Graph-based Detection of Multiuser Impulse Radio Systems

TL;DR

This paper introduces a graph-based approach for multiuser detection in impulse radio systems, leveraging factor graphs and iterative algorithms to improve performance with low complexity.

Contribution

It presents a novel graphical model framework for multiuser detection in impulse radio, accommodating general coding schemes and demonstrating significant performance gains.

Findings

Large performance gains observed in simulations

Detectors applicable to any binary linear coding scheme

Low-complexity iterative detection methods

Abstract

Impulse-Radio (IR) is a wideband modulation technique that can support multiple users by employing random Time-Hopping (TH) combined with repeated transmissions. The latter is aimed at alleviating the impact of collisions. This work employs a graphical model for describing the multiuser system which, in turn, facilitates the inclusion of general coding schemes. Based on factor graph representation of the system, several iterative multiuser detectors are presented. These detectors are applicable for any binary linear coding scheme. The performance of the proposed multiuser detectors is evaluated via simulations revealing large gains with low complexity.