhance indoor localization and positioning for different commercial purposes. We proposed two methods: Time Of Arrival (TOA) and fingerprinting-deep learning, considering a simple Single Input-Single Input (SISO) system at five Gigahertz with the highest standard allowable bandwidth. The behavior of TOA had very low performance considering a realistic multi-path case. On the other hand, the deep learning algorithm achieved ultra-high indoor positioning resolution (around twelve centimeters). Although TOA is a technique that relies on a simple hardware algorithm relative to deep learning, this paper proved the failure of TOA in a simple indoor environment even using the latest IEEE positioning standard compared with the deep learning method.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130623693","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019036
Abdelfatah Ali, M. Mokhtar, M. Shaaban
Electricity theft is a worldwide issue that adversely impacts companies and users. This issue disrupts the expansion of utility companies, produces electric dangers, and affects the high-level cost of electricity for users. The extensive penetration of advanced metering infrastructure networks gives a chance to identify theft cyberattacks by examining the collected data of the energy consumption from smart meters. This work presents a detection approach based on statistical and machine learning to measure theft confidence. An anomaly detection approach is adopted, in which, to detect suspicious data, a theft detection unit based on a fine tree regression model is constructed. Historical data of average load consumption per unit area, smart meter readings, and temperature are employed in the training stage of the proposed approach. The error between the true and estimated data is fitted by a probability density function to identify suspicious data and determine the theft confidence. Different electricity theft cyberattacks are studied to evaluate the efficacy of the developed approach. The obtained results demonstrate the effectiveness of the developed detection approach.
{"title":"Theft Cyberattacks Detection in Smart Grids Based on Machine Learning","authors":"Abdelfatah Ali, M. Mokhtar, M. Shaaban","doi":"10.1109/ICCSPA55860.2022.10019036","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019036","url":null,"abstract":"Electricity theft is a worldwide issue that adversely impacts companies and users. This issue disrupts the expansion of utility companies, produces electric dangers, and affects the high-level cost of electricity for users. The extensive penetration of advanced metering infrastructure networks gives a chance to identify theft cyberattacks by examining the collected data of the energy consumption from smart meters. This work presents a detection approach based on statistical and machine learning to measure theft confidence. An anomaly detection approach is adopted, in which, to detect suspicious data, a theft detection unit based on a fine tree regression model is constructed. Historical data of average load consumption per unit area, smart meter readings, and temperature are employed in the training stage of the proposed approach. The error between the true and estimated data is fitted by a probability density function to identify suspicious data and determine the theft confidence. Different electricity theft cyberattacks are studied to evaluate the efficacy of the developed approach. The obtained results demonstrate the effectiveness of the developed detection approach.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130670500","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019249
Hend K. Gedawy, Khaled A. Harras, A. Erbad
Federated Learning (FL) was introduced to over-come traditional Machine Learning data privacy concerns, and thus, enable us to gain access to more data. Data owners, clients, are orchestrated by a central FL-server to train data locally and only share their model weights. FL approaches have mainly relied on Cloud and/or Edge to aggregate these model weights and propagate training knowledge across clients. However, several issues hinder the scalability of these approaches, especially in communication-challenged environments. In this paper, we propose a novel semi-distributed system to improve FL training accuracy and time, as well as resource-efficiency at the clients. We leverage co-located clusters of high-end IoT devices, known as FemtoClouds, to propagate training knowledge beyond the Edge. We only leverage Edge/Cloud opportunistically to prop-agate knowledge across FemtoCloud pools. Our evaluation shows that our semi-distributed FemtoClouds system achieves competitive accuracy to state-of-the-art FL approaches, with up to 95% time savings and up to 84% energy savings.
