A two-phase and two-period spectrum sensing scheme using high-layer information for cognitive radio networks

Spectrum sensing is important for dynamic spectrum allocation in cognitive radio networks (CRNs). Spectrum sensing algorithms are used to determine the frequency of spectrum sensing and the order for the channels to be scanned. In this paper, we propose a two-phase and two-period spectrum sensing (TTSS) scheme using high-layer information for CRNs. Benefiting from advanced physical-layer technologies, the TTSS scheme uses the results of coarse sensing to predict the best candidate channels for fine sensing and it adopts two types of sensing periods to optimize the network performance. When a node is not transmitting data, a long sensing period is used to reduce the sensing overhead. When a node is transmitting data, a short sensing period is used to reduce the average time interval where secondary users collide with primary users. Moreover, high-layer information is used to adjust the two sensing periods to further reduce the sensing overhead and increase the sensing accuracy. On average, simulation results show that the TTSS scheme with smaller short sensing period has lower packet dropping rate of primary users by 3.6% than coarse sensing scheme and by 1.17% than fine sensing scheme. Also, the TTSS scheme has medium sensing overhead and higher throughput of secondary user by 14% than coarse sensing scheme and by 197.21% than fine sensing scheme.