Mitigating the Effects of Irregular Reflection Beam Patterns Through Data Association in RIS-Aided mmWave Positioning

Despite promising potential for boosting or even simply enabling the localization capabilities of future wireless communication networks, low-complexity reflective Reconfigurable Intelligent Surfaces (RISs) have recently been shown to suffer from practical limitations, such as irregular beam patterns and grating lobes caused by element-wise phase quantization or hardware impairments. The latter phenomena can introduce geometric ambiguities in the RIS-aided positioning problem, when estimating the RIS Angle of Departure (AoD) through beam sweeping based on downlink received power. In this context, our paper introduces an algorithmic framework that combines user position estimation and data association between radio observations and underlying channel model parameters, while still requiring the same basic RIS hardware and UE channel estimation capabilities. Validations based on field measurements centered around 28GHz are provided in a reference indoor environment for different bandwidths, showing the relevance of the proposed solution in terms of both position accuracy and geometric disambiguation.