Throughput and Delay Analysis of Cognitive Go-Back-N Hybrid Automatic Repeat reQuest Using Discrete-Time Markov Modelling

Cognitive radio (CR) techniques have been proposed for improving the spectral efficiency by exploiting the temporarily unoccupied segments of the licensed spectrum, provided that the transmission of primary users (PUs) is not hampered. In this paper, we propose a cognitive Go-Back-N Hybrid Automatic Repeat reQuest (CGBN-HARQ) scheme that enables the cognitive user (CU) to opportunistically transmit data over a primary radio (PR) channel. Based on the sensing decisions by the CU, it decides about the availability of the PU’s channel for its own transmission using the proposed CGBN-HARQ scheme. In addition, it enables the CR transmitter to receive feedback concerning the success/failure of its prior transmissions during the sensing and transmission phases of the time-slot (TS). A discrete time Markov chain model is invoked for the theoretical analysis of the proposed system, where we conceive an algorithm to generate all possible states of the CR transmitter. Both the throughput and delay of the CGBN-HARQ scheme is analyzed by deriving a range of closed-form formulas, which are validated by simulation results. The occupation of the channel by the PU and the reliability of the CU’s channel significantly affect both the achievable throughput and the delay of the CGBN-HARQ scheme. Finally, our studies show that the number of packets transmitted within a TS should be adapted according to the communication channel for attaining the maximum throughput and the lowest average transmission delay.