A Density-Based Decision-Making Data Fusion Method for Multiapplication Wireless Sensor Networks

Wireless sensor network (WSN) design has moved towards scenarios with multiple applications, allowing several concurrent applications to share sensing and communication resources. The new multiapplication paradigm made earlier WSN limitations such as limited processing power and energy supply even more challenging. Several studies have tackled the challenge of extending the operational lifetime of WSNs and a promising approach is applying Multisensor data fusion (MDF). MDF denotes a set of methods to enable synergistic combination of sensing data from multiple sources, augmenting the quality of the inferences performed from such data. As a secondary benefit, MDF decreases the number of messages transmitted in the network, thus the energy spent in communication, prolonging the WSN lifetime. MDF techniques can be applied at different levels regarding the complexity and semantics of input and output data used in the process. To date, most of the approaches for WSNs dealt with low-level data. We claim that a decision-level MDF for multiapplication WSN (MWSN) has the potential to improve the accuracy of the fusion process and to reduce resource consumption. Therefore, our contribution in this paper is to propose a decision-level multiapplication MDF technique based on data density that extends the algorithm presented in a previous work, augmenting it with the ability to make decisions and later analyze the decisions taken to avoid conflicts. The goal is to achieve high data accuracy and reduce energy consumption in the MWSN. The performed experiments shown that we achieved the stated goal.