Skip to Main Content
In this paper, we investigate minimum mean squared error (MMSE) based amplify-and-forward cooperative multiple antenna relaying systems where a non-negligible direct link exists between the source and the destination. First, we provide a new design strategy for optimizing the relay amplifying matrix. Instead of conventional optimal design approaches resorting to an iterative gradient method, we propose a near optimal closed-form solution which provides an insight. As relay systems with a direct link incur a non-convex problem in general, we exploit the decomposable property of the error covariance matrix and a relaxation technique imposing a structural constraint on the problem. Next, we study the error performance limit of the proposed scheme using diversity-multiplexing tradeoff analysis, which leads to several interesting observations on MMSE-based cooperative relaying systems. Finally, through numerical simulations, we confirm that the proposed solution shows the performance very close to the optimum with much reduced complexity and the analysis closely matches with simulation results.