Low complexity resource allocation for QF VMIMO receivers with a shared backhaul

In this paper, we consider the resource allocation problem of a quantize-and-forward virtual MIMO receive cooperation scheme, where N sources simultaneously communicate over N relays to a destination. Different to previous work, the second hop links are considered to be non-orthogonal (i.e. a shared backhaul) but with individual rate constraints. This requires a careful resource allocation, leading to an involved non-convex optimization problem. The goal of this paper is to provide and evaluate computationally simple and robust resource allocation algorithms. In this context, we devise two low complexity schemes which simply select a subset of relays and assign equal resources to them. In thorough numerical simulations we then demonstrate that for moderate backhaul rates (as to be expected in wireless virtual MIMO), it is beneficial to serve only a subset of the relays. Furthermore, while large gains are achievable by selecting the optimal relays and assigning equal resources to them, only minor additional gains can be achieved by further optimizing the resource allocation among them, e.g., with a gradient search. Hence, the proposed relay selection schemes lead to very good performance and are a reasonable alternative to costly optimization techniques if low complexity is required.