The energy-efficient design of multiple antenna communication systems

Minimizing the energy consumption of a portable wireless system is a very critical issue due to the finite capacity of a battery. In this dissertation, the energy-efficient design of multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) based systems is addressed. The energy-aware link adaptation and energy-optimized power amplifier (PA) clipping are presented as system level energy optimization techniques. In addition, a low complexity MIMO detector design is provided for algorithm/architecture level energy optimization. The objective of energy-aware link adaptation is to choose the optimal mode that maximizes energy efficiency subject to a given quality of service constraint (QoS). The link adaptation problem is formulated with the set of optimization variables including: number of spatial streams, number of transmit/receive antennas, selection of space time coding scheme, constellation size, bandwidth, transmit power and choice of the MIMO detection algorithm. The resulting solution allows us to systematically search the space of system parameters to deliver on the QoS with minimal energy consumption. Application of the results shows that the proposed strategy can provide an order of magnitude improvement in energy efficiency relative to a static strategy. For a MIMO-OFDM system that is being pushed into PA saturation, this dissertation investigates power optimized PA clipping. The goal is to analyze the tradeoff resulting from adaptive PA clipping and identify the optimum clipping that delivers the desired bit error rate (BER) with minimum energy consumption in the PA. A complete theoretical framework for the optimal PA clipping is provided based on an analytical expression of BER subject to clipping. Application of this technique shows that as much as 78% savings in power consumption can be achieved by proper choice of clipping level. Finally, this dissertation presents a highly optimized minimum mean square error (MMSE) MIMO detector implementation. Design of an energy-efficient MIMO detector requires co-optimization of the algorithm with the underlying hardware architecture. Special attention must be paid to application requirements such as throughput and latency while minimizing the algorithm complexity. The proposed MIMO detector has resulted in a field programmable gate array based implementation, which can deliver over 420Mbps sustained throughput with a small 2.77 μs latency. The designed 4x4 MMSE MIMO detector is capable of complying with the IEEE 802.11n standard.