Access point selection and auction-based scheduling in uplink MU-MIMO WLANs

Abstract

We study the uplink multi-user multiple-input multiple-output (UL MU-MIMO) channel in a wireless local area network (WLAN) with multiple access points (APs) and multiple stations (STAs). APs are allocated with different orthogonal frequency-division multiplexing (OFDM) subcarriers. Each AP has multiple antennas and applies zero-forcing successive interference cancellation (ZF-SIC) as in [1] for spatial multiplexing in its own MU-MIMO channel. STAs arrive and depart dynamically. The joint AP selection and STA scheduling problem is modeled as an auction game. STAs perform AP selection to maximize their utility, which is a function of the expected rate, the valuation and the price charged by APs. APs schedule STAs to maximize the social welfare and set price to ensure truthful revelation of STAs' valuation. The auction is conducted in a distributed fashion and requires subtle information exchange. Shown by the simulation, the proposed mechanism has superior performance in respect of social welfare, throughput and load balancing.