TBDRS: Threshold Based Data Reduction System for Data Transmission and Computation Reduction in WSNs

Abstract

Wireless sensor networks, consisting of sensor nodes, monitoring the real-world phenomena generate a huge amount of data. However, the transmission of these measured values consumes most of the battery power of energy constrained sensor nodes. In this paper, threshold based data reduction system (TBDRS) is proposed in order to reduce the transmissions from the sensor node by exploiting the correlation between the measured values. Unlike other systems, TBDRS uses a control mechanism for threshold based switching on the basis of estimation error. This switching allows the system to achieve better trade-off between convergence speed and misadjustment, thereby improving the lifetime of the wireless sensor network by reducing the number of transmissions and computations. Due to the settling time, misadjustment trade-off capability, TBDRS can also be used in disaster management applications where the sensor data variations are more. The performance analysis of the system is evaluated in terms of mean square error. Moreover, the performance of TBDRS is evaluated for different parameters including length, step size and scaling factor. The simulation results validate that the proposed system achieves 70.56% and 83.01% better performance than the state-of-the-art system for an accuracy of 0.1 in terms of data reduction and number of computations, respectively.