A Novel IDS to Detect Multiple DoS Attacks with Network Lifetime Estimation Based on Learning-Based Energy Prediction Algorithm for Hierarchical WSN

Lifetime of sensor network plays a crucial part in designing any WSN-based application. On the other hand, securing them against malicious activity is also important. Security and lifetime should work hand in hand to protect and conserve the network. For example, imparting any security solution should not decrease WSN’s lifetime. Thus, a novel energy prediction algorithm is developed and deployed in WSN structure that can actively look and report for any adversaries present within the network and can also report the network’s lifetime stating the remaining amount of time the network can work perfectly without losing its energy. The key idea is that by retrieving the residual energy of the nodes at an instance of time, the proposed algorithm can predict its energy consumption at various other time instances; in such a way how long the nodes can stay alive is predicted. Moreover, we observe that different types of DoS attack consume abnormal amount of energy; by predicting its energy consumption at various instances and comparing with its actual energy consumption, attacks can be identified and classified. This novel mechanism can look for and detect 5 types of DoS attacks at a time. By formulating the network’s lifetime, the design of WSN can be optimized according to the application requirements. Detailed performance analysis is done using NS-2 Mannasim framework. Simulation studies are done under various scenarios to prove its efficiency and accuracy.