Optimal energy consumption and Security enhancement in wireless Sensor network using machine Learning approach

Abstract

With the increasing adoptions and applications of Wireless Sensor Networks in various fields, it has to be admitted that Remote sensing with constrained nodes demands an extremely optimized procedures for resource utilization and security-rich network with minimum trade-off. One of the significant considerations for deploying WSN lies in its capability of resource consumption and its robustness to withstand various threats and attacks. Many of the literature focuses on dealing either to address the energy utilization or providing security aspects. newlineA Wireless Sensor Network is considered to be alive and active whenever any of the n feasibly estimated transaction could be succeeded in finite time with minimal or desired resource consumption. Henceforth, whether the problem is enhancing the life time or enforcing a consumption policy or routing policy to provide better security and reliability, all ends up in addressing a common root cause named Denial of Service. In this work, instead of fixing the DoS attack and its associated threats, an Optimal energy- based routing policy has been formulated which is then feed into an Adaptive Machine learning methodology to read the behavioral features for n normal transactions. Finally, with this data as multiple hidden layers, using deep learning methodology whereas every behavioral feature is monitored individually to keep malicious nodes out of the hop. This is the major contribution of the work, that addresses all the DoS attacks by successfully keeping the malicious node out of our current route.The proposed novel cover-set formation of nodes has been designed to customize multiple characteristics of nodes which influences cover-set. newline