Combining Channel Sounding and Guessing Random Additive Noise Decoding

Abstract

This thesis investigates how channel estimation can be used to improve the performance of Guessing Random Additive Noise Decoding. The trade-off between devoting resources to channel sounding and data transmission is investigated for pilot symbol assisted modulation schemes. Using a soft-information variant of the GRAND algorithm called Ordered Reliability Bit Guessing Random Additive Noise Decoding- Approximate Independence (ORBGRAND-AI), it is shown that by accounting for the correlation between received symbols bit and block error rate improvements can be obtained. This thesis also considers the achievable communications rate of ORBGRAND-AI when different estimators are used to provide channel estimates. Finally, this thesis investigates the use of ORBGRAND-AI in channels subjected to inter-symbol interference (ISI).