Enhancing Iterative Receivers with Spatial Coupling

Research Project, 2018 – 2021

Iterative (turbo) receivers, dealing with receiver tasks of various kinds such as decoding, equalization, synchronization, and detection for multi-antenna and multiuser systems, have become ubiquitous in modern communication systems. While the optimal receiver is prohibitely complex, iterative receivers achieve high performance gains with respect to the classical approach where the receiver tasks are performed independently after each other, yet with affordable complexity. 

Surprisingly, besides iterative decoding, whose theoretical foundations and performance limits are nowadays quite well established, the limitations of more general iterative receiver structures are still far from being understood. Fundamental questions such as how much is lost by iterative processing compared to an optimal receiver, what are the achievable limits for low-complexity iterative algorithms, and how do they depend on the components involved, remain still open. The aim of this project is to unveil these challenging questions using analysis tools borrowed from modern coding theory. A major step in this project is to investigate the recent concept of spatial coupling to enhance the performance of iterative receivers, and as a tool to derive fundamental limits on their asymptotic performance.