Deadline-Constrained Random Access in Cognitive Radio Networks with Multiple-Packet Reception.

Chengran Li,Yidong Guo,Yuan Shen,Ting Fang,Aoyu Gong,Yijin Zhang
DOI: https://doi.org/10.1109/icct56141.2022.10072590
2022-01-01
Abstract:Deadline-constrained random access plays an essential role in cognitive radio networks (CRNs) where secondary users (SUs) opportunistically access the licensed spectrum unoccupied by primary users (PUs) to transmit packets for real-time services. In this paper, we focus on the frame-synchronized traffic under multiple-packet reception that allows more than one packet to be successfully received simultaneously. A basic protocol structure of dynamic slotted-ALOHA that jointly accounts for the current delivery urgency, the local knowledge of contention intensity, the MPR capability and the PUs activity is presented. Assuming always incompletely known contention intensity, in order to maximize the timely delivery ratio (TDR), we formulate this access problem as a partially observable Markov decision process (POMDP), and design two low-complexity suboptimal policies based on this POMDP formulation. Simulation results verify the TDR advantage of the proposed policies, and show that the myopic policy is in general not optimal.
What problem does this paper attempt to address?