Research On Hybrid Resource Allocation Algorithm For OFDMA-based Cognitive Radio Networks

Radio spectrum is a scarce resource in wireless communication networks in recent years. However, according to the survey of FCC(Federal Communications Commission) shows that the utilization of traditional spectrum allocation is very low. In order to realize maximum use of this resource, cognitive radio(CR) has been proposed to improve the spectral efficiency by exploiting unused spectrum. CR system can dynamically sense changing environment, and make secondary users(SUs) use the unoccupied spectrum by primary users(PUs) to communicate under certain transmit power control scheme. During the communication, CR system needs to determine some transmission characteristics(e.g., capacity, power, and bandwidth). The interference generated by SUs to PUs is the main factor affecting the transmission quality of cognitive radio networks.Orthogonal Frequency Division Multiplexing(OFDM) has been considered as a promising modulation candidate for CR networks, mainly because of its flexibility in dynamically allocating radio resource to multiple users as well as its low interference between adjacent subcarriers. In OFDM system, due to the non-orthogonality of the transmitted signals, there is mutual interference between PUs and SUs. The traditional resource allocation algorithm will generate excessive interference to PUs.In classical CR networks, there are two spectrum sharing approaches for SUs: spectrum overlay, which is an opportunistic spectrum access mode where SUs are allowed to access the licensed bands only when they are not used by PUs. And maximize the downlink capacity of SUs while keeping the interference introduced to the band of PUs and the total transmit power under a specified threshold. Spectrum underlay, which is a spectrum sharing paradigm, where SUs coexist with PUs and transmit power under the interference temperature constraints. Because the only spectrum overlay and underlay cannot make full use of the spectrum, some scholars propose the resource allocation in hybrid overlay/underlay mode.In the existing hybrid resource allocation algorithms, they only consider the interference introduced by SUs to the band of PUs without taking the interference caused by the underlay fashion to the sub-channel into account. And ignore the imperfect spectrum sensing. The subcarrier allocation is the first step in resource allocation. Some subcarrier allocation algorithms according to the channel gain from the SU transmitter to the SU receiver allocate the subcarrier for SU. And some subcarrier allocation algorithms consider both the channel gain and the transmission rate. But in these subcarrier allocation algorithms, they all ignore the effect of the initial power allocation. Currently, research on joint resource allocation algorithms all assume that the channel gain is known, but in actual communication scenario, the uncertainty of the transmission channel will have an impact on cognitive radio networks.In this paper, a resource allocation problem for a hybrid overlay and underlay spectrum sharing OFDMA-based CRN is studied under imperfect spectrum sensing. Combing the interference introduced by adjacent users, channel conditions and the initial power allocation, a subcarrier allocation scheme is designed to improve the fairness of each SU. Then, a power allocation algorithm is proposed to maximize the sum capacity of secondary network while keeping the QoS for both PUs and SUs under the consideration of sensing errors. Specifically, according to the Bayes formulation, both false-alarm probability and mis-detection probability are simultaneously considered in the interference temperature constraint of each subcarrier. The analytical solutions of power allocation are obtained by Lagrange multiplier and KKT conditions. In addition, the analysis of the robustness of this power allocation algorithm is presented when considering the uncertainty of the channel gain between SU and PU in the sense of the Worst-Case.Simulation results show that, under the same conditions, the optimal resource allocation algorithm and the suboptimal algorithm proposed in this paper is greater than that of the overlay only and the underlay only schemes. Subcarrier allocation algorithm can ensure the interests of each SU. Through analyzing the uncertainty of channel between the SUs and the PUs, can make the communication of cognitive radio networks more stable.