Decentralized Relay Selection in Multi-User Multihop Decode-and-Forward Relay Networks

Abstract

This paper analyzes the performance of a multi-user multihop relay network using a low complexity decentralized relay selection (DRS) scheme for decode-and-forward cooperative networks. We carry out a rigorous diversity order analysis, with Nakagami- m fading and pathloss and show that the DRS scheme achieves full diversity while maintaining a complexity that is quadratic in the number of users, quadratic in the number of relays and independent of the number of hops. For a special case of two-user networks we derive exact closed-form expressions for the outage probability by considering the order statistics. Furthermore, we extend our analysis to consider interfering relay networks and derive an accurate lower bound on the outage of an arbitrary network user. Based on the lower bound we also show how the outage probability saturates in the high signal-to-interference-plus-noise ratio regime. Extensive numerical examples are used to illustrate the accuracy of the analysis and to highlight the use of the DRS scheme in multi-user multihop relay networks. © 2018 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Optimal Routing for Multi-user Multi-hop Relay Networks via Dynamic Programming

“…In addition, due to zero SNR associated with dummy relay nodes, they will not be selected during the relay selection process. As commonly used in the literature, we assume that each S-D pair is assisted by only one relay in each hop and each relay assists at most one S-D pair to minimize the synchronization requirements, to avoid too much processing complexity in any single relay and to minimize power consumption in the network [2], [12]. As one relay node supports at most one S-D pair, it is important to consider M ≥ N to ensure that all S-D pairs are able to communicate via the relay network.…”

“…For such networks, relay selection is critical in gaining advantages in terms of performance, complexity and overhead. As a result, much research has focused on relay selection (RS) in cooperative relay networks [2]- [5].…”

“…This involves a computational complexity which is exponential in the number of hops and the number of S-D pairs. As a result, different sub-optimal RS strategies have been considered to offer less computational complexity [12], decentralization [2], [5] and practical constraints [4].…”

“…Taking a different approach, in [12], the authors propose a sub-optimal RS scheme with linear complexity with respect to the number of relays for a multi-user dual-hop relay network. Later, this was extended to a multi-user, multi-hop relay network in [2], where the authors propose a decentralized RS scheme which performs RS in each hop independently, with the last two hops combined together to achieve full diversity.…”

“…The naive or brute-force approach to the optimal relay selection involves exhaustively searching through all possible paths from source nodes to destination nodes. Therefore, in many of the previous works that provide sub-optimal solutions, the simulation results were either limited to smaller networks or performance was not compared against the optimal solution [2]. The main contribution of this work is to provide a centralized solution that is guaranteed to obtain the optimal relay selection with linear complexity in terms of the number of hops.…”