Network Functions Virtualization (NFV) leverages virtualization and cloud computing technologies to make networks more flexible, manageable, and scalable. Instead of using traditional hardware middleboxes, NFV uses more flexible Virtual Network Functions (VNFs) running on commodity servers. One of the key challenges in NFV is to ensure strict reliability and low latency while also improving energy efficiency. Any software or hardware failures in an NFV environment can disrupt the service provided by a chain of VNFs, known as a Service Function Chain (SFC), resulting in significant data loss, delays, and wasted resources. Due to the sequential nature of SFC, latency increases linearly with the number of VNFs. To address this issue, researchers have proposed parallelized SFC or VNF parallelization, which allows multiple independent VNFs in an SFC to run in parallel. In this work, we propose a method to solve the parallelized SFC deployment problem as an Integer Linear Program (ILP) that minimizes energy consumption while ensuring reliability and delay constraints. Since the problem is NP-hard, we also propose a heuristic scheme named ERASE that determines the placement of VNFs and routes traffic through them in a way that minimizes energy consumption while meeting capacity, reliability, and delay requirements. The effectiveness of ERASE is evaluated through extensive simulations and it is shown to perform better than benchmark schemes in terms of total energy consumption and reliability achieved.
{"title":"Energy- and Reliability-Aware Provisioning of Parallelized Service Function Chains With Delay Guarantees","authors":"Venkatarami Reddy Chintapalli;Balaprakasa Rao Killi;Rajat Partani;Bheemarjuna Reddy Tamma;C. Siva Ram Murthy","doi":"10.1109/TGCN.2023.3317927","DOIUrl":"10.1109/TGCN.2023.3317927","url":null,"abstract":"Network Functions Virtualization (NFV) leverages virtualization and cloud computing technologies to make networks more flexible, manageable, and scalable. Instead of using traditional hardware middleboxes, NFV uses more flexible Virtual Network Functions (VNFs) running on commodity servers. One of the key challenges in NFV is to ensure strict reliability and low latency while also improving energy efficiency. Any software or hardware failures in an NFV environment can disrupt the service provided by a chain of VNFs, known as a Service Function Chain (SFC), resulting in significant data loss, delays, and wasted resources. Due to the sequential nature of SFC, latency increases linearly with the number of VNFs. To address this issue, researchers have proposed parallelized SFC or VNF parallelization, which allows multiple independent VNFs in an SFC to run in parallel. In this work, we propose a method to solve the parallelized SFC deployment problem as an Integer Linear Program (ILP) that minimizes energy consumption while ensuring reliability and delay constraints. Since the problem is NP-hard, we also propose a heuristic scheme named ERASE that determines the placement of VNFs and routes traffic through them in a way that minimizes energy consumption while meeting capacity, reliability, and delay requirements. The effectiveness of ERASE is evaluated through extensive simulations and it is shown to perform better than benchmark schemes in terms of total energy consumption and reliability achieved.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135599196","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-22DOI: 10.1109/TGCN.2023.3317829
Yang Li;Anzhong Hu
This paper studies the channel estimation of millimeter-wave massive multiple-input multiple-output orthogonal frequency division multiplexing systems with a wide bandwidth and a hybrid structure. By considering the influence of the frequency-spatial wideband effect and the channel sparsity, the subcarriers are grouped, the support as well as the angle of each group are estimated. Based on the angle estimate, the support of each subcarrier is deduced. The computational complexities are analyzed and compared, which show the proposed method has a lower computational complexity. The Cramér-Rao bound (CRB) and the spectral efficiency are also analyzed. The analysis shows that the spectral efficiency increases logarithmly with the increase of the signal-to-noise ratio when channel estimation is perfect but is limited otherwise. The simulations show the proposed method gets closer to the CRB in channel estimation error and achieves higher spectral efficiency than the existing one.
