PROVENANCE BASED MALICIOUS NODE DETECTION IN WIRELESS SENSOR NETWORK USING BLOOM FILTER

Abstract

Wireless Sensor Networks (WSN) are essential in modern day to gain information about the environmental studies that are useful in decision making. Data is transferred through wireless medium, so the integrity of the data has to be maintained. We have created a model that helps to securely transmit the data from source node to base station. We have implied provenance based bloom filter and cryptographic algorithm to securely transmit data to the base station. We have also created a model to detect packet dropping malicious node when data is in transit. In our model, we have used AODV protocol as a routing algorithm and AES-128 as a cryptographic algorithm to encrypt the data during transmission. Bloom filter requires a hashing algorithm so in our case SHA-224 cryptographic hashing algorithm has been deployed. We used a light weight Bloom filter model to transmit the data and to detect any packet dropping node which act as an intermediate nodes. We used provenance information to help detect any malicious packet dropping nodes. We relied on provenance encoding and decoding methods to our model. We analyzed our model using different parameters like Verification failure rate (VFR), True false positive (TFP), throughput, end to end delay, etc on different number of nodes, packet size and bloom filter size.