An adaptive algorithm for multipath-assisted simultaneous localization and mapping using belief propagation

In multipath-assisted simultaneous localization and mapping (SLAM), position-related information (delays and angles) in multipath components of radio signals is used to simultaneously localize user equipment (UE) and map the environment. Multipath-assisted SLAM often involves unknown parameters that are potentially time varying, such as measurement noise. Knowledge of the measurement noise is of critical importance to multipath-assisted SLAM, and uncertainty in such knowledge will seriously affect estimation accuracy. We address this challenge by improving the belief propagation (BP)-based SLAM algorithm and proposing an adaptive multipath-assisted SLAM algorithm in a Bayesian tracking framework, which enables accommodation of a model mismatch of the measurement noise online. Specifically, we describe the evolution of the measurement noise standard deviation via a Markov chain and integrate it into the factor graph representing the Bayesian model of the multipath-assisted SLAM. Then, the BP message passing algorithm is leveraged to calculate the marginal posterior distributions of the UE, environmental features and the measurement noise standard deviation to achieve SLAM and the adaptive adjustment of the measurement noise. Finally, the experimental results verify the robustness of the proposed adaptive multipath-assisted SLAM algorithm against the uncertainty of the measurement noise.