{"title":"Channel Estimation via Subcarrier Grouping for Wideband mmWave Hybrid Massive MIMO-OFDM Systems","authors":"Yang Li;Anzhong Hu","doi":"10.1109/TGCN.2023.3317829","DOIUrl":"10.1109/TGCN.2023.3317829","url":null,"abstract":"This paper studies the channel estimation of millimeter-wave massive multiple-input multiple-output orthogonal frequency division multiplexing systems with a wide bandwidth and a hybrid structure. By considering the influence of the frequency-spatial wideband effect and the channel sparsity, the subcarriers are grouped, the support as well as the angle of each group are estimated. Based on the angle estimate, the support of each subcarrier is deduced. The computational complexities are analyzed and compared, which show the proposed method has a lower computational complexity. The Cramér-Rao bound (CRB) and the spectral efficiency are also analyzed. The analysis shows that the spectral efficiency increases logarithmly with the increase of the signal-to-noise ratio when channel estimation is perfect but is limited otherwise. The simulations show the proposed method gets closer to the CRB in channel estimation error and achieves higher spectral efficiency than the existing one.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135599187","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-20DOI: 10.1109/TGCN.2023.3317674
Manobendu Sarker;Abraham O. Fapojuwo
In this paper, two centralized power allocation schemes are proposed for data transmission during the uplink phase in the user-centric cell-free (CF) massive multiple-input multiple-output (mMIMO) systems. The proposed schemes solve two non-convex power allocation problems of maximizing the summation of spectral efficiency (SE) (max-sum-SE) and that of maximizing the minimum SE (max-min-SE) to improve the overall SE and fairness performance while simultaneously reducing the per-user equipment (UE) transmission power. To solve the max-sum-SE problem, we utilize the fractional programming (FP) method to transform the non-convex problem into a series of convex problems. Furthermore, the max-min-SE problem is solved after reformulating it with the help of the FP method along with the alternating direction method of multipliers (ADMM) technique. The proposed schemes are computationally efficient as they solve the aforementioned problems iteratively by using only closed-form updates for the decision variables, which is one of their strongest features, and suitable for allocating power in large-scale CF mMIMO systems. Numerical results demonstrate that, compared to the no power control scheme, the proposed schemes improve the average SE performance by up to 47% while reducing the average transmission power by up to 95%.
{"title":"Fractional Programming-Based Uplink Transmission Power Allocation for User-Centric Cell-Free Massive MIMO Systems","authors":"Manobendu Sarker;Abraham O. Fapojuwo","doi":"10.1109/TGCN.2023.3317674","DOIUrl":"10.1109/TGCN.2023.3317674","url":null,"abstract":"In this paper, two centralized power allocation schemes are proposed for data transmission during the uplink phase in the user-centric cell-free (CF) massive multiple-input multiple-output (mMIMO) systems. The proposed schemes solve two non-convex power allocation problems of maximizing the summation of spectral efficiency (SE) (max-sum-SE) and that of maximizing the minimum SE (max-min-SE) to improve the overall SE and fairness performance while simultaneously reducing the per-user equipment (UE) transmission power. To solve the max-sum-SE problem, we utilize the fractional programming (FP) method to transform the non-convex problem into a series of convex problems. Furthermore, the max-min-SE problem is solved after reformulating it with the help of the FP method along with the alternating direction method of multipliers (ADMM) technique. The proposed schemes are computationally efficient as they solve the aforementioned problems iteratively by using only closed-form updates for the decision variables, which is one of their strongest features, and suitable for allocating power in large-scale CF mMIMO systems. Numerical results demonstrate that, compared to the no power control scheme, the proposed schemes improve the average SE performance by up to 47% while reducing the average transmission power by up to 95%.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135557446","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Low latency communication is one of the fundamental requirements for 5G wireless networks and beyond. In this paper, a novel approach for joint caching, user scheduling and resource allocation is proposed for minimizing the queuing latency in serving users’ requests in cloud-aided wireless networks. Due to the slow temporal variations in user requests, a time-scale separation technique is used to decouple the joint caching problem from user scheduling and radio resource allocation problems. To serve the spatio-temporal user requests under storage limitations, a Reinforcement Learning (RL) approach is used to optimize the caching strategy at the small cell base stations by minimizing the content fetching cost. A spectral clustering algorithm is proposed to speed-up the convergence of the RL algorithm for a large content caching problem by clustering contents based on user requests. Meanwhile, a dynamic mechanism is proposed to locally group coupled base stations based on user requests to collaboratively optimize the caching strategies. To further improve the latency in fetching and serving user requests, a dynamic matching algorithm is proposed to schedule users and to allocate users to radio resources based on user requests and queue lengths under probabilistic latency constraints. Simulation results show the proposed approach significantly reduces the average delay from 21% to 90% compared to random caching strategy, random resource allocation and random scheduling baselines.
