Capacity of distributed MIMO with finite size

Abstract

Previous work has shown that distributed cooperative MIMO systems (e.g., the hierarchical MIMO cooperation scheme) can provide large gains in capacity if the size of the MIMO systems is a function of the total number of nodes in the network (n). However, no results have been reported on the scaling laws of distributed cooperative MIMO systems when the number of transmit and receive antennas are of finite size, which is the case in real networks. This paper uses the extended network model to demonstrate that, if the size of distributed MIMO systems is restricted to a finite size, then there is at most a constant gain compared to point-to-point communications. Opportunistic interference management (OIM) is introduced as an alternative to distributed MIMO, and it is shown that it provides higher order gains than distributed cooperative MIMO systems under the same assumption of having a finite number of transmit and receive antennas. More specifically, OIM achieves a throughput capacity of C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> (n) = Θ (D log log 2/θ log n/√n log n) in fading channels when the transmission range T(n) is Ω (√2/θ n), where θ is a constant parameter close to zero. This constitutes an order gain of Θ(log log 2/Θ log n) compared to point-to-point communications and distributed MIMO systems! Given that it is much easier to implement OIM schemes than distributed cooperative MIMO systems, these results indicate that OIM is a far better choice to making wireless ad hoc networks scale than distributed cooperative MIMO systems.