Research On Antenna Selection And Multi-user Diversity For Imperfect Channel State Information In Cognitive Radio

With the development of mobile Internet, radio spectrum resources is in short supply because of the increasement of the number of users. In the cognitive radio(CR)network, secondary users(SUs) are allowed to use the spectrum even when primary users(PUs) are active may improve spectral efficiency. Multi-user diversity(MUD)technology can effectively improve capacity in multiple-input multiple-output(MIMO)systems. However, multiple antennas must be configured in MIMO, which significantly increasing the hardware cost and difficulty in this process. Antenna selection(AS) technique, a method select the transmitter and receiver antennas which performs better in each time slot to communicate, not only takes advantage of the MIMO, but also saves costs.This paper studies the AS and MUD based on robust downlink beamforming in multi-user MISO CR networks with imperfect channel state information(CSI).AS is a non-convex problem about integer programming which is difficult to sovle directly. In the AS, we discuss three algorithms: optimal, iterative and non-iterative norm-based. The optimal AS algorithm is an exhaustive algorithm which has the best performance but the most complex. Hence, the latter algorithm will be key point. In addition, the non-iterative norm-based AS algorithm is the least complexity algorithm, but its performance is not so good. Iterative AS algorithm which has two optimization variables could be solved by convex approximation. The simulation result shows that the performance of iterative AS algorithm which has low-complexity is close to that of optimal AS algorithm. So, the former should be the best compromise term.In the MUD, we take simultaneously scheduling multi-user into consideration,and propose two scheduling algorithms: optimal scheduling and fast scheduling.Optimal scheduling, a exhaustive algorithm, which has difficulty in achieving real-time scheduling because of the complexity when the number of users that access the system is numerous. The fast scheduling is an algorithm based on maximum norm strategy which can effectively schedule while obtains satisfactory performance. The simulation result shows that fast diversity scheduling algorithm not only reduce the amount of calculations,but also get good performance.