Optimized Asymmetric Feedback Detection for Rate-adaptive HARQ with Unreliable Feedback

Abstract

This work considers downlink incremental redundancy Hybrid Automatic Repeat Request (IR-HARQ) over unreliable feedback channels. Since the impact of positive feedback (i.e., ACK) error is smaller than that of negative feedback (i.e., NACK) error, an asymmetric feedback detection scheme is proposed to protect NACK and further reduce the outage probability. We formulate the HARQ process as a Markov Decision Process (MDP) model to adapt to the transmission rate of each transmission attempt without enriched feedback and additional feedback cost. We aim to optimize the performance of HARQ process under certain outage probability requirements by finding optimal asymmetric detection thresholds. Numerical results obtained on the downlink Rayleigh fading channel and 5G new radio (NR) PUCCH feedback channel show that by applying asymmetric feedback detection and adaptive rate allocation, higher throughput can be achieved under outage probability limitations.