Performance Analysis of Data Detection in the THz-Band under Channel-Correlated Noise

TL;DR

This paper develops a detailed SER analysis framework for THz-band communication systems, accounting for channel and noise correlations, hardware impairments, and providing more accurate performance predictions.

Contribution

It introduces a novel SER analysis method that considers correlated channel and noise effects, improving upon previous independent assumptions.

Findings

The framework accurately predicts SER deviations caused by channel-noise correlation.

Correlated noise from hardware impairments significantly impacts SER performance.

The analysis corrects multi-dB SER errors from traditional independence assumptions.

Abstract

We present a comprehensive symbol error rate (SER) analysis framework for link-level terahertz (THz)-band communication systems under linear zero-forcing (ZF) data detection. First, we derive the mismatched SER for indoor THz systems under independent channel and noise assumptions, calculating the probability density function of the ratio of Gaussian noise to $\alpha$-$\mu$ channels resulting from ZF filtering. Next, we derive the precise SER under correlated channel and noise conditions, modeling dependencies using the copula method. Finally, we evaluate the SER for THz channels with correlated distortion noise from hardware impairments. Simulations demonstrate that the proposed framework corrects for multi-dB SERs resulting from the channel-noise independence assumption.