Efficient Link Scheduling for MIMO Ad Hoc Networks in Time-Varying Channels

In an ad hoc network, efficiently scheduling the optimal multi-user multiple-input multiple-output links is challenging. Specifically, optimal scheduling requires full understanding of the subspace structure of the channel matrices, but exploiting this detailed information during scheduling requires significant computational and algorithmic complexity. This paper demonstrates efficient yet effective scheduling based on link signal-to-noise ratio coupled with the average ratio of the maximum to minimum channel singular values and the channel Doppler frequency. Computational results show that this scheduling technique is able to achieve average network sum rates that approach those of the optimal link topology.