On the Optimal Detection and Error Performance Analysis of the Hardware Impaired Systems

Abstract

The conventional minimum Euclidean distance (MED) receiver design is based on the assumption of ideal hardware transceivers and proper Gaussian noise in communication systems. Throughout this study, an accurate statistical model of various hardware impairments (HWIs) is presented. Then, an optimal maximum likelihood (ML) receiver is derived considering the distinct characteristics of the HWIs comprised of additive improper Gaussian noise and signal distortion. Next, the average error probability performance of the proposed optimal ML receiver is analyzed and tight bounds are derived. Finally, different numerical and simulation results are presented to support the superiority of the proposed ML receiver over MED receiver and the tightness of the derived bounds.