Downlink Adaptive Array Algorithms for Cellular Mobile Communications

The capacity of wireless downlink communication to mobile receivers in a dense urban environment is limited primarily by co-channel interference. Downlink adaptive arrays can be used to mitigate this limitation, thereby offering the potential for increased capacity. The basic objective is to maximize the power transmitted to desired in-cell mobiles in the reference cell while minimizing power to co-channel mobiles in neighboring cells. 

This is accomplished by using uplink measurements to estimate downlink signal and interference covariance matrices. In this memo, a local objective function is proposed to approximate global optimality in a cooperative network and is minimized by the solution to a generalized eigenvalue problem. 

Several algorithms are proposed to adaptively estimate the optimal solution and are evaluated using a simplified signal model that allows efficient deterministic performance calculations. Of the several algorithms considered, a simple iterative matrix power method is shown to realize near-optimal performance for this application.