{"title":"Latency-Aware Radio Resource Optimization in Learning-Based Cloud-Aided Small Cell Wireless Networks","authors":"Syed Tamoor-ul-Hassan;Sumudu Samarakoon;Mehdi Bennis;Matti Latva-aho","doi":"10.1109/TGCN.2023.3317128","DOIUrl":"10.1109/TGCN.2023.3317128","url":null,"abstract":"Low latency communication is one of the fundamental requirements for 5G wireless networks and beyond. In this paper, a novel approach for joint caching, user scheduling and resource allocation is proposed for minimizing the queuing latency in serving users’ requests in cloud-aided wireless networks. Due to the slow temporal variations in user requests, a time-scale separation technique is used to decouple the joint caching problem from user scheduling and radio resource allocation problems. To serve the spatio-temporal user requests under storage limitations, a Reinforcement Learning (RL) approach is used to optimize the caching strategy at the small cell base stations by minimizing the content fetching cost. A spectral clustering algorithm is proposed to speed-up the convergence of the RL algorithm for a large content caching problem by clustering contents based on user requests. Meanwhile, a dynamic mechanism is proposed to locally group coupled base stations based on user requests to collaboratively optimize the caching strategies. To further improve the latency in fetching and serving user requests, a dynamic matching algorithm is proposed to schedule users and to allocate users to radio resources based on user requests and queue lengths under probabilistic latency constraints. Simulation results show the proposed approach significantly reduces the average delay from 21% to 90% compared to random caching strategy, random resource allocation and random scheduling baselines.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135555673","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, a multipath component aggregation (MCA) mechanism is introduced for spatial scattering modulation (SSM) to overcome the limitation in conventional SSM that the transmit antenna array steers the beam to a single multipath (MP) component at each instance. In the proposed MCA-SSM system, information bits are divided into two streams. One is mapped to an amplitude-phase-modulation (APM) constellation symbol, and the other is mapped to a beam vector symbol which steers multiple beams to selected strongest MP components via an MCA matrix. In comparison with the conventional SSM system, the proposed MCA-SSM enhances the bit error performance by avoiding both low receiving power due to steering the beam to a single weak MP component and inter-MP interference due to MP components with close values of angle of arrival (AoA) or angle of departure (AoD). For the proposed MCA-SSM, a union upper bound (UUB) on the average bit error probability (ABEP) with any MCA matrix is analytically derived and validated via Monte Carlo simulations. Based on the UUB, the MCA matrix is analytically optimized to minimize the ABEP of the MCA-SSM. Finally, numerical experiments are carried out, which show that the proposed MCA-SSM system remarkably outperforms the state-of-the-art SSM system in terms of ABEP under a typical indoor environment.
