Performance of Cognitive Hybrid Automatic Repeat reQuest: Go-Back-N

In this paper, we propose a cognitive Go-Back-N Hybrid Automatic Repeat reQuest (CGBN-HARQ) scheme for a cognitive radio (CR) system to opportunistically transmit data over a primary radio (PR) channel. We model the activity of PR users (PRUs) occupying the PR channel as a Markov chain with two states: `ON' and `OFF'. In order to use the PR channel, the CR system first senses the availability/unavailability of the PR channel. Once it finds that the PR channel is free, the CR system transmits data packets over the PR channel's spectrum, whilst relying on the principles of GBN-HARQ. In this paper, we investigate both the throughput and delay of CGBN-HARQ, with a special emphasis on the impact of various system parameters involved in the scenarios of both perfect and imperfect spectrum sensing. Our studies demonstrate that the activity of PRUs, the transmission reliability of the CR system as well as the number of packets transmitted per time-slot may have a substantial impact on both the throughput and the delay of the CR system.