Network Association in Machine-Learning Aided Cognitive Radar and Communication Co-Design

Jingjing Wang,Sanghai Guan,Chunxiao Jiang,Dimitrios Alanis,Yong Ren,Lajos Hanzo

DOI: https://doi.org/10.1109/jsac.2019.2933778

IF: 16.4

2019-01-01

IEEE Journal on Selected Areas in Communications

Abstract:In order to beneficially exploit the scarce wireless spectral resources, spectrum sharing between communication and radar systems has become a promising research topic. However, traditional network association strategies may not result in efficient hybrid communication and radar systems. We circumvent this problem by formulating a partially observable Markov decision processes (POMDP) aided network association scheme, where the radar user acts as the primary user (PU), while the cognitive communication user is the secondary user (SU). For maximizing the network throughput, whilst minimizing the interference imposed on the radar user, the communication user is configured for adaptively selecting its underlay or overlay access mode. Moreover, a low-complexity near-optimal reinforcement learning algorithm is proposed for the co-design by considering both its complexity and feasibility. Finally, we quantify the performance of our proposed POMDP based network association scheme.