{"title":"Spatial Scattering Modulation With Multipath Component Aggregation","authors":"Jiliang Zhang;Wei Liu;Alan Tennant;Weijie Qi;Jiming Chen;Jie Zhang","doi":"10.1109/TGCN.2023.3315247","DOIUrl":"10.1109/TGCN.2023.3315247","url":null,"abstract":"In this paper, a multipath component aggregation (MCA) mechanism is introduced for spatial scattering modulation (SSM) to overcome the limitation in conventional SSM that the transmit antenna array steers the beam to a single multipath (MP) component at each instance. In the proposed MCA-SSM system, information bits are divided into two streams. One is mapped to an amplitude-phase-modulation (APM) constellation symbol, and the other is mapped to a beam vector symbol which steers multiple beams to selected strongest MP components via an MCA matrix. In comparison with the conventional SSM system, the proposed MCA-SSM enhances the bit error performance by avoiding both low receiving power due to steering the beam to a single weak MP component and inter-MP interference due to MP components with close values of angle of arrival (AoA) or angle of departure (AoD). For the proposed MCA-SSM, a union upper bound (UUB) on the average bit error probability (ABEP) with any MCA matrix is analytically derived and validated via Monte Carlo simulations. Based on the UUB, the MCA matrix is analytically optimized to minimize the ABEP of the MCA-SSM. Finally, numerical experiments are carried out, which show that the proposed MCA-SSM system remarkably outperforms the state-of-the-art SSM system in terms of ABEP under a typical indoor environment.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135443290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-13DOI: 10.1109/TGCN.2023.3314870
Elio Faddoul;Ghassan M. Kraidy;Constantinos Psomas;Symeon Chatzinotas;Ioannis Krikidis
In this paper, we present a non-coherent energy detection scheme for spatial modulation (SM) systems. In particular, the use of SM is motivated by its low-complexity implementation in comparison to multiple-input multiple-output (MIMO) systems, achieved through the activation of a single antenna during transmission. Moreover, energy detection-based communications restrict the channel state information to the magnitude of the fading gains. This consideration makes the design applicable for low-cost low-powered devices since phase estimation and its associated circuitry are avoided. We derive an energy detection metric for a multi-antenna receiver based on the maximum-likelihood (ML) criterion. By considering a biased pulse amplitude modulation, we develop an analytical framework for the SM symbol error rate at high signal-to-noise ratios. Numerical results show that the diversity order is proportional to half the number of receive antennas; this result stems from having partial receiver channel knowledge. In addition, we compare the performance of the proposed scheme with that of the coherent ML receiver and show that the SM energy detector outperforms its coherent counterpart in certain scenarios, particularly when utilizing non-negative constellations. Ultimately, we implement an SM testbed using software-defined radio devices and provide experimental error rate measurements that validate our theoretical contribution.
本文针对空间调制（SM）系统提出了一种非相干能量检测方案。与多输入多输出（MIMO）系统相比，空间调制（SM）系统的实施复杂度较低，在传输过程中只需激活一根天线即可实现。此外，基于能量检测的通信将信道状态信息限制为衰减增益的大小。由于避免了相位估计及其相关电路，这种设计适用于低成本、低功率设备。我们基于最大似然（ML）准则，为多天线接收器推导出一种能量检测指标。通过考虑偏置脉冲幅度调制，我们为高信噪比条件下的 SM 符号错误率建立了一个分析框架。数值结果表明，分集阶与接收天线数量的一半成正比；这一结果源于部分接收器信道知识。此外，我们还将所提方案的性能与相干 ML 接收器的性能进行了比较，结果表明 SM 能量检测器在某些情况下的性能优于相干接收器，尤其是在使用非负星座时。最后，我们利用软件定义无线电设备实现了 SM 测试平台，并提供了实验误差率测量结果，验证了我们的理论贡献。
{"title":"Spatial Modulation With Energy Detection: Diversity Analysis and Experimental Evaluation","authors":"Elio Faddoul;Ghassan M. Kraidy;Constantinos Psomas;Symeon Chatzinotas;Ioannis Krikidis","doi":"10.1109/TGCN.2023.3314870","DOIUrl":"10.1109/TGCN.2023.3314870","url":null,"abstract":"In this paper, we present a non-coherent energy detection scheme for spatial modulation (SM) systems. In particular, the use of SM is motivated by its low-complexity implementation in comparison to multiple-input multiple-output (MIMO) systems, achieved through the activation of a single antenna during transmission. Moreover, energy detection-based communications restrict the channel state information to the magnitude of the fading gains. This consideration makes the design applicable for low-cost low-powered devices since phase estimation and its associated circuitry are avoided. We derive an energy detection metric for a multi-antenna receiver based on the maximum-likelihood (ML) criterion. By considering a biased pulse amplitude modulation, we develop an analytical framework for the SM symbol error rate at high signal-to-noise ratios. Numerical results show that the diversity order is proportional to half the number of receive antennas; this result stems from having partial receiver channel knowledge. In addition, we compare the performance of the proposed scheme with that of the coherent ML receiver and show that the SM energy detector outperforms its coherent counterpart in certain scenarios, particularly when utilizing non-negative constellations. Ultimately, we implement an SM testbed using software-defined radio devices and provide experimental error rate measurements that validate our theoretical contribution.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135403386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-07DOI: 10.1109/TGCN.2023.3312707
