Clustering Based Centralized Provenance Forgery Detection System To Ensure Secured Wireless Sensor Network

Abstract

In the wireless sensor network environment, Provenance forgery is the most found threat, which affects the transmission of
packet rate. To rectify this issue in the earlier work, we brought-in the Improved Merkle-Hellman knapsack algorithm
(IMHKP), where the security is assured by fixing watch dog for each node, which is included in the routing process. This
watch dog would oversee the broadcasting of packet preceding and forwarding nodes. So, according to this monitored
behavior, the malicious node’s contribution will be recognized. Nevertheless this process is deficient in an exact
identification of malicious nodes, which estimates the malicious nodes only by assuming the previous and forwarding
nodes. Provenance would be built by every node, whereas they were acquiring the broadcasted packet, which will get
colluded at the time of un-recognized malicious nodes. More computation overhead will be created by controlling the
provenance by each node. In order to rectify these issues, in our proposed system we bring-in the Clustering Based
Centralized Provenance Forgery Detection System (CBC-PFDS). Here, by means of firefly algorithm, Clustering of sensor
nodes and optimal cluster head selection is performed, in order to optimize the resource handling and to eliminate the DoS
attacks. Here Cluster head will be chosen by considering the enhanced network lifetime, that observe the sensor node that
exist within its cluster periodically, so that attacker nodes are found efficiently, in keeping with their transmission activity.
Next to clustering, centralized authentication is performed with the help of pair wise key approach, to eliminate the
malicious node involvement that can clone the malicious attacks. Here, the security of provenance sharing is assured by
bringing-in the Elliptic Curve cryptography system, where the provenance would be enciphered. Then centralized server is
managed to build and control the provenance dictionary. Modification which is identified by the cluster head would
distinguish the receiving packets provenance with the centralized provenance dictionary according to the malicious
activities, which will be eliminated later. In the NS2 simulation environment, the entire implementation of the presented
research technique is performed and it is confirmed that the suggested system leads to give optimal result when compared
with the current research methodologies, with respect to enhanced security level.