Malicious Nodes Detection and Avoidance Using Trust-based Routing in Critical Data Handling Wireless Sensor Network Applications

With the ever-evolving wireless technological domain, large number of applications are developed which utilizes the advantages of wireless technology. Healthcare, defense, intrusion detection, agriculture, logistics etc. are examples of domains which witnessed a complete paradigm shift due to the incorporation of wireless technology. Wireless Sensor Networks (WSN) have provided the most suitable platform for developing applications in various domains utilizing the advantageous of wireless transmission of data between devices. Along with the numerous advantages offered by the wireless communication technology, applications developed over WSN faces the most critical challenge of managing the battery for performing the various operations including computations and communication. Due to this constraint, generally applications deployed over WSNs are focused on optimizing energy consumption to improve the lifetime of the WSN nodes. But when WSNs are used to deploy applications from defense or healthcare sectors, an additional requirement has been taken the focus which is data security. Most of the applications in healthcare or defense sectors handle confidential data and communication of such data must be secured. The traditional security schemes deployed for secured data communication are computationally complex and consumes larger amount of energy this making it less feasible for WSN based applications. This opens the possibility of attacks on WSN based applications. Cardinal impairment is regarded as attacks emanating from compromised nodes and changing data. The paper analyses several attacks and proposes a new security approach to prevent data change due to compromised nodes. The suggested technique is a non-cryptographic scheme which is computationally less complex. In this scheme, nodes compute trust of the neighbors and exclude hostile nodes from the probable list of forwarding nodes based on low trust value. Trust computation is carried out considering various parameters like packet drop rate, packet rejection ratio and remaining energy level of the node. The proposed system is simulated using Network Simulators (NS2) and performance of the proposed system is compared against traditional schemes.