Random Black Hole Attack Modelling and Mitigation Using Trust- Confidence Aware OLSR in MANETs for Private Data Communications

Abstract

Background and Objective: Random Black Hole (BH) attack significantly degrades MANET’s performance. For strategic applications, the performance parameters like Packet Delivery Ratio, Routing Overheads, etc. are important. The objectives are: (a) To model random BH attack, (b) To propose a routing strategy for the protocol to mitigate random BH attack, (c) To evaluate and compare the network performance of modified protocol with the standard protocol. Methods: The random BH attack is modelled probabilistically. The analysis is carried out by varying Black Hole Attack (BHA) time as Early, Median, Late occurrences and mix of these three categories. The blocking performance is also analysed by varying the percentages of malicious presence in the network. Normal Optimized Link State Routing (OLSR) protocol is used to simulate the MANET performance using a typical medium size network. The protocol has then been modified using Trust- Confidence aware routing strategy, named as TCAOLSR, with a view to combat the degradations due to the random BH attack. Results: The random behavior of Black Hole attack is analyzed with all the possible random parameters, like deployment of mobile nodes, number of malicious nodes and timing instances at which these nodes change their state. From the results of individual type- Early, Median and Late, it is observed that the TCAOLSR protocol gives stable performance for Packet Delivery Ratio (PDR) and Routing Overheads (RO), whereas for OLSR protocol PDR gradually reduces and RO increases. For individual and mix type, Average Energy Consumption (AEC) per node increases marginally for TCAOLSR protocol. For the mix type, PDR for TCAOLSR is 40-60% better whereas RO for TCAOLSR is very less compared to OLSR protocol. The efficacy of the TCAOLSR protocol remains stable for different categories of BH attack with various percentages of malicious nodes compared to OLSR with the same environment. Conclusion: Simulations reveal that the modified protocol TCAOLSR, effectively mitigates the network degradation for Packet Delivery Ratio and Routing Overheads considerably, at the cost of a slight increase in Average Energy Consumption per node of the network. Efficacy of the OLSR and TCAOLSR protocols has also been defined and compared to prove robustness of the TCAOLSR protocol.