Security in Pervasive Computing

Author(s):  
Sunita Prasad ◽  
Rakesh Chouhan

Pervasive computing has wide application in military, medical and smart home domain. In pervasive computing, a large number of smart objects interact with one another without the user intervention. Although the technology is promising but security needs to be addressed before the technology is widely deployed. Pervasive networks are formed spontaneously and the devices communicate via radio. Thus, mobile ad hoc networking is an essential technology for pervasive computing. An ad hoc network is a collection of wireless mobile nodes, which acts as a host as well as a router. The communication between the nodes is multihop without any centralized administration. AODV (Ad Hoc On demand Distance Vector) is a prominent on-demand reactive routing protocol for mobile ad hoc networks. But in existing AODV, there is no security provision against well-known attack known as “Black hole attack”. Black hole nodes are those malicious nodes that agree to forward the packets to destination but do not forward the packets intentionally. Thischapter extends the watchdog mechanism for the AODV routing protocol to detect such misbehavior based on promiscuous listening. The proposed method first detects a black hole node and then gives a new route bypassing this node. The experimental results show that in a lightly loaded, hostile environment, the proposed scheme improves the throughput compared to an unprotected AODV protocol.

Mobile Ad Hoc Networks (MANET) are typically wireless networks that do not have any fixed network architecture. This makes the task of routing layers difficult. A popular reactive routing protocol – Ad hoc On-demand Distance Vector (AODV) used in MANETs has certain vulnerabilities which make it susceptible to black hole attacks. This paper proposes an innovative technique to detect and prevent the possibility of a black hole attack in AODV based MANETs. Using the approach of detecting multiple Route Reply (RREP) messages, authentication of nodes is done by the use of a randomly generated id unique to each route discovery process. This method was tested using simulations in NS2 software and compared with previous attempts. The proposed method showed marked improvement in performance in the presence of malicious nodes.


Mobile ad hoc networks (MANETs) are collection of nodes connected through wireless medium and do not require infrastructure for operation. Network Topology keeps on changing because mobility of nodes are high. Therefore, it is important for MANETs to provide excellent routing and security features. Since MANETs do not require any pre-existing infrastructure, they are extensively used in emergency and rescue and military applications. MANETs thus will form essentially an important part in wireless networks. In this paper, Ad hoc On-Demand Distance Vector (AODV) and Greedy Perimeter Stateless Routing (GPSR) routing protocol performance is compared with respect to Throughput and E2ED and observed that there is an improvement in throughput by 11% in case of GPSR. Simulation is performed using NS3.


Sign in / Sign up

Export Citation Format

Share Document