Search

Opportunistic spectrum access (OSA), a promising technology to resolve radio spectrum scarcity, is still faced with some challenges, of which one is potentially frequent channel handoff (CH) even handoff blocking (HB) for secondary user (SU) to avoid primary user (PU). Reaccess is a direct way to resolve HB. However, multiple handoff blocking‐then‐reaccess (H‐BTR) are time consuming. Whether it is worthwhile to make multiple H‐BTR, is an interesting but hardly mentioned issue. To this end, the multiple H‐BTR‐based OSA are focused on here. Three key indices, the average transmission probability, handoff delay, and average reaccess times of SU in the H‐BTR‐based OSA, are deduced as close form first. Then, the effects of H‐BTR frequency and service traffic rate of SU on OSA are discussed. Finally, the H‐BTR‐based OSA is compared to two other schemes, the handoff without BTR (HWBTR)‐based OSA and the stop‐and‐waiting (SW)‐based OSA. Theoretical and simulated results show that the H‐BTR‐based OSA performs best among three schemes. Making H‐BTR in moderate frequency according to service traffic rate can increase transmission opportunity while additional delay of SU is tolerable, thus the performance of OSA can be well improved. [ABSTRACT FROM AUTHOR]

The internet of things accelerates the wireless connections of massive devices to provide all kinds of new services, and thus intensifies the spectrum scarcity and access delay. Cognitive radio (CR) technology brings a solution for the issue. As a typical CR scheme, opportunistic spectrum access (OSA) has been addressed widely in the past decade. However, the tradeoff between two key indexes, the throughput of cognitive radio network (CRN) and the delay of secondary user (SU), is rarely mentioned so far, which ignites the authors' work in this paper. Taking into account a channel handoff (CH) based multi‐channel OSA scenario, the authors first analyze the opportunistic transmission performance of SU, and model the throughput of CRN as well as the handoff delay of SU. Then, the authors build up a delay‐constraint throughput optimization problem, and thus formulate the throughput‐delay tradeoff for OSA. Finally, the optimal traffic rates of SU for a good throughput‐delay tradeoff are derived according to maximizing the throughput of CRN. Theoretical and simulated results show that to enhance the throughput and to reduce the delay do not conflict always. By well adjusting the traffic rates of SU according to the traffic rates of primary user, the throughput of CRN can be improved while the handoff delay can be kept under a given level. [ABSTRACT FROM AUTHOR]