Energy efficient reputation mechanism for defending different types of flooding attack

Delay tolerant network solves technical challenges in the heterogeneous network that may lack end-to-end connectivity. However, due to the disconnected paths, message delivery is much dependent on the cooperation of intermediate nodes, but malicious nodes may inject other nodes with either bogus messages or copies of good messages. This causes relaying of unwanted packets, which in turn leads to draining the energy of the intermediate nodes. This scenario may be termed as flooding attack. The paper discusses three types of flooding attacks, namely breadth attack (type 1), breadth attack (type 2), and depth attack. Breadth attack (type 1) refers to attack by those malicious nodes that relay only bogus messages, breadth attack (type 2) refers to the attack by those malicious nodes that relay both bogus and good messages, and depth attack refers to the attack by those malicious nodes that create copies of its own good messages and floods in the network. In this paper, we present a novel reputation based schemas that detect the flooding type of malicious nodes in DTNs. We propose three algorithms where first algorithm Reputation Algorithm handles a breadth attack (type 1), second algorithm Reputation with Good Messages over Total Messages Algorithm handles both breadth attacks (type 1 and type 2), and third algorithm, Reputation using Good Messages over Total Messages with Check message Generation Rate (RepGMTMwithCGR) is robust for (depth attack \(+\) breadth attack) flooding attack. The simulation study shows RepGMTMwithCGR defends all categories of flooding attack considered in this paper and there is the improvement of 20% in message delivery, 61% decrease in relaying, 49% decrease in message dropping and 63% decrease in energy consumption in the presence of 30% malicious nodes in the network. The algorithm shows increase in message latency by 2% but decreases message latency by 25% when compared to existing work. The protocols can be employed in monitoring systems using base stations in open environments.