GMM-based Symbol Error Rate Prediction for Multicarrier Systems with Impulsive Noise Suppression

TL;DR

This paper develops a GMM-based analytical framework for predicting the symbol error rate in OFDM systems with impulsive noise suppression, covering various noise models and channel conditions, validated by simulations.

Contribution

It introduces a unified GMM-based approach for SER prediction in OFDM systems with impulsive noise suppression, including multiple noise models and fading channels, providing closed-form expressions.

Findings

GMM accurately models impulsive noise in OFDM systems.

Analytical SER expressions match simulation results across scenarios.

Method applies to various noise models and channel conditions.

Abstract

Theoretical analysis of orthogonal frequency division multiplexing (OFDM) systems equipped at the receiver by a non-linear impulsive noise suppressor is a challenging topic in communication systems. Indeed, although an exact closed-form expression for the output signal-to-noise ratio (SNR) of such OFDM systems is available for widely used impulsive noise models, theoretical analysis of the associated symbol error rate (SER) is still open. So far, the analytical SER expressions available in the literature approximate the time-domain impulsive noise, as Gaussian distributed in the discrete frequency domain. Conversely, this work presents an accurate analysis of the distortion noise at the nonlinearity output exploiting a Gaussian mixture model (GMM). By using GMMs we unified the approach of SER prediction for unmitigated systems, as well as for the mitigated ones, equipped by non-linear impulsive noise suppressors, including blanking, clipping, clipping-blanking and attenuating processors. Closed form expressions for the SER are derived both for non-fading and frequency-selective Rayleigh and Rician fading channels affected by impulsive noise which is represented by GMMs, thus including Bernoulli-Gaussian (BG), Middleton Class-A, as well as (approximated) alpha-stable noise. Theoretic SER performance are compared with simulations, showing very good agreement for all the impulsive noise scenarios and the non-linear suppressors.