Improving phaseless DoA estimation in multipath-impaired scenarios by exploiting dual-band operations

This paper presents an approach improving phaseless Direction-of-Arrival (DoA) estimation accuracy for indoor environments. The positioning system's anchor is equipped with a dual-band transceiver and a switched beam antenna, operating in circular polarization at both 2.45GHz and 5.7 GHz, hence compatible with legacy IEEE 802.11abg connectivity. The estimation of DoA of a nomadic node equipped with a monopole antenna, is the result of a likelihood criteria driven by the expected signal partition. Considering the weak correlation of the multi-path at incommensurate frequencies, the data fusion collected at both 2.45 GHz and 5.7 GHz for the likelihood algorithm is expected to mitigate noise and in turn it allows for better estimation performance. The experimental validations demonstrate the performance of the proposed approach positioning the node at the distance of 3.5m inside a 5×4×3 m3 setup, being the latter perturbed with a large and invasive conductive ground aimed at the multi-path generation. Despite a very critical localization result at each single frequency, with a worst case of 90% of the observed domain affected by an error up to 50°, the data fusion approach boosts the performance getting the error below 12° in the same conditions. Similar improvements are observed for both the horizontal and vertical polarization of the nomadic node and for multiple configuration of the reflecting plane, demonstrating the effectiveness of the proposed approach.