Abstract
Utilizing ultra-wide bandwidths is a promising approach to achieve the terabits per second (Tbps) wireless links required to unlock emerging mobile applications such as holographic telepresence. However, conventional digital systems face significant challenges to exploit such ultra-wide bandwidths efficiently. On the one hand, utilizing ultra-fast, high-precision analogue-to-digital converters (ADCs) required for ultra-wide bandwidth systems becomes impractical in terms of power consumption. On the other hand, performing ultra-fast receiver processing becomes extremely challenging in terms of power consumption and processing latency not only due to the complexity of state-of-the-art signal processing algorithms but also because the increased sampling rates challenge the speed capabilities of modern digital processors. To overcome these bottlenecks, the Digilogue processing has been recently proposed, which suggests to perform signal processing directly in the analogue domain without requiring ADCs and traditional digital signal processing. While initial results show that DigiLogue receivers can potentially achieve substantial power gains, they focus on oversimplified evaluations that are far from being practical since they only account for simple repetition codes and additive white Gaussian noise (AWGN) channels. In this work, we propose a DigiLogue receiver design suitable for flat-fading channels, which is of very high interest in the case of systems operating at higher frequencies where ultra-wide bandwidths can be available. In contrast to existing analogue solutions, the proposed design can perform "soft" channel equalization/detection and decoding directly in the analogue domain and achieve the error rate performance of the corresponding digital systems, but with power savings of more than36 × .