Nonlinear Equalization for RF Receivers

This paper describes the need for high performance computing (HPC) to facilitate the development and implementation of a nonlinear equalizer that is capable of mitigating and/or eliminating nonlinear distortion to extend the dynamic range of radar front-end receivers decades beyond the analog state-of-the-art. The search space for the optimal nonlinear equalization (NLEQ) solution is computationally intractable using only a single desktop computer. However, we have been able to leverage a combination of an efficient greedy search with the high performance computing technologies of LLGrid and MatlabMPI to construct an NLEQ architecture that is capable of extending the dynamic range of radar front-end receivers by over 25dB