Research On The Degree Of Freedom Of The Cognitive Radio Based On Interference Alignment

With the development of wireless communication, spectrum is facing greater pressurethan before. On the one hand, the use of unlicensed spectrum has been saturated. In face ofthis situation, cognitive radio technology was proposed because it can make part ofunlicensed users communicate by using the bandwidth of licensed users, under the premiseof no effect on the communication among primary network. On the other hand, along withthe increase of the factor of frequency reuse, the interference signal in wireless network isbecoming more complex. Interference alignment (IA) can separate the desired signal andthe interference signal by designing the pre-coding matrix and the post-processing matrixjointly, thus bring more multiplexing gain. The spectrum efficiency can be maximized bycombining cognitive radio and interference alignment, which attracts more attention,recently.In this paper, we explore the interference alignment technology in cognitive radio, andour main work is the feasibility of interference alignment and the analysis of degree offreedom. First, we divide the problem into two cases according to ignoring the interferenceamong the primary users or not. Then, we discuss the feasibility of IA in both the two casesseparately.When ignoring the interference among the primary users, we propose a necessarycondition and further prove it by using mapping theory and algebraic geometry. Then,based on the necessary condition, we can obtain the upper bound of the average degree offreedom (Dof) of secondary users. The analysis of some cases and the numerical resultsshow that the necessary condition and the upper bound of Dof obtained by the conditionare in line with previous conclusion.When considering the interference among the primary users, the feasibility problem ofIA is divided into two sub-problems—the feasibility problem of IA in primary network andthe feasibility problem of IA in secondary network under the premise that the feasibilitycondition of IA in primary network is satisfied. After investigate the whole problem, wefind that the former sub-problem is equivalent to the feasibility of IA in multi-user MIMO network, while the latter is consistent with the feasibility of IA in cognitive radio whenignoring the interference among the primary network. According to the previousconclusion, we present the necessary conditions of the two sub-problems respectively.