Low-Complexity Adaptive Spatial Processing of ESPAR Antenna Systems

Accessing the spatial domain of wireless environments through antenna arrays has become key to improve the use of energy and, in light of this, there is an increasing need for low cost multi-antenna architectures. As one alternative, the Electronically Steerable Parasitic Array Radiator (ESPAR) was proposed as a coupling-based inexpensive option; allowing to significantly reduce the amount of required radio frequency (RF) front-ends. As a caveat, due to its inherent non-linear behavior, the required computational complexity can be prohibitive. Additionally, the unavailability of precise mutual coupling and channel state information (CSI) becomes a further issue difficult to avoid in practical setups. In this regard, the current work has two main contributions: a) it deals with the computational complexity by proposing the linearization of ESPAR's system model through the truncated Taylor expansion of the admittance matrix; facilitating the spatial processing algorithm of interest. Also, the authors propose b) to face the mutual coupling and CSI unavailability issue via the joint estimation of the channel-ESPAR parameters as directly observed through the single RF front-end. Relying on the optimization of the linearized system model, the pilot-based algorithm to be introduced allows ESPAR to track the configuration that synthesizes coherent combination at a significantly low complexity. Particularly, as shown through Monte-Carlo simulation, the latter allows to obtain 4 dB of array gain with a single RF front-end via a 5-element ESPAR.