Dynamic Base Station Selections for QoS Provisioning over Distributed MIMO Links

Abstract

We propose the dynamic BS-selection schemes for distributed wireless MIMO links, which aim at minimizing the average base-station (BS) usage while satisfying the specified statistical delay quality-of-service (QoS) constraint characterized by the delay-bound violation probability. In particular, the distributed transmission system consists of a number of location-independent BS's, a central server, and one mobile station (MS). Based on the channel state information (CSI), the MS dynamically selects the subsets of BS's with variable subset cardinalities to construct the distributed MIMO link, and feeds back the selected BS's indices to the central server. The central server controls these selected BS's to cooperatively transmit data to the MS. Under the above framework, we develop the QoS-guaranteed dynamic BS-selection schemes for scenarios with and without feeding back the selected BS subsets' CSI, respectively. Simulation results show that our proposed dynamic schemes significantly decrease the average BS usage and interfering range as compared to baseline schemes which fix the cardinality of the selected BS subsets.