Cognitive MIMO Radio: Incorporating Dynamic Spectrum Access in Multiuser MIMO Network

In this paper, we develop a general mathematical framework to incorporate dynamic spectrum access in a multiuser MIMO network. This framework is particularly helpful in computing the maximum achievable system capacity of a resulting multiple-band multiuser MIMO network. The mathematical formulation to maximize the system capacity is shown to be quite similar to that of a well studied single-band multiuser MIMO network. It is further shown that the capacity maximization problem is equivalent to finding the optimal eigenvalues of the input symbol covariance matrices of the users in each frequency band. Due to the dependence of the eigenvalues on the physical characteristics of the system, such as orientation of the antennas and the channel conditions, it is difficult to achieve their optimal values in general. Because of this difficulty in achieving the optimal capacity, we also consider the suboptimal MIMO techniques (specifically beamforming) and study their capacity performance in a multiple-band multiuser MIMO system.