Wireless Channel Prediction in Partially Observed Environments
This work addresses the challenge of RF propagation prediction for dynamic robotic navigation, but it is incremental as it builds on existing ray tracing and ML techniques.
The paper tackles the problem of predicting wireless channels in partially observed environments by introducing a method that combines partial ray tracing with machine learning to estimate channel characteristics and uncertainty. The method is demonstrated in simulated robotic navigation, showing it can interpolate between statistical and deterministic models based on observation completeness.
Site-specific radio frequency (RF) propagation prediction increasingly relies on models built from visual data such as cameras and LIDAR sensors. When operating in dynamic settings, the environment may only be partially observed. This paper introduces a method to extract statistical channel models, given partial observations of the surrounding environment. We propose a simple heuristic algorithm that performs ray tracing on the partial environment and then uses machine-learning trained predictors to estimate the channel and its uncertainty from features extracted from the partial ray tracing results. It is shown that the proposed method can interpolate between fully statistical models when no partial information is available and fully deterministic models when the environment is completely observed. The method can also capture the degree of uncertainty of the propagation predictions depending on the amount of region that has been explored. The methodology is demonstrated in a robotic navigation application simulated on a set of indoor maps with detailed models constructed using state-of-the-art navigation, simultaneous localization and mapping (SLAM), and computer vision methods.