Optimal Control And Performance Analysis Of Spectrum Access In Cognitive Radio System

As a non-renewable resource, the scarcity of radio spectrum becomes more and more serious with the rapid growth of wireless subscribers and the fast development of wireless services. Cognitive radio is a novel technique which is able to make efficient use of the existing wireless spectrum. It allows the unlicensed (cognitive) user to opportunistically exploit the licensed spectrum without causing significant interference to the licensed (primary) user, and therefore the overall spectral efficiency can be economically increased. However, just as every coin has two sides, this dynamic spectrum access technique becomes a challenge for the traditional spectrum access algorithm. The optimal control and performance analysis of spectrum access in cognitive radio system based on different protection models of primary user is intensively studied in this dissertation. The contents of this work are listed as follows:The resourse optimal allocation problems in both of the centralized and distributed cognitive network based on the interference temperature model are studied respectively. For the centralized cognitive network, the interference temperature constraint and the outage probability constraint of each cognitive link are modeled. The power control problem is posed as a geometric program and the quasi-optimal solution is obtained using the convex optimization theory. Then a centralized access control algorithm based on the tree pruning algorithm is proposed, and an approximate rule is given to reduce the computational complexity. For the distributed cognitive network, a distributed fixed-step power control algorithm is proposed. The protection of the primary user and the active links is proved, and the convergence property is shown. Based on this power control algorithm, a distributed access control algorithm for cognitive network is also given.For cognitive MIMO system based on the queuing stability model, the power control optimization problem and the system performance of spectrum access is analyzed. The maximum stable throughput of the cognitive link using two transmission antennas and the corresponding optimal transmission power of each antenna are derived with a fixed packet arriving rate selected by the primary link. The results are then extended to the case in which the cognitive link has N transmission antennas.We also analyze the spectrum access control problem in cognitive radio network based on the transmission collision model. For the cognitive network with limited and adjustable sensing ability, we study the optimal cooperative spectrum sensing algorithm when the OR-rule is used as the fusion rule. A heuristic algorithm is given to reduce the computational complexity, while the centralized and distributed cooperative spectrum sensing strategies based on this algorithm are designed. For the cognitive network with limited and fixed sensing ability, we proposed the opportunistic spectrum access algorithms when the OR-rule and the voting rule are used as the fusion rules respectively. In the former case, the spectrum access control strategy contains two parts:an access policy to find the tradeoff between transmission collision and overlooked opportunity in each channel, and a cooperative spectrum sensing strategy to maximize the overall transmission throughput of the cognitive network. For the latter case, we first analyze the optimal access policy of each channel when the performance of the cooperative spectrum sensing algorithm is given. Then, the optimal fusing threshold is presented and an approximated expression is also obtained to reduce the computational complexity. At last, we prove that the spectrum access control problem can be treated as a convex optimization problem with some reasonable approximation, and the spectrum access algorithm is proposed using the greedy algorithm.