Estimation of Received Signal Strength Distribution for Smart Meters with Biased Measurement Data Set

TL;DR

This paper introduces a novel modeling approach combining Rician fading with a bias function to accurately estimate the distribution of received signal strength in smart meters, accounting for measurement bias due to receiver sensitivity thresholds.

Contribution

The paper presents a new method that models the biased signal strength measurements of smart meters using a combined Rician and bias function approach, improving distribution estimation accuracy.

Findings

The proposed model outperforms naive Rician fitting in approximating signal strength distribution.

Experimental validation with two datasets demonstrates the effectiveness of the combined model.

The approach accounts for measurement bias caused by receiver sensitivity thresholds.

Abstract

This letter presents an experimental study and a novel modelling approach of the wireless channel of smart utility meters placed in basements or sculleries. The experimental data consist of signal strength measurements of consumption report packets. Since such packets are only registered if they can be decoded by the receiver, the part of the signal strength distribution that falls below the receiver sensitivity threshold is not observable. We combine a Rician fading model with a bias function that captures the cut-off in the observed signal strength measurements. Two sets of experimental data are analysed. It is shown that the proposed method offers an approximation of the distribution of the signal strength measurements that is better than a na\"ive Rician fitting.