Linear transceiver design for multiuser spectrum sharing MIMO networks

We design a linear transceiver in the downlink for a spectrum sharing multiuser multiple-input-multiple-output (MIMO) networks. This linear transceiver ensures that the required quality of service (QoS) is achieved at each user while satisfying the interference constraints on the primary users. Here, certain predefined throughput is considered as a QoS requirement for each user. The original QoS based transceiver design in the downlink is solved by exploiting the stream-wise mean-square error (MSE) duality and based on alternating optimization framework between the uplink and the downlink. From the MSE duality, we propose an iterative algorithm to obtain transceiver filters. In each iteration, the downlink receiver filters are designed based on minimum mean square errors (MMSEs). The virtual uplink receiver filters are designed by solving a quadratically constrained quadratic programming (QCQP), which incorporates the interference constraints on the primary users. The power allocation problem in the virtual uplink is solved using a geometric programming (GP), where the required total transmit power is minimized to satisfy the required rate constraints. The performance of the proposed algorithm is validated through the simulation results.