He-Hsuan Liu;Kuang-Hsun Lin;Hung-Yu Wei
The Discontinuous Reception (DRX) mechanism is designed by the Third Generation Partnership Project (3GPP) for power saving in the Long-Term Evolution and the New Radio technologies. In Release 16, 3GPP designed an advanced signal called Downlink Control Information of Power Saving (DCP). Energy-critical devices can conserve more energy by turning off the receiving module for a long period according to the DCP indication from the next-generation base station (gNB). However, the possible failure of DCP potentially causes extra power consumption or packet loss. To handle the problem, 3GPP allows the gNB to configure the default behavior when the DCP is not detected. In this paper, we accurately model the two DCP mechanisms with different default settings with a three-state discrete-time Markov model. Moreover, we also propose a method that efficiently finds a proper parameter configuration under latency and reliability constraints. Through simulation, we validate the accuracy of the model. In addition, we conduct another simulation to show that the proposed method finds a parameter set yielding near-optimal performance within a short execution time. The derived model, along with the proposed parameter setting method, serves as a solid basis for future research on power saving and sustainability.
{"title":"NR DCP: An Enhanced Power Saving Mechanism With Wake-Up Signal Enabled DRX","authors":"He-Hsuan Liu;Kuang-Hsun Lin;Hung-Yu Wei","doi":"10.1109/TGCN.2023.3312707","DOIUrl":"10.1109/TGCN.2023.3312707","url":null,"abstract":"The Discontinuous Reception (DRX) mechanism is designed by the Third Generation Partnership Project (3GPP) for power saving in the Long-Term Evolution and the New Radio technologies. In Release 16, 3GPP designed an advanced signal called Downlink Control Information of Power Saving (DCP). Energy-critical devices can conserve more energy by turning off the receiving module for a long period according to the DCP indication from the next-generation base station (gNB). However, the possible failure of DCP potentially causes extra power consumption or packet loss. To handle the problem, 3GPP allows the gNB to configure the default behavior when the DCP is not detected. In this paper, we accurately model the two DCP mechanisms with different default settings with a three-state discrete-time Markov model. Moreover, we also propose a method that efficiently finds a proper parameter configuration under latency and reliability constraints. Through simulation, we validate the accuracy of the model. In addition, we conduct another simulation to show that the proposed method finds a parameter set yielding near-optimal performance within a short execution time. The derived model, along with the proposed parameter setting method, serves as a solid basis for future research on power saving and sustainability.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62571503","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-04DOI: 10.1109/TGCN.2023.3311590
Muchen Jiang;Kwan-Wu Chin
This paper studies sample delivery in a multi-hop network where a power beacon charges devices via radio frequency (RF) signals. Devices forward samples with a deadline from a source to a sink. The goal is to minimize the power beacon’s transmit power and guarantee that samples arrive at the sink with probability $(1-epsilon)$ by their deadline, where $epsilon $ is a given probability of failure. A key challenge is that the power beacon does not have instantaneous channel gains information to devices and also between devices; i.e., it does not know the energy level of devices. To this end, we formulate a chance-constrained stochastic program for the problem at hand, and employ the sample-average approximation (SAA) method to compute a solution. We also outline two novel approximation methods: Sampling based Probabilistic Optimal Power Allocation (S-POPA) and Bayesian Optimization based Probabilistic Optimal Power Allocation (BO-POPA). Briefly, S-POPA generates a set of candidate solutions and iteratively learns the solution that returns a high probability of success. On the other hand, BO-POPA applies the Bayesian optimization framework to construct a surrogate model to predict the reward value of transmit power allocations. Numerical results show that the performance of S-POPA and BO-POPA achieves on average 86.91% and 79.25% of the transmit power computed by SAA.