{"title":"Leveraging Semi-Connected Devices To Enhance Federated Learning","authors":"Hend K. Gedawy, Khaled A. Harras, A. Erbad","doi":"10.1109/ICCSPA55860.2022.10019249","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019249","url":null,"abstract":"Federated Learning (FL) was introduced to over-come traditional Machine Learning data privacy concerns, and thus, enable us to gain access to more data. Data owners, clients, are orchestrated by a central FL-server to train data locally and only share their model weights. FL approaches have mainly relied on Cloud and/or Edge to aggregate these model weights and propagate training knowledge across clients. However, several issues hinder the scalability of these approaches, especially in communication-challenged environments. In this paper, we propose a novel semi-distributed system to improve FL training accuracy and time, as well as resource-efficiency at the clients. We leverage co-located clusters of high-end IoT devices, known as FemtoClouds, to propagate training knowledge beyond the Edge. We only leverage Edge/Cloud opportunistically to prop-agate knowledge across FemtoCloud pools. Our evaluation shows that our semi-distributed FemtoClouds system achieves competitive accuracy to state-of-the-art FL approaches, with up to 95% time savings and up to 84% energy savings.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134614480","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019204
Mahdi Beitollahi, Mingrui Liu, Ning Lu
Federated Learning (FL) is considered the key, enabling approach for privacy-preserving, distributed machine learning (ML) systems. FL requires the periodic transmission of ML models from users to the server. Therefore, communication via resource-constrained networks is currently a fundamental bottleneck in FL, which is restricting the ML model complexity and user participation. One of the notable trends to reduce the communication cost of FL systems is gradient compression, in which techniques in the form of sparsification are utilized. However, these methods utilize a single compression rate for all users and do not consider communication heterogeneity in a real-world FL system. Therefore, these methods are bottlenecked by the worst communication capacity across users. Further, sparsification methods are non-adaptive and do not utilize the redundant, similar information across users' ML models for compression. In this paper, we introduce a novel Dynamic Sparsification for Federated Learning (DSFL) approach that enables users to compress their local models based on their communication capacity at each iteration by using two novel sparsification methods: layer-wise similarity sparsification (LSS) and extended top- $K$ sparsification. LSS enables DSFL to utilize the global redundant information in users' models by using the Centralized Kernel Alignment (CKA) similarity for sparsification. The extended top-$K$ model sparsification method empowers DSFL to accommodate the heterogeneous communication capacity of user devices by allowing different values of sparsification rate $K$ for each user at each iteration. Our extensive experimental results11All code and experiments are publicly available at: https://github.com/mahdibeit/DSFL. on three datasets show that DSFL has a faster convergence rate than fixed sparsification, and as the communication heterogeneity increases, this gap increases. Further, our thorough experimental investigations uncover the similarities of user models across the FL system.
{"title":"DSFL: Dynamic Sparsification for Federated Learning","authors":"Mahdi Beitollahi, Mingrui Liu, Ning Lu","doi":"10.1109/ICCSPA55860.2022.10019204","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019204","url":null,"abstract":"Federated Learning (FL) is considered the key, enabling approach for privacy-preserving, distributed machine learning (ML) systems. FL requires the periodic transmission of ML models from users to the server. Therefore, communication via resource-constrained networks is currently a fundamental bottleneck in FL, which is restricting the ML model complexity and user participation. One of the notable trends to reduce the communication cost of FL systems is gradient compression, in which techniques in the form of sparsification are utilized. However, these methods utilize a single compression rate for all users and do not consider communication heterogeneity in a real-world FL system. Therefore, these methods are bottlenecked by the worst communication capacity across users. Further, sparsification methods are non-adaptive and do not utilize the redundant, similar information across users' ML models for compression. In this paper, we introduce a novel Dynamic Sparsification for Federated Learning (DSFL) approach that enables users to compress their local models based on their communication capacity at each iteration by using two novel sparsification methods: layer-wise similarity sparsification (LSS) and extended top- $K$ sparsification. LSS enables DSFL to utilize the global redundant information in users' models by using the Centralized Kernel Alignment (CKA) similarity for sparsification. The extended top-$K$ model sparsification method empowers DSFL to accommodate the heterogeneous communication capacity of user devices by allowing different values of sparsification rate $K$ for each user at each iteration. Our extensive experimental results11All code and experiments are publicly available at: https://github.com/mahdibeit/DSFL. on three datasets show that DSFL has a faster convergence rate than fixed sparsification, and as the communication heterogeneity increases, this gap increases. Further, our thorough experimental investigations uncover the similarities of user models across the FL system.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134027714","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019047
Abdelrahman Tawfik, Shehab Hosny, Sara Hisham, Ali Amr Farouk, Doha Mustafa, Samaa Abdel Moaty, A. Gamal, Khaled Salah
Multimedia compression is a fundamental and significant research topic in the industrial field in the past several decades attempting to improve compression techniques. It is always a trade-off between size and quality where the growth rate of image, audio and video data is far beyond the improvement of the compression ratios achieved so far. Here, we are aiming to explore the potential of neural networks to achieve data compression, making use of multilayer neural networks providing a more efficient solution. In this paper, we present a lossy compression architecture, which utilizes the advantages of convolutional autoencoder (CAE) to replace the conventional transforms. Experimental results demonstrate that our method outperforms traditional coding algorithms, by achieving better compression ratios over the related work.
