Methods For Detecting Attacks In Mobile/Wireless Ad-Hoc Networks: A Survey

摘要

Use of Mobile/Wireless ad hoc networks (MANET) is growing mostly in situations there is need for temporary data exchange such as in emergency services, conference meetings, virtual classrooms due to their low cost and easy in setting up, since they are infrastructure-less. However none existence of centralized administration of standard support services, dynamic topology, and air interface makes them highly vulnerable to attacks. The autonomous nature of the network results in security solutions being implemented mainly at the nodes, which are resource constrained i.e. limited computing power and memory, low bandwidth, and low battery life, thereby limiting the tightness of these solutions. It is extremely important to have flexible and robust methods of detecting attacks and their sources looking at the diversity of MANET application areas, so that appropriate countermeasures can be put in place. Detection based schemes can be incorporated to complement prevention techniques implemented in the network protocols. These methods should work through the whole protocol stack since attacks target specific layers. This paper discusses MANET attacks detection methods that are currently in use taking note of their performance. In the paper we also suggest a hybrid cross layer approach capable of detecting more than one known and new attacks, which can be researched on to complement existing methods in building effective security solutions for MANET.