Exploration of Security Threat Analysis in Wireless Mobile Adhoc Network

摘要

Objective: Mobile Adhoc Networks (MANETs) due to its principle characteristics of network infrastructure, limited resources and transmission range are more vulnerable to multitude of attacks. The objective is to classify the attacks and its counter measures to detect and eliminate malicious nodes. Methods/Analysis: Each attack has been analyzed briefly based on its own characteristics and behavior. Also, the defeating methodologies against network attacks have been described and evaluated as a part of the measurements. We have also proposed an Algorithm to study and analyze networks on affected conditions. We presented analytics and classification of the attacks on the different layers of the network. Findings: On attacked situation in network there will be no data available as the characteristics of the network are unknown. We need to simulate such types of conditions in the network. The performance of the network gets degraded at the time of attack. It was observed that the impact of attack depends on the proximity of the attacker to the source node, it is severe when close and least when far from the source. Each malicious node uses network feature (Distributed Network, Non-centralized, Hop-by-Hop communications, Open network boundary or Wireless media) to break the security. The goal is to violate security service (Availability, Data Confidentiality and Integrity). In our algorithm, we showcased the normalization of the data set such that we get the maximum and minimum values for the classification of the network. The primary groups of data types for classification are: Delay, bandwidth utilization, and drop rate and packet type). There are some secondary classification like conjunction, status of process, running services and utilization of processor. The condition of system was presented as a vector by storing the normalized values in an array. We arrived at simulating a network in attacked situations. Applications/Improvements: The work can be extended to find ways to calculate the threshold effectively. Group attacks can be studied and derive the relationship between the average detection delay and the mobility of the nodes.