Exploiting Sensed Radio Strength and Precipitation for Improved Distance Estimation

The received signal strengths (RSSs) of wireless radios are susceptible to environmental dynamics, such as weather conditions, which can severely degrade the performance of RSS-based distance estimators. This paper proposes an enhanced RSS-based distance estimator that accounts for precipitation. The rain attenuation effect is incorporated into the path loss model to derive an analytical expression for maximum-likelihood-based distance estimation by using an iterative form of Newton's method. In addition, the Cramér-Rao lower bound for the proposed distance estimator is proposed. This study performed experiments on the mobile networks at a campus, sensing the actual RSS using smartphones and deploying local weather stations to record transmission conditions. This is a pioneer work to overcome the rain attenuation problems and real measurements verify the effectiveness of an RSS-based distance estimator. Simulation and experimental results indicated that the proposed algorithm provides robust distance estimation during the periods of precipitation. The parameters of radio propagation and rain rate models can be extracted offline. With these parameters and the real-time measured RSSs and rain rate, our proposed distance estimation method achieves an 85%-90% error reduction rate over the conventional approach.