Optimization methodology for designing 2-CPAs exploiting pattern diversity in clustered MIMO channels

Arrays designed for multiple-input multiple-output (MIMO) communication systems must provide good performance in terms of information theoretic and microwave theoretic performance measures for a large number of channel scenarios. Consequently, the general MIMO array design problem is challenging. This paper proposes a novel optimization methodology for designing circular patch arrays (CPAs), under a clustered MIMO channel model assumption. Based on the ergodic capacity, it is found that the single-cluster scenario can approximate CPA performance in clustered channels, reducing the number of channel scenarios that need to be considered. Using a new definition of channel spatial correlation that allows for pattern diversity, metrics are derived for evaluating average and outage capacity performance of CPAs, which can be expressed as a function of only the antenna parameters. With this expression, an optimization is formulated and solved numerically for optimizing the parameters of a 2-CPA on a capacity based objective function and microwave theory performance constraints. An example design is given corresponding to a 2-CPA in the 2.4 GHz unlicensed frequency band. Index Terms--MIMO systems, antenna arrays, pattern diversity, microstrip antennas, spatially correlated channels.