Abstract
In cognitive radio (CR) networks, to avoid collision with the primary users, the secondary users (SUs) must sense the existence of the primary signals in each fixed time period, before transmitting their signals. For a multiple primary channels (PCs) scenario, the SUs are generally allowed to utilize all the idle PCs in order to maximize the system throughput. In the conventional approach, the SUs sense all the PCs to maximize the number of channels available for data transmission. However, the increase in the number of PCs sensed by the CR network will increase the time consumption in the sensing phase, and thus degrades the time available for data transmission. The goal of this work is to find the pre-determined optimal number of channels utilized for data transmission to maximize the system throughput. Compared to the real-time adaptation approach, our proposed scheme can be easily applied to practical systems and achieve a throughput very close to the ultimate system throughput, but without the need of real-time computation burden. © 2012 IEEE.