Source Localization from Received Signal Strength Under Log-Normal Shadowing: Bias and Variance

This paper considers statistical issues of source localization from received signal strength (RSS) measurements affected by log-normal shadowing with focus on bias and variance. We characterize the class of all unbiased estimates of the source position and show that their error variances grow exponentially with the noise power. Finally, we propose an estimate the bias and error variance of which are both bounded in the noise power.