{"title":"Optimizing Sample Delivery in RF-Charging Multi-Hop IoT Networks","authors":"Muchen Jiang;Kwan-Wu Chin","doi":"10.1109/TGCN.2023.3311590","DOIUrl":"10.1109/TGCN.2023.3311590","url":null,"abstract":"This paper studies sample delivery in a multi-hop network where a power beacon charges devices via radio frequency (RF) signals. Devices forward samples with a deadline from a source to a sink. The goal is to minimize the power beacon’s transmit power and guarantee that samples arrive at the sink with probability \u0000<inline-formula> <tex-math>$(1-epsilon)$ </tex-math></inline-formula>\u0000 by their deadline, where \u0000<inline-formula> <tex-math>$epsilon $ </tex-math></inline-formula>\u0000 is a given probability of failure. A key challenge is that the power beacon does not have instantaneous channel gains information to devices and also between devices; i.e., it does not know the energy level of devices. To this end, we formulate a chance-constrained stochastic program for the problem at hand, and employ the sample-average approximation (SAA) method to compute a solution. We also outline two novel approximation methods: Sampling based Probabilistic Optimal Power Allocation (S-POPA) and Bayesian Optimization based Probabilistic Optimal Power Allocation (BO-POPA). Briefly, S-POPA generates a set of candidate solutions and iteratively learns the solution that returns a high probability of success. On the other hand, BO-POPA applies the Bayesian optimization framework to construct a surrogate model to predict the reward value of transmit power allocations. Numerical results show that the performance of S-POPA and BO-POPA achieves on average 86.91% and 79.25% of the transmit power computed by SAA.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-09-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62571491","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-31DOI: 10.1109/TGCN.2023.3310569
Li Zeng;Dingzhu Wen;Guangxu Zhu;Changsheng You;Qimei Chen;Yuanming Shi
Federated learning (FL) is a promising technique for distilling artificial intelligence from massive data distributed in Internet-of-Things networks, while keeping data privacy. However, the efficient deployment of FL faces several challenges due to, e.g., limited radio resources, computation capabilities, and battery lives of Internet-of-Things devices. To address these challenges, in this work, the energy harvesting technique is first enabled on Internet-of-Things devices for supporting their sustainable lifelong learning. Then, the convergence rate of the FL algorithm is derived, which is shown to depend on the data utility (defined as the number of used training samples) in each training iteration. Thus, to accelerate the convergence rate and reduce the training latency, a data utility maximization problem for each iteration is formulated, under several practical constraints on the limited time, bandwidth (i.e., number of subcarriers), computation frequency, and energy supply. The problem is mixed-integer and non-convex, and hence NP-hard. To solve the problem, an optimal joint device selection and resource allocation (JDSRA) scheme is proposed. In this scheme, a distributed on-device resource allocation problem is first solved to determine the minimum required number of subcarriers for each device, followed by a dynamic programming approach for attaining the optimal device selection policy. In particular, no global channel state information (CSI) sharing is needed to execute the scheme. Finally, extensive experiments are presented to demonstrate the performance of the proposed optimal algorithm.
