SNACK: An efficient intrusion detection system in Mobile Ad-Hoc Network based on the Selective-Negative Acknowledgement algorithm

Abstract

The Mobile Ad-Hoc Network (MANET) consists of independent devices connected together, which can change their locations and configure themselves without being controlled by a central unit. This autonomous topology of MANET makes it vulnerable against the internal attacks, such as black hole, wormhole, and flooding, from inside the system. One existing solution to this problem has been achieved by using Negative Acknowledgement (NACK) as an Intrusion Detection System (IDS). NACK method is easy to implement and has a high level of packet delivery with lightweight security monitoring. However, although packet delivery is guaranteed in NACK, its high rate of routing overhead and high level of energy consumption become as two big weaknesses of the network, especially when it comes to increasing the mobility and the number of insider attacks. In this study the performance of NACK in this regard has been challenged and investigated in different scenarios. Then a new approach, called Selective Negative Acknowledgement (SNACK), based on NACK and Selective Acknowledgement (SACK) is proposed. It is shown that the proposed acknowledgement method outperforms NACK with much less packet overhead, by comparing the results of simulations in Network Simulator v-2.35 (NS-2).