Abstract
Wireless sensor nodes are low cost tiny devices deployed in unsupervised environment. These devices require some kind of security mechanism to tackle different kinds of attacks. Intrusion detection is one such methodology used where the malicious activity is observed, and the target is isolated. Many authors have suggested different methodologies which are detailed in the work. The article also uses intrusion detection scheme to provide reliable network. A mobile agent and an assistance node aids in increasing security and efficiency of the work. The illegitimacy of the node is inferred using the data provided by the assistance and mobile agent. The work minimizes 12.9% of energy consumption, increases security by 3.025%, and increases reliability by 7.07% compared to the previous work.
TopBackground
Many authors have provided their own insights (Butun, Ismail, Morgera, & Sankar, 2014) and suggested different ways to detect an intrusion in the sensor network. The details of the same are explained in this section.
In (mehmood, et al., 2018) the network is divided into clusters and each cluster is headed by the cluster head. The base station has a blank knowledge database installed. The cluster heads are provided with inference engines. The heads monitor node-related events and data transmission in the respective clusters. The data monitored is considered as events and the same is transmitted to the base station. This received data is analysed to detect suspicious activity. Routine and redundant events are eliminated and same is notified to the respective cluster heads. The threats detected in the received message are notified to the cluster heads. The same is broadcasted to the other cluster members. Any unknown event is analysed and alerted to the cluster heads to avoid transmission. The cluster head is rotated among the members of the cluster for effective and energy consumption of the nodes.
Three stages are proposed by the authors in (Silva, et al., 2005) to detect an intrusion in the network. Filtered messages are collected by the promiscuous node in data acquisition phase. Using this approach the energy is saved. The array storage structure is used to store the discarded messages. In the rule application phase the rules are applied to stored data in array. If the tested data again results in failure, the data is discarded. The approach reduces detection latency. Failure alert is raised if the stored data does not obey the stated rules. An intrusion detection alert is raised if the failure alert is raised beyond a threshold. The failure history is maintained by the respective node. The previous and the present cumulative failure value are combined to obtain new cumulative value. The approach follows deviation failure.
Key Terms in this Chapter
Smart Environment: Smart environment is an assembly of sensors, actuators and many computational elements providing services for betterment of human life.
Wireless Sensor Network: Wireless sensor network (WSN) refers to a group of spatially dispersed and dedicated sensors deployed to monitor/track any object of interest.
Cluster-Based Routing: The nodes in the network are grouped together in the form of clusters. The nodes transmit their readings to the group head which later transmits the aggregated data to the next available hop.
Mobile Agent: Mobile agents are the set of dynamic nodes capable to moving from one end of the network to another and doing their regular routines similar to static nodes of their calibre.
Knowledge-Based Anomaly IDS: The nodes after certain amount of observation of the network are provided with certain observations that aid them in locating illegitimate nodes.
Intrusion Detection Scheme: The methodology adopted aids in detecting the compromised nodes in the network.
Assistance Node: These are nodes deployed with other nodes in the network to aid the other nodes in their doings.