Performance Analysis of a Cooperative Wireless Network with Adaptive Relays

TL;DR

This paper analyzes a cooperative wireless network with adaptive relays and multipacket reception, focusing on stability, throughput, and delay performance, and introduces a state-dependent retransmission protocol for improved efficiency.

Contribution

It presents a novel analysis of a relay-assisted cooperative network with adaptive relays and MPR, deriving stability regions, delay expressions, and employing boundary value problem techniques.

Findings

Enhanced delay performance with adaptive relay protocol

Explicit delay expressions for symmetric case

Numerical results demonstrate system stability and throughput gains

Abstract

In this work, we investigate a slotted-time relay assisted cooperative random access wireless network with multipacket (MPR) reception capabilities. MPR refers to the capability of a wireless node to successfully receive packets from more than two other modes that transmit simultaneously at the same slot. We consider a network of $N$ saturated sources that transmit packets to a common destination node with the cooperation of two infinite capacity relay nodes. The relays assist the sources by forwarding the packets that failed to reach the destination. Moreover, the relays have also packets of their own to transmit to the destination. We further assume that the relays employ a state-dependent retransmission control mechanism. In particular, a relay node accordingly adapts its transmission probability based on the status of the other relay. Such a protocol is towards self-aware networks and leads to substantial performance gains in terms of delay. We investigate the stability region and the throughput performance for the full MPR model. Moreover, for the asymmetric two-sources, two-relay case we derive the generating function of the stationary joint queue-length distribution with the aid of the theory of boundary value problems. For the symmetric case, we obtain explicit expressions for the average queueing delay in a relay node without solving a boundary value problem. Extensive numerical examples are presented and provide insights on the system performance.