{"title":"A Generic Real Time Autoencoder-Based Lossy Image Compression","authors":"Abdelrahman Tawfik, Shehab Hosny, Sara Hisham, Ali Amr Farouk, Doha Mustafa, Samaa Abdel Moaty, A. Gamal, Khaled Salah","doi":"10.1109/ICCSPA55860.2022.10019047","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019047","url":null,"abstract":"Multimedia compression is a fundamental and significant research topic in the industrial field in the past several decades attempting to improve compression techniques. It is always a trade-off between size and quality where the growth rate of image, audio and video data is far beyond the improvement of the compression ratios achieved so far. Here, we are aiming to explore the potential of neural networks to achieve data compression, making use of multilayer neural networks providing a more efficient solution. In this paper, we present a lossy compression architecture, which utilizes the advantages of convolutional autoencoder (CAE) to replace the conventional transforms. Experimental results demonstrate that our method outperforms traditional coding algorithms, by achieving better compression ratios over the related work.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114692651","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10018978
Vijaya Nirmala Mitnala, M. Reed, Ian Kegel, J. Bicknell
Sustained growth in the smart speaker market has helped establish high quality, far-field speech communications as a viable alternative to the handset. Seamless handover offers a simple but effective way of improving the far-field communication experience by automatically switching to the best available device regardless of where a user is located. While the basic concept of seamless handover has been proven in a lab environment, this paper proposes two significant enhancements: reduction in media disruption during handover by introducing a parallel session on multiple devices through session initiation protocol (SIP) call forking; and, coherence-based signal processing to more accurately determine the most suitable device for the user. The solution proposed uses the magnitude square coherence (MSC) and results verified through simulation and real datasets show it has excellent performance. However, the raw MSC is found to have high variation due to room effects, consequently this work shows that a smoothing predictor is needed to significantly reduce the extraneous transitions that would otherwise be subjectively poor. Unlike a purely location based approach, the proposed solution selects the best smart device without any environment specific calibration making it ideal for straightforward deployment of a pervasive speech application that uses smart speakers.
{"title":"Seamless device handover for pervasive speech communication","authors":"Vijaya Nirmala Mitnala, M. Reed, Ian Kegel, J. Bicknell","doi":"10.1109/ICCSPA55860.2022.10018978","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10018978","url":null,"abstract":"Sustained growth in the smart speaker market has helped establish high quality, far-field speech communications as a viable alternative to the handset. Seamless handover offers a simple but effective way of improving the far-field communication experience by automatically switching to the best available device regardless of where a user is located. While the basic concept of seamless handover has been proven in a lab environment, this paper proposes two significant enhancements: reduction in media disruption during handover by introducing a parallel session on multiple devices through session initiation protocol (SIP) call forking; and, coherence-based signal processing to more accurately determine the most suitable device for the user. The solution proposed uses the magnitude square coherence (MSC) and results verified through simulation and real datasets show it has excellent performance. However, the raw MSC is found to have high variation due to room effects, consequently this work shows that a smoothing predictor is needed to significantly reduce the extraneous transitions that would otherwise be subjectively poor. Unlike a purely location based approach, the proposed solution selects the best smart device without any environment specific calibration making it ideal for straightforward deployment of a pervasive speech application that uses smart speakers.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120890903","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019245
A. Albaseer, M. Abdallah
Honeypot is considered a powerful complement to the Network Intrusion Detection System (NIDS) in smart grid (SG) systems, which minimizes the workload of NIDSs while providing access to information about the attacker's actions. This assists in further tracing the attack surface and, in return, enables the NIDSs to prevent such behaviors. Machine learning (ML) has recently attracted considerable attention in the SG security domain as a stringent technique for designing and implementing algorithms to predict security threats. However, large data sets collected by honeypots require more effort for faster response, real-time processing, and decision-making, especially for limited resources SG's devices. Thus, this paper proposes an approach to address this challenge, including feature extraction, oversampling and weak label combinations. We demonstrate that all classic ML algorithms cannot maintain the desired performance level when reducing the number of selected features (i.e., using only 25% of the features). As a result, we resort to the Deep Learning approach and propose an LSTM-based model that outperforms the state-of-the-art in terms of accuracy, precision, recall, and f1-score. We conduct extensive simulations using a realistic dataset that includes large log files. The proposed approach can employ just 25% of the features from each collected network packet while attaining 99.8% testing accuracy with a 13% improvement compared to the benchmarks.
