Limiting Queueing Models for Scheduling in Multi-user MIMO Wireless Systems

M. Airy, S. Shakkottai, and R.W. Heath, Jr (USA)

Keywords

MIMO scheduling, Timescale separation

Abstract

Opportunistic scheduling based on full channel state information at the transmitter can provide sig nificant performance gains for wireless networks by ex ploiting the independence of fading statistics across the user population. While opportunistic schedulers typically pick the single "best" user for transmis sion, MIMO (multiple-input multiple-output) systems can support transmissions to multiple users simulta neously. This paper compares (i) greedy multi-user sum rate maximizing scheduling, where multiple users are served simultaneously; and (ii) single-user schedul ing, where all antenna resources are allocated to a sin gle user depending on both the channel state and job length information. Based on two limiting arguments, where we consider fast and slow channel variations, we conclude that in most regimes greedy sum rate maxi mizing scheduling strategy performs better than single user strategies that use a combination of channel state and job length information.

Important Links:



Go Back