The enormous developments in the field of wireless communication technologies have made Q1 the unlicensed spectrum bands crowded, resulting uncontrolled interference to Q2 the traditional wireless network applications. On the other hand, licensed spectrum bands are almost completely allocated to the licensed users also known as Primary users (PUs). This dilemma became a blackhole for the upcoming innovative wireless network applications. To mitigate this problem, the cognitive radio (CR) concept emerges as a promising solution for reducing the spectrum scarcity issue. The CR network is a low cost solution for efficient utilization of the spectrum by allowing secondary users (SUs) to exploit the unoccupied licensed spectrum. In this paper, we model the PU's utilization activity by a two-state Discrete-Time-Markov Chain (DTMC) (i.e., Free and busy states), for identifying the temporarily unoccupied spectrum bands,. Furthermore, we propose a Cognitive Radio Sense-and-Wait assisted HARQ scheme, which enables the Cluster Head (CH) to perform sensing operation for the sake of determining the PU's activity. Once the channel is found in free state, the CH advertise control signals to the member nodes for data transmission relying on Stop-and-Wait Hybrid-Automatic Repeat-Request (SW-HARQ). By contrast, when the channel is occupied by the PU, the CH waits and start sensing again. Additionally, the proposed CRSW assisted HARQ scheme is analytical modeled, based on which the closed-form expressions are derived both for average block delay and throughput. Finally, the correctness of the closed-form expressions are confirmed by the simulation results. It is also clear from the performance results that the level of PU utilization and the reliability of the PU channel have great influence on the delay and throughput of CRSW assisted HARQ model.