Spectrum access strategies based on the ratio of throughput to spectrum consumption in the context of spectrum fragmentation in distributed cognitive radio networks

Because of the random presence of primary users in cognitive radio networks, the available spectrums are fragmented with the size of each fragment varying from one channel to several channels. Due to the spectrum fragmentation, the cross-band interference between spectrally adjacent users is considered harmful in the distributed cognitive radio networks. To eliminate the interference, frequency guard-bands are inserted between spectrum fragments. The guard-bands not used for transmission are considered as the spectrum consumption. Accordingly, the achieved throughput is at the cost of the guard-bands consumption. Thus, to utilize the spectrum fragments efficiently, the ratio of throughput to spectrum consumption is defined for the dynamic spectrum access problem. Therefore, under the conditions of prescribed user quality-of-service (QoS) requirements and the frequency agility limit, a novel mathematic model in which the ratio of throughput to spectrum consumption is maximized is proposed in this paper, which is an integer programming problem. To this end, three spectrum access strategies with lower complexities than the traditional exhaustive search are proposed. Simulation results show that the ratio can be used as the index whether the spectrum fragments are effectively utilized. Meanwhile, the spectrum consumption is mainly affected by the size of guard-band, number of fragments and available channels, so the ratios are varying with these factors. Moreover, as the different spectrum utilizations of the four strategies, these factors also have significant influences on the performances of the strategies.