Hybrid Approach of KNN and Euclidean Distance to Tackle Sybil Attack in the Network

ABSRACT The wireless sensor network uses batteries which decay as data is being transmitted from source and destination. This decay in batteries requires to be minimised. Reasons for decay in batteries could be congestion and attacks. The attacks which are common in WSN is Sybil attack which is multiple identity attack. In such a situation, attacker node copies the identities of other nodes and data which is transmitted delivered to the wrong node causing threat to secure data. In order to solve the problem KNN mechanism is used in the proposed system. KNN is used in order to form clusters of the minimum distance nodes. These clusters then can be examined for similarity in terms of identities. In case similarity in terms of identities is found then Sybil attack is detected. The result is presented in terms of classification accuracy and mean square error. Classification accuracy is obtained by subtracting the actual value from the approximate value. The error rate is obtained by subtracting the classification accuracy from 100. The proposed approach uses Euclidean distance to determine the neighboring nodes. The simulation of the proposed system is conducted in MATLAB 2017. The mechanism employed detects the Sybil attack with more precision. The result is improved by the margin of 10 % proving worth of the study.