Over the past several years, there has been a great deal of research to improve performance of wireless communications in fading environments by exploiting transmitter and/or receiver diversity and spatial multiplexing. The pioneering work by Telatar, Foschini and Gans showed that multiple antenna radios (MAR) in a wireless communication system can greatly improve the performance. The major roadblock to MAR being a financially viable option for wireless communications is the cost of the analog portions of the radio. Point--to--point MAR systems employing complex signaling and signal processing algorithms have been designed and built and have achieved performance very close to the outage capacity. The remaining gains that are achievable in MAR will only be exploited by considering the cost and complexity of analog circuitry in MAR signalling and signal processing, by considering more complex models for wireless propagation, and by putting MAR in networking context. The focus of this effort is to push MAR technology by exploiting the structure in these more complex models and networking scenarios.
The objectives of the effort can be succinctly stated as: 1) Communication theory is often interested in normalizing performance for a fixed average power but in wireless implementations the peak power is a more critical driver than the average power. The research is examining optimization of performance and throughput with a peak power constraint in multiple antenna radio. 2) Feedback communication is likely going to enable a significant improvement in throughput and demodulator complexity in a networked scenario especially in environments with low mobility (e.g., wireless computer networks). This research is approaching feedback communication with a classic scientific method of iterating between experimentation on real channels and more realistic theoretical problem formulations based on the results derived in experimentation. 3) The cost driver in wireless communication currently is the analog circuitry. This research is examining techniques that would minimize the complexity of the analog systems in a multiple antenna radio. The objectives are to being met by the following research tasks a) Improved channel parameterization for long range prediction for multiple antenna radio, b) Design and field testing of a variety of MAR automatic repeat request algorithms, c) Experimentation with the proposed feedback algorithms for MAR, d) Feedback algorithms for MAR with constant envelop modulations, e) Feedback algorithms for MAR with limited analog capabilities,
