Differential data processing technique to improve the performance of wireless sensor networks

Abstract

A wireless sensor network is a network composed of various types of sensors for the detection of magnetic, thermal, infrared, and acoustic fields, in addition to earthquakes and radio detecting and ranging (radar), among others. The size of the data collected from the various sensors is significant, and it is utilized in various applications such as artificial intelligence, data prediction, and analyses. However, the hardware of a wireless sensor node has limited energy and consumes a significant amount of energy for data transmission. Several studies have been conducted on multi-hop communication, clustering, and the compression/merging of data to increase energy efficiency in data transmission. In this study, the differential data processing (DDP) method is employed to reduce the size of the transmission data and improve the performance of wireless sensor networks. We propose node identification (ID)-based DDP and cluster header (CH)-based DDP. The node ID-based DDP transmits the initial collected aggregate data to the CH at the start of the collection, compares the previous collected data with the currently collected data, and then transfers the difference value to the CH. The CH is transmitted to the base station based on the smallest value of the collected data. The CH-based DDP collects data from the CH, generates reference data for the difference, and transmits the reference data at the time of cluster broadcasting. The member node performs differential processing on the collected data using the reference data transmitted to the CH. The performances of low-energy adaptive clustering hierarchy and DDP were compared. The simulation results revealed that the performance of the wireless sensor networks was improved by efficiently using the energy of the sensor nodes and by decreasing energy consumption in data transmission, given the reduction in the data size.