Cooperative Relaying in Wireless Networks under Spatially and Temporally Correlated Interference

TL;DR

This paper investigates the performance of cooperative relaying in wireless networks with interference that varies across space and time, providing analytical insights and design guidelines for different environmental conditions.

Contribution

It derives analytical expressions for success probability and transmission distribution considering interference correlation, and highlights the impact of relay placement, mobility, and combining schemes.

Findings

Maximal ratio combining benefits are location-dependent.

More relays improve performance in harsh interference environments.

Interferer mobility enhances system performance.

Abstract

We analyze the performance of an interference-limited, decode-and-forward, cooperative relaying system that comprises a source, a destination, and $N$ relays, placed arbitrarily on the plane and suffering from interference by a set of interferers placed according to a spatial Poisson process. In each transmission attempt, first the transmitter sends a packet; subsequently, a single one of the relays that received the packet correctly, if such a relay exists, retransmits it. We consider both selection combining and maximal ratio combining at the destination, Rayleigh fading, and interferer mobility. We derive expressions for the probability that a single transmission attempt is successful, as well as for the distribution of the transmission attempts until a packet is transmitted successfully. Results provide design guidelines applicable to a wide range of systems. Overall, the temporal and spatial characteristics of the interference play a significant role in shaping the system performance. Maximal ratio combining is only helpful when relays are close to the destination; in harsh environments, having many relays is especially helpful, and relay placement is critical; the performance improves when interferer mobility increases; and a tradeoff exists between energy efficiency and throughput.