In this project we examine the ultimate network throughput that can be achieved with the use of multi-element antennas at each node, and allowing multiple MIMO links to transmit one or more streams simultaneously. Network topologies used in this study were focused on local area wireless networking technologies such as IEEE 802.11 and Bluetooth.

Previous research findings include distributed link adaptation algorithms that yield suboptimal transmit and receive weights for interfering closed-loop MIMO links, and provide distributed resource allocation capabilities within a network [5-7]. A stream control algorithm, also distributed, assures that the receivers are not overloaded with excessive desired and interference streams [8]. Using transmit power, desired capacities, or the number of parallel streams for each link as control parameters, these algorithms produce higher network throughputs through spatial multiplexing than possible with TDMA. Moreover, these parameters allow control of the capacity of each link according to different data rate or quality of service (QoS) requirements.

Interfering MIMO links

Simulation results showed that the use of channel state information (CSI) at the transmitter makes a significant difference in the link capacities in the presence of interference, which makes closed-loop operation (i.e. with channel feedback to the transmitter) feasible for static and low-mobility networks [8-10].

It is also shown that by using simulated and measured indoor MIMO channels, the proposed network model with interfering MIMO links provides improvement over the TDMA throughput even when the desired and interfering streams are spatially correlated (i.e. coming from the same direction). The improvement is more significant with less correlation. For a two-link network, the improvement of the proposed model with interfering links over TDMA is in the range from 15 to 75% [9].