Abstract:
The MANET would soon replace existing wireless technology, due to its easy to deploy and bare
minimum infrastructure requirement and dynamic topology. Emergence of faster high-speed hand held mobile device, mail delivery drones, drone camera footage of live cricket match etc. have
reserved MANET in the reach of common man. There for it is imperative to address ubiquitous
issue existing in MANET. These are safeguard from attacks by intruder nodes, consideration of
tradeoffs between speed and efficient communication and selection of reliable and secure network
paradigm. Analysis of the mobile ad hoc networks system from security stand point is crucial in
order to construct a robust and counteractive system.
MANETs are surrounded by various attacks, each with different behavior and aftermaths. One of
the serious attacks that affect the normal work of MANET is DoS attack. One of DoS is jellyfish
attack, which is quite hard because of its foraging behavior and exploit the behavior of closed loop
k2protocol and disturb the communication process without disobeying any protocol rules, thus the
detection process becomes challenging. Consequently, traffic is disrupted leading to degradation
in network throughput which degrades over all network performance. The jellyfish attack is
regarded as one of the most difficult attack to detect and degrade the overall network performance.
In order to mitigate jellyfish attack in MANET this paper propose a novel technique called holding
period foundation and accurately detecting and preventing jellyfish attack node in the path.
Support vector machine is utilized for learning packet forwarding behavior. The proposed
technique chooses the node in the network for performing routing of packet on the bases of
hierarchical trust evaluation property of node. The technique is tested using NS2 simulator
algorithms using various parameters; throughput, packet deliver ratio, end to end delay. The result
proves that finding holding period for detecting and preventing jellyfish attack node is highly
efficient in jellyfish attack detection and also perform well as compared to another algorithm.
