eduzhai > Applied Sciences > Engineering >

Optimization-driven Hierarchical Learning Framework for Wireless Powered Backscatter-aided Relay Communications

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 29 pages

Abstract: In this paper, we employ multiple wireless-powered relays to assistinformation transmission from a multi-antenna access point to a single-antennareceiver. The wireless relays can operate in either the passive mode viabackscatter communications or the active mode via RF communications, dependingon their channel conditions and energy states. We aim to maximize the overallthroughput by jointly optimizing the access point s beamforming and the relays radio modes and operating parameters. Due to the non-convex and combinatorialstructure, we develop a novel optimization-driven hierarchical deepdeterministic policy gradient (H-DDPG) approach to adapt the beamforming andrelay strategies dynamically. The optimization-driven H-DDPG algorithm firstlydecomposes the binary relay mode selection into the outer-loop deep Q-network(DQN) algorithm and then optimizes the continuous beamforming and relayingparameters by using the inner-loop DDPG algorithm. Secondly, to improve thelearning efficiency, we integrate the model-based optimization into the DDPGframework by providing a better-informed target estimation for DNN training.Simulation results reveal that these two special designs ensure a more stablelearning and achieve a higher reward performance, up to nearly 20 , compared tothe conventional DDPG approach.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×