Particle Filtering Applied to Turbo Equalization with Imperfect Channel Estimation

Alberto Guimarães

DOI: 10.14209/sbrt.2007.31193

Evento: XXV Simpósio Brasileiro de Telecomunicações (SBrT2007)

Keywords: Equalization Turbo Equalization Particle Filtering Sequential Importance Sampling Fixed-lag Smoothing

Abstract

In this article we propose applying the particle filtering technique to derive new turbo equalization schemes under conditions of imperfect channel estimation. We develop a Soft Input-Soft Output (SISO) equalizer that exploits the soft information provided in the iterative process in order to perform fixed-lag smoothing of the transmitted bits, on the grounds of Sequential Importance Sampling (SIS). Preliminary simulation results of performance evaluation and comparison against another alternative recently proposed are provided, for some scenarios of channel estimation errors. It is shown that the scheme herein presented performs significantly better when the estimation error is high.

Download