{"title":"Fine-tuned LSTM-Based Model for Efficient Honeypot-Based Network Intrusion Detection System in Smart Grid Networks","authors":"A. Albaseer, M. Abdallah","doi":"10.1109/ICCSPA55860.2022.10019245","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019245","url":null,"abstract":"Honeypot is considered a powerful complement to the Network Intrusion Detection System (NIDS) in smart grid (SG) systems, which minimizes the workload of NIDSs while providing access to information about the attacker's actions. This assists in further tracing the attack surface and, in return, enables the NIDSs to prevent such behaviors. Machine learning (ML) has recently attracted considerable attention in the SG security domain as a stringent technique for designing and implementing algorithms to predict security threats. However, large data sets collected by honeypots require more effort for faster response, real-time processing, and decision-making, especially for limited resources SG's devices. Thus, this paper proposes an approach to address this challenge, including feature extraction, oversampling and weak label combinations. We demonstrate that all classic ML algorithms cannot maintain the desired performance level when reducing the number of selected features (i.e., using only 25% of the features). As a result, we resort to the Deep Learning approach and propose an LSTM-based model that outperforms the state-of-the-art in terms of accuracy, precision, recall, and f1-score. We conduct extensive simulations using a realistic dataset that includes large log files. The proposed approach can employ just 25% of the features from each collected network packet while attaining 99.8% testing accuracy with a 13% improvement compared to the benchmarks.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123271149","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019048
Adel Mahmoud, Samuel Medhat, Mark Maged, Othman Mohamed, Reham Karam, Khaled Salah, M. El-Kharashi
Data compression is a trending field that is used in data storage and data transmission systems. Lossy compression means that data cannot be completely retrieved while in lossless compression the compressed data must be reconstructed exactly. Lossless data compression is used in compressing binary files, telemetry data and high-fidelity medical and scientific images where details are crucial. There is no generic compression algorithm that gives best compression ratio on all data pattern. In this paper, we propose a hybrid lossless hardware architecture that compresses most of data patterns such as repeated data, Gaussian distribution data and images. A profiling-before-compressing and then choosing the right compression hardware is proposed. The proposed design is a highly parallelized architecture that can compress/decompress 64 bytes/cycle with minor overhead. Moreover, it provides high compression ratio on small block sizes as well as large ones.
{"title":"An Efficient Hardware Accelerator For Lossless Data Compression","authors":"Adel Mahmoud, Samuel Medhat, Mark Maged, Othman Mohamed, Reham Karam, Khaled Salah, M. El-Kharashi","doi":"10.1109/ICCSPA55860.2022.10019048","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019048","url":null,"abstract":"Data compression is a trending field that is used in data storage and data transmission systems. Lossy compression means that data cannot be completely retrieved while in lossless compression the compressed data must be reconstructed exactly. Lossless data compression is used in compressing binary files, telemetry data and high-fidelity medical and scientific images where details are crucial. There is no generic compression algorithm that gives best compression ratio on all data pattern. In this paper, we propose a hybrid lossless hardware architecture that compresses most of data patterns such as repeated data, Gaussian distribution data and images. A profiling-before-compressing and then choosing the right compression hardware is proposed. The proposed design is a highly parallelized architecture that can compress/decompress 64 bytes/cycle with minor overhead. Moreover, it provides high compression ratio on small block sizes as well as large ones.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131497429","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}
Pub Date : 2022-12-27DOI: 10.1109/ICCSPA55860.2022.10019147
N. K. Mallat, A. Jafarieh, M. Nouri, H. Behroozi
Wider impedance bandwidth (IBW), and lower latency rate than older mobile communication systems possess are required for fifth-generation (5G) mobile communication systems. Furthermore, with respect to the high operation frequency of 5G systems, a high released gain is necessary to compensate for the high path loss on these frequencies. With respect to the requirements mentioned above, millimeter-wave (MMW) antennas seem to be a good solution for 5G applications. The low wavelength of MMW frequency bands, makes it practical to use large array antennas for massive multi input multi-output (MIMO) 5G systems with high gain. The high number of design variables of antennas makes an optimum antenna harder to design. Using machine learning (ML) approaches, however, alleviates this challenge. However, most ML approaches entail high computational complexity. Therefore, surrogate-based optimization (SBO) approaches must be used to handle the high computational complexity of ML approaches.
{"title":"A Machine Learning Based Design of mmWave Compact Array Antenna for 5G Communications","authors":"N. K. Mallat, A. Jafarieh, M. Nouri, H. Behroozi","doi":"10.1109/ICCSPA55860.2022.10019147","DOIUrl":"https://doi.org/10.1109/ICCSPA55860.2022.10019147","url":null,"abstract":"Wider impedance bandwidth (IBW), and lower latency rate than older mobile communication systems possess are required for fifth-generation (5G) mobile communication systems. Furthermore, with respect to the high operation frequency of 5G systems, a high released gain is necessary to compensate for the high path loss on these frequencies. With respect to the requirements mentioned above, millimeter-wave (MMW) antennas seem to be a good solution for 5G applications. The low wavelength of MMW frequency bands, makes it practical to use large array antennas for massive multi input multi-output (MIMO) 5G systems with high gain. The high number of design variables of antennas makes an optimum antenna harder to design. Using machine learning (ML) approaches, however, alleviates this challenge. However, most ML approaches entail high computational complexity. Therefore, surrogate-based optimization (SBO) approaches must be used to handle the high computational complexity of ML approaches.","PeriodicalId":106639,"journal":{"name":"2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2022-12-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130821944","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：