Research On Power Allocation In Cognitive Radio Networks

With the rapid development of wireless communication technology, the demand for wireless services is increasing. As relatively wide bandwidth must be occupied in many wireless applications, spectrum that is unrenewable and precious becomes rare resource. Researchers have to face a dilemma:on one hand, the confliction between spectrum supply and demand is increasingly obvious; on the other hand, current spectrum management policy leaves a large portion of licensed spectrum severely under-utilized. Under this background, cognitive radio technology, which can reuse the licensed spectrum, is proposed. Cognitive radio is an intelligent wireless communication system, which can finish dynamic spectrum access and recycle the allocated spectrum with no interference to primary users, leading to improved efficiency of spectrum utilization.To share spectrum with primary users (PUs), interference from secondary users (SUs) must be deminished or limited, thus power control or power allocation becomes an effective way to avoid the interference to PUs. For the purpose of guaranteeing the quality of service (QoS) of PUs, the transmit power of SUs must be controlled at sufficiently low levels. However, relatively low transmit power of SUs might affect its corresponding QoS. As a result, the QoS of SUs should be improved under the interference temperature limit. Researches on power control and allocation technology become extremely significant, as they are viewed as one of the key technologies in cognitive radio. In this dissertation, power control and allocation technologies in cognitive radio networks are researched thoroughly with the aids of mathematical optimization theory and game theory.The main work and contributions of this dissertation can be generalized as follows:(1) With game theory applied to power control in cognitive networks, two distributed power control algorithms are proposed.1) Based on the non-cooperative game mechanism, the secondary users are regarded as game players, and effects of both the SUs’throughput and its interference temperature limit are considered in the utility function. Through the game of SUs, the Nash equilibrium can be achieved. The sufficient and unique conditions for the existence of Nash equilibrium are derived, and a distributed power control algorithm is proposed under this background.2) In order to slove the near and far unfair problem, Stackelberg game mechanism is adopted furtherly to build a new spectrum renting model. In this model, the PUs act as leaders, and SUs act as followers. When SUs reuse the licensed spectrum, they are charged per unit of interference power induced to corresponding PU. By price adjusting, PU can achieve maximal benefits and keep the total induced interference power from SUs under the interference temperature limit. According to the price specified by PU, all SUs compete with each other based on the non-cooperative game theory. Finally a simple distributed power control algorithm is proposed.Simulations show that, compared with the centralized global optimal power control algorithm, the proposed distributed power control algorithm not only can obtain almost same performance, but also only few communication overhead are required, making our algorithm superior to the global optimal one.(2) For resource allocation problem in amplified and forward (AF) cognitive one-way relay network scenario, several simple suboptimal and equal power allocation algorithms are presented. Meanwhile, a near opimal power allocation algorithm is advanced for decode and forward (DF) cognitive one-way relay network scenario. Based on all these power allocation algorithms, an optimal relay selection scheme is also proposed.1) In the multiuser OFDMA cognitive one-way AF relay networks, the power allocation problem is discussed in order to maximize the sum rate of SUs under the constraint of interference temperature limit. Considering the power allocation problem among multiusers sharing the same subchannel is very difficult to solve, it is resolved into two steps: first, channel allocation is implemented, and then, power allocation is implemented. At the same time, the original optimization problem is decomposed into two suboptimal subproblems so as to simplify the process of power allocation. Because the closed form solution of the power allocation problem is not attainable, the bisection search iteration algorithm is adopted here to get the optimal transmit power for cognitive nodes. Furthermore, to make the power allocation easy to implement in real-time, a much simpler suboptimal equal power allocation scheme is proposed. Simulation results show that, although the performances of proposed suboptimal power allocation algorithms are slightly worse than the optimal power allocation algorithm, our algorithms are simple, fast and applicable in real-time implemention.2) In a OFDM cognitive DF relay network with multiple cognitive relay nodes, how to optimize cognitive nodes transmit power is investigated so as to maximize the sum rate under the constraints of interference temperature limit and sum transmit power of cognitive nodes. As this optimization problem categorized as nonlinear convex optimization problem with multiple constraints, solving this problem is very time-consuming and difficult for real-time implementation. For this reason, the optimization problem is decomposed, and a simpler near optimal power allocation algorithm is proposed. Consequently, an optimal relay selection scheme is also proposed. Theoretical analysis and simulation results show that, the proposed scheme is easy to implement and can satisfy the interference temperature limit and sum power constraints of cognitive nodes, making it more suitable for cognitive one-way relay network scenarios.(3) In the cognitive two-way AF relay networks scenario, the optimal and suboptimal power allocation algorithms are proposed, in which the rate optimization problem with fairness constraint is considered. Besides, an optimal relay selection scheme is also proposed. Meanwhile, an adaptive relaying protocol is adopted in OFDM cognitive two-way relay network, and the power allocation problems are effectively solved accordingly.1) In the cognitive two-way AF relay networks with multiple cognitive relay nodes, the SINR balancing algirthm is adopted to solve the optimization problem in which the rate optimization problem with fairness is considered under the constraints of sum transmit power and interference temperature limit and the closed form solution is attaninted. Two suboptimal power allocation algorithms are also proposed for engineering. The original optimization problem is divided into two steps in suboptimal algorithm:first, a simple power allocation is completed under the constraint of maximal sum of transmit power, the bisection method is then adopted to adjust the power according to interference temperature limit. Finally, the relay node that leads to best attainable rate is selected. Simulation results show that, compared with other schemes, the proposed scheme can achieve superior system performance and is more suitable for cognitive two-way relay network scenario.2) An adaptive relaying strategy is presented for OFDM cognitive two-way relay networks. Based on it, the power allocation problem is discussed, which is aimed at maximizing the sum rate of cognitive radio system and making sure that the interference introduced to the primary user is below the interference temperatured limit. By exploiting the Lagrange dual decomposition technique, the original problem can be decomposed into N subproblems for single carriers. According to different relaying protocol (AF or DF), each subproblem is analyzed in two cases. For AF, the subproblem is transformed into equivalent concave optimization problem in that it is very difficult to solve. For DF, the subproblem is divided into two cases according to different decoding order of cognitive relay nodes, the projected subgradient method is adopted to solve this problem. Subsequently, AF or DF relaying protocol will be selected adaptively depending on the obtained power allocation results. Simulation results show that, the adaptive relaying scheme can achieve better system performance at high SNR and low SNR levels in comparison with AF and DF relaying schemes.