{"title":"Federated Learning With Energy Harvesting Devices","authors":"Li Zeng;Dingzhu Wen;Guangxu Zhu;Changsheng You;Qimei Chen;Yuanming Shi","doi":"10.1109/TGCN.2023.3310569","DOIUrl":"10.1109/TGCN.2023.3310569","url":null,"abstract":"Federated learning (FL) is a promising technique for distilling artificial intelligence from massive data distributed in Internet-of-Things networks, while keeping data privacy. However, the efficient deployment of FL faces several challenges due to, e.g., limited radio resources, computation capabilities, and battery lives of Internet-of-Things devices. To address these challenges, in this work, the energy harvesting technique is first enabled on Internet-of-Things devices for supporting their sustainable lifelong learning. Then, the convergence rate of the FL algorithm is derived, which is shown to depend on the data utility (defined as the number of used training samples) in each training iteration. Thus, to accelerate the convergence rate and reduce the training latency, a data utility maximization problem for each iteration is formulated, under several practical constraints on the limited time, bandwidth (i.e., number of subcarriers), computation frequency, and energy supply. The problem is mixed-integer and non-convex, and hence NP-hard. To solve the problem, an optimal joint device selection and resource allocation (JDSRA) scheme is proposed. In this scheme, a distributed on-device resource allocation problem is first solved to determine the minimum required number of subcarriers for each device, followed by a dynamic programming approach for attaining the optimal device selection policy. In particular, no global channel state information (CSI) sharing is needed to execute the scheme. Finally, extensive experiments are presented to demonstrate the performance of the proposed optimal algorithm.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-08-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62571450","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-08-29DOI: 10.1109/TGCN.2023.3309657
Jianyang Ren;Wanli Ni;Hui Tian;Gaofeng Nie
As the number of sensors becomes massive in Internet of Things (IoT) networks, the amount of data is humongous. To process data in real-time while protecting user privacy, federated learning (FL) has been regarded as an enabling technique to push edge intelligence into IoT networks with massive devices. However, FL latency increases dramatically due to the increase of the number of parameters in deep neural network and the limited computation and communication capabilities of IoT devices. To address this issue, we propose a semi-federated learning (SemiFL) paradigm in which network pruning and over-the-air computation are efficiently applied. To be specific, each small base station collects the raw data from its served sensors and trains its local pruned model. After that, the global aggregation of local gradients is achieved through over-the-air computation. We first analyze the performance of the proposed SemiFL by deriving its convergence upper bound. To reduce latency, a convergence-constrained SemiFL latency minimization problem is formulated. By decoupling the original problem into several sub-problems, iterative algorithms are designed to solve them efficiently. Finally, numerical simulations are conducted to verify the effectiveness of our proposed scheme in reducing latency and guaranteeing the identification accuracy.
{"title":"Convergence Analysis and Latency Minimization for Semi-Federated Learning in Massive IoT Networks","authors":"Jianyang Ren;Wanli Ni;Hui Tian;Gaofeng Nie","doi":"10.1109/TGCN.2023.3309657","DOIUrl":"10.1109/TGCN.2023.3309657","url":null,"abstract":"As the number of sensors becomes massive in Internet of Things (IoT) networks, the amount of data is humongous. To process data in real-time while protecting user privacy, federated learning (FL) has been regarded as an enabling technique to push edge intelligence into IoT networks with massive devices. However, FL latency increases dramatically due to the increase of the number of parameters in deep neural network and the limited computation and communication capabilities of IoT devices. To address this issue, we propose a semi-federated learning (SemiFL) paradigm in which network pruning and over-the-air computation are efficiently applied. To be specific, each small base station collects the raw data from its served sensors and trains its local pruned model. After that, the global aggregation of local gradients is achieved through over-the-air computation. We first analyze the performance of the proposed SemiFL by deriving its convergence upper bound. To reduce latency, a convergence-constrained SemiFL latency minimization problem is formulated. By decoupling the original problem into several sub-problems, iterative algorithms are designed to solve them efficiently. Finally, numerical simulations are conducted to verify the effectiveness of our proposed scheme in reducing latency and guaranteeing the identification accuracy.","PeriodicalId":13052,"journal":{"name":"IEEE Transactions on Green Communications and Networking","volume":null,"pages":null},"PeriodicalIF":4.8,"publicationDate":"2023-08-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48625459","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","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：