Power Allocation for NOMA-Based Visible Light Communication Systems with DQN

Jiawei Deng, Xuan Tang, Xian Wei, Pu Li, Jiaqi Li, Xicong Li, Zabih Ghassemlooy

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    1 Citation (Scopus)

    Abstract

    The spectral efficiency of the visible light communication system can be enhanced by using the non-orthogonal multiple access (NOMA) scheme. In this paper, we propose a deep Q network (DQN) framework-based power allocation scheme that maximizes the sum rate of a NOMA-based cellular VLC network with mobility support. The numerical results indicate that the optimisation process achieves improved performance in terms of sum data rate (SDR) by approximately 8.8, 7, and 4% compared with conventional algorithms, such as fixed power allocation, gain ratio power allocation, and genetic algorithms.
    Original languageEnglish
    Title of host publication2024 14th International Symposium on Communication Systems, Networks and Digital Signal Processing (CSNDSP)
    PublisherIEEE
    Pages512-517
    Number of pages6
    ISBN (Electronic)9798350348743
    ISBN (Print)9798350348750
    DOIs
    Publication statusPublished - 19 Jul 2024
    Event2024 14th International Symposium on Communication Systems, Networks and Digital Signal Processing (CSNDSP) - Rome, Italy
    Duration: 17 Jul 202419 Jul 2024
    Conference number: 14

    Conference

    Conference2024 14th International Symposium on Communication Systems, Networks and Digital Signal Processing (CSNDSP)
    Abbreviated titleCSNDSP
    Country/TerritoryItaly
    CityRome
    Period17/07/2419/07/24

    Keywords

    • Deep reinforcement learning
    • NOMA
    • visible light communication
    • power allocation

    Cite this