Scaling Laws for Ergodic Spectral Efficiency in MIMO Poisson Networks

In this paper, we examine the benefits of multiple antenna communication in random wireless networks, the topology of which is modeled by stochastic geometry. The setting is the Poisson bipolar model introduced in [1], which is a natural model for ad-hoc and device-to-device networks. The primary finding is that, with the knowledge of channel state information between a receiver and its associated transmitter, by zero-forcing successive interference cancellation, and for appropriate antenna configurations, the ergodic spectral efficiency can be made to scale linearly with both: 1) the minimum of the number of transmit and receive antennas and 2) the density of nodes. This scaling law is achieved by using the multiple transmit antennas to send multiple data streams (e.g., through an openloop transmission method) and by exploiting the receive antennas to cancel interference. Furthermore, when a receiver is able to learn channel state information from a certain number of near interferers, higher scaling gains can be achieved when a successive interference cancellation method is used. Both results require rich scattering environments. A major implication of the derived scaling laws is that, under this scattering assumption, spatial multiplexing transmission methods are essential for obtaining better and eventually optimal scaling laws in random wireless networks with multiple antennas.