Link Quality Estimation for Adaptive Data Streaming in WSN

Abstract

In an industrial environment, reliable data transmission between wireless sensor nodes is a challenging factor, because the link quality is constantly degraded by industrial EM noise and inter-technology interference. Data loss as a result of massive transmission over such degraded link significantly affects the network lifetime. Therefore, this work proposes link quality based adaptive data streaming as a solution for effective deployment of low power Zigbee. Initially to determine the quality using RSSI and LQI indicator, an enhanced link quality estimation technique (ELQET) is designed with an intuitive combination of the Kalman filter and fuzzy logic. The quality score returned by fuzzy utilizing four efficient link metrics PRR, ASNR, ALQI, and SA is further smoothened with exponential weighted moving average filter for stability. Consequently the estimated quality is categorized into good/poor quality to stream data at high/low transmission rate respectively between the LPC2148s via CC2550 transceiver. Here, two contrasting RSSI and LQI data sets are furnished as input to ELQET and categorized into good link with quality of about 61 % and poor link with a quality of about 50 %. The straight forward low computation technique is WSN propitious and exhibits high performance with RMSE of 0.0133. The environment adaptive data streaming enhance the quality of transmission accompanied by reducing energy and data loss.