中文摘要：

研究了认知MIMO系统频谱检测与吞吐量折中问题，以最大化SU（Secondary User）吞吐量为目标，求解出最佳的频谱检测时间和最优功率分配方案。仿真结果表明，存在最佳检测时间使su的吞吐量达到最大，并且通过注水法功率分配，能进一步提高SI的吞吐量。此外，通过仿真比较了能量检测和基于特征值的频谱检测算法对SU吞吐量的影响，结果表明，采用两种算法SU的吞吐量几乎相等，由于基于特征值的频谱检测算法更具有健壮性，因而更适合应用于认知MIMO系统。

英文摘要：

Spectrum sensing-throughput tradeoff is investigated for a cognitive MIMO system. The objective is to find the optimal spectrum sensing time and optimal power allocation to maximize the throughput of secondary user. Simulation results show that there is indeed a unique optimal sensing time to make thethroughput achieve the maximum. Further, the throughput of SU can be improved by using optimal power allocation. The throughput of SU with different spectrum sensing algorithms including energy detection and eigenvalue-based detection is compared by simulations. The results show that the eigenvalue-based detection is very appropriate for cognitive MIMO system, which is not only more robust than energy detec- tion, but also satisfies the requirement of throughput.