Analysis of additive noise characteristics in indoor wireless sensor networks

Indoor wireless sensor networks (WSN) with high scalability and availability are suitable for multiple applications. They can be found in a variety of modifications sensitive to both network and communication design. Proper simulation models are required in order to produce realistic transmission conditions for WSN design optimization. The presented paper documents research results of time series analysis performed on the physical additive noise trace of real indoor WSN under different signal-to-noise ratio (SNR). The primary focus lies in identifying noise characteristics exploitable in error models of physical channels during the design phase of a custom error control mechanisms or cross-layer optimization. The results lead to a identification of a single and easily obtainable characteristic of the indoor WSN additive noise for the physical layer modeling. Furthermore, applicability of autoregressive (AR), moving-average (MA) and autoregressive moving-average (ARMA) models to this characteristic is provided along with suggestions for future improvements to the analysis.