Deep Learning Framework for Two-Way MISO Wireless-Powered Interference Channels

被引:4
作者
Lee, Kisong [1 ]
Lee, Woongsup [2 ]
机构
[1] Dongguk Univ, Dept Informat & Commun Engn, Seoul 04620, South Korea
[2] Yonsei Univ, Grad Sch Informat, Seoul 03722, South Korea
基金
新加坡国家研究基金会;
关键词
Deep learning; energy harvesting; simultaneous wireless information and power transmission; two-way communication; MISO wireless-powered interference channel; RESOURCE-ALLOCATION; COMMUNICATION-NETWORKS; SECURE COMMUNICATION; RELAY NETWORKS; TRANSMIT POWER; INFORMATION; SWIPT; OPTIMIZATION; MAXIMIZATION; QOS;
D O I
10.1109/TWC.2023.3243611
中图分类号
TM [电工技术]; TN [电子技术、通信技术];
学科分类号
0808 ; 0809 ;
摘要
In this paper, we present a realistic and novel protocol for two-way communication in multiple-input-single-output (MISO) wireless-powered interference channels, namely, a simultaneous wireless information and power transfer (SWIPT) then wireless information transfer (WIT) protocol. In the protocol considered, transmitters first perform SWIPT in a forward link (FL) and receivers and then execute WIT using the harvested energy in a backward link (BL). Given that the operation of SWIPT in the FL affects the performance of WIT in the BL, our aim is to find a resource allocation strategy that maximizes the sum spectral efficiency (SE) of the FL and BL, which requires the joint optimization of the transmit beamforming vector, transmit power, and energy harvesting (EH) ratio in the FL and the receive beamforming vector in the BL. To deal with the non-convexity of the optimization problem, a deep learning (DL) framework is devised, in which the optimal resource allocation strategy is approximated by a well-designed deep neural network (DNN) model consisting of six independent DNN modules where the sigmoid and softmax functions are jointly utilized to properly model each control parameter. Furthermore, a two-stage training method is proposed where the DNN model is initialized using suboptimal solutions that are found using a low complexity algorithm in a supervised manner before it is fine-tuned using the main training based on unsupervised learning. Through intensive simulations performed in various environments, we confirm that the proposed method improves the training performance of the DNN model while reducing the training overhead. As a result, the proposed DL-based resource allocation achieves a near-optimal performance in terms of the sum SE with a low computation time.
引用
收藏
页码:6459 / 6473
页数:15
相关论文
共 65 条
  • [41] Wireless Information Transfer with Opportunistic Energy Harvesting
    Liu, Liang
    Zhang, Rui
    Chua, Kee-Chaing
    [J]. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, 2013, 12 (01) : 288 - 300
  • [42] On Downlink Resource Allocation for SWIPT in Small Cells in a Two-Tier HetNet
    Lohani, Sudha
    Hossain, Ekram
    Bhargava, Vijay K.
    [J]. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, 2016, 15 (11) : 7709 - 7724
  • [43] Wireless Networks With RF Energy Harvesting: A Contemporary Survey
    Lu, Xiao
    Wang, Ping
    Niyato, Dusit
    Kim, Dong In
    Han, Zhu
    [J]. IEEE COMMUNICATIONS SURVEYS AND TUTORIALS, 2015, 17 (02): : 757 - 789
  • [44] Throughput Maximization in Full-Duplex Dual-Hop Wireless Powered Communication Networks
    Lyu, Bin
    Qi, Ting
    Guo, Haiyan
    Yang, Zhen
    [J]. IEEE ACCESS, 2019, 7 : 158584 - 158593
  • [45] Relaying Protocols for Wireless Energy Harvesting and Information Processing
    Nasir, Ali A.
    Zhou, Xiangyun
    Durrani, Salman
    Kennedy, Rodney A.
    [J]. IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, 2013, 12 (07) : 3622 - 3636
  • [46] WIRELESS POWERED COMMUNICATION NETWORKS: RESEARCH DIRECTIONS AND TECHNOLOGICAL APPROACHES
    Niyato, Dusit
    Kim, Dong In
    Maso, Marco
    Han, Zhu
    [J]. IEEE WIRELESS COMMUNICATIONS, 2017, 24 (06) : 88 - 97
  • [47] UAV-Aided Wireless Powered Communication Networks: Trajectory Optimization and Resource Allocation for Minimum Throughput Maximization
    Park, Junhee
    Lee, Hoon
    Eom, Subin
    Lee, Inkyu
    [J]. IEEE ACCESS, 2019, 7 : 134978 - 134991
  • [48] In Search of Energy-Efficient Mobile Networking
    Pentikousis, Kostas
    [J]. IEEE COMMUNICATIONS MAGAZINE, 2010, 48 (01) : 95 - 103
  • [49] Learning to Optimize: Training Deep Neural Networks for Interference Management
    Sun, Haoran
    Chen, Xiangyi
    Shi, Qingjiang
    Hong, Mingyi
    Fu, Xiao
    Sidiropoulos, Nicholas D.
    [J]. IEEE TRANSACTIONS ON SIGNAL PROCESSING, 2018, 66 (20) : 5438 - 5453
  • [50] Minimum Throughput Maximization for Multi-UAV Enabled WPCN: A Deep Reinforcement Learning Method
    Tang, Jie
    Song, Jingru
    Ou, Junhui
    Luo, Jingci
    Zhang, Xiuyin
    Wong, Kai-Kit
    [J]. IEEE ACCESS, 2020, 8 : 9124 - 9132