You are on page 1of 4


Dr.G.K.D.Prasanna Venkatesan1, S.Jegadeesan 2. 1 Professor, Department of ECE, PGP College of Engineering and Technology, Namakkal. 2 Assistant Professor, Department of ECE, V.S.B Engineering College, Karur. Email, Abstract In wireless sensor networks (WSNs) the main attacks are compromised-node and denial-of-service. The existing multipath routing approaches are easily vulnerable to those attacks. So if the adversary finds the routing algorithm then it can easily identify the routes and the information is made available to such attacks. In this paper, randomized multipath routing is used. In this mechanism the shares of packets change the routes with respect to time. So it cannot find the routes which is travelled by each packet even the adversary knows the routing algorithm. Secure Message Transmission (SMT) mechanism proposed in continuously updates the rating of the routes. It is more energy efficient and highly dispersive. The end to end energy consumption is minimized. Index TermsRandomized multi-path routing, wireless sensor network, secure data delivery, energy efficiency, multicast tree. I INTRODUCTION In wireless sensor networks the two key attacks are compromised-node (CN) and denial-of-service (DOS). In the compromised node attack, the adversary is used to compromise the node but in the denial of service attack, the adversary disturbs the normal operation of the network by changing, the functions of a subset of nodes. These two attacks generate black holes within which the adversary can block information delivery. It is carried through a two-step process. First, the packet is splitted into M shares threshold secret-sharing mechanism such as the Shamirs algorithm. The information can be recovered from T shares, Second, multiple routes from the source to the destination are determined according to multi-path routing algorithm. These routes are node-disjoint. The M shares are then distributed over these routes and it reaches the destination. SEQUENCE DIAGRAM

Secure Message Transmission (SMT) mechanism proposed in continuously is used to update the rating of the routes. For each successful share, the rating of the particular route goes up. By adjusting the random propagation and secret-sharing parameters as N and M, the different security levels can be given by these algorithms at different energy costs. CASE DIAGRAM

random relays. When the TTL value is down to 0, the last node which has to receive this share starts to route it towards the sink by using minimum-hop routing. Once the sink collects T shares, it can regenerate the original packet. The information cannot be reconstructed from less than T shares. In routes of higher dispersiveness the black hole is avoided but in routes of lower dispersiveness the packet is blocked by the black hole and it stops forwarding it to the destination. II RANDOMIZED DELIVERY MULTI-PATH III MULTICAST TREE-ASSISTED RANDOM PROPAGATION Multicast tree involves only directionality in its tree process and it does not need of location. Muli-cast tree is constructed from itself to every node network. Such tree construction is usual in existing protocols, and is conducted by flooding a hello message from the sink to every node. Once if a multicast tree is constructed between the sink and the node, a node knows is distance to the sink and the id of its parent node on the tree. Each entry in the neighbour list maintained by a node has a field that records a number of hops to the sink from the corresponding neighbor. IV ANALYSIS OF THE PRP SCHEME The random routes generated by the algorithms are not node-disjoint. Even it is node-disjoint it avoids the black hole by conducting asymptotic analysis of the PRP scheme. It is useful in serving as a lower bound on the performance of the propagations like NRRP, DRP, and MTRP schemes. The security analysis for the CN and DOS attacks are determined as similar because both of them involves in the calculation of the packet interception probability. The same treatment can be applied to the DOS attack with a straight forward modification. Network and Attack Models

We use the approaches like randomized propagation of each information share, and normal routing (e.g., minhop routing) toward the sink. More specifically, when a node has to send a packet to the sink, it first splits the packet into M shares according to a threshold secret sharing algorithm like Shamirs algorithm. Then each share is transmitted to neighbour which is selected randomly. The neighbour which receives the share will send the share to other randomly selected neighbours, and it goes on.

In this transmission TTL field concept is used, where the initial value is set by the node which sends the first packet to control the total number of

An area S is considered which is uniformly covered by sensors with density . A unit-disk model is assumed for the communication of sensor i.e., the transmitted signal from a sensor can be successfully received by any sensor which is at Rh meters away. Multi-hop relay is used if the destination is more than Rh away from the source. The link-level security is assumed that it has been established through a conventional cryptography-based bootstrapping algorithm. Security Definition For a given source sensor node, the protocol provides the security which is defined as the worst-case (maximum) probability that for the M shares of an information packet which is sent from the source node, T of them are intercepted and attacked by the black hole. Mathematically, this can be defined as follows. The distance between the source s and the sink o be taken as ds. It is defined that a series of N+1 circles is co-centered at s. For the ith circle, 1 i N, the radius is iRh. For circle 0, its radius is 0. These N +1 circles will be given to as the N-hop neighborhood of s. More intentionally, we say that a node is i hops away from s if it is located within the intersection between circles i 1 and i. We refer to this intersection as ring i. For an arbitrary share, after the random propagation phase, the parent id of the ring in which the last receiving node w, is located is a discrete random variable with state space f1; : : : ;Ng. The actual path from w to the sink is decided by the specific routing protocol employed by the network. According to that, different packet interception rates are obtained under different routing protocols. However, the route given by mininimum-hop routing, which is under high node density can be approximate by

the line between w and the sink, which is used to give an upper bound on the packet interception rates under all other routing protocols. This can be justified by min-hop routing tends not to distribute traffic over various intermediate nodes and only selects those nodes that are closest to the sink. The path-concentration effect makes min-hop routing have a smaller traversing area of the paths, especially when it is compared to the power-balancing routing protocols that build dispersive routes. The interception probability for an arbitrary share of information is given by

Accordingly, the worst case probability that at least T out of M shares are intercepted by E is given by

To proceed with the security analysis, we have to calculate the shaded area in each ring of Si V SIMULATION STUDIES In this simulation setup is used in order to evaluate the PRP, NRRP, DRP, and MTRP performance under more sttings of real. To understand the capability of these randomized multi-path routing algorithms a better manner in bypassing black holes, we compare the performance of schemes like PRP, NRRP, DRP, and MTRP against H-SPREAD algorithm, which generates node-disjoint multi-path routes. End to end delay

Packet delivery ratio

VI CONCLUSIONS Thus it is concluded that by using the randomized multipath routing algorithm the packet transmission is made more secure. The adversary cannot determine the path taken by the routes even if it knows the routing algorithm. In the proposed algorithm the packet retransmission does not occurs so the energy consumption is reduced. Hence this algorithm is highly dispersive and energy efficient.

The authors thank the management of V.S.B Engineering College, Karur for providing the necessary facilities to undertake the above work. REFERENCES

[1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci. A survey on sensor networks. IEEE Communications Magazine, 40(8):102 114, Aug. 2002. [2] C. L. Barrett, S. J. Eidenbenz, L. Kroc, M. Marathe, and J. P. Smith. Parametric probabilistic sensor network routing. In Proceedings of the ACM International Conference on Wireless Sensor Networks and Applications (WSNA), pages 122131, 2003. [3] M. Burmester and T. V. Le. Secure multipath communication in mobile ad hoc networks. In Proceedings of the International Conference on Information Technology: Coding and Computing, pages 405409, 2004. [4] T. Claveirole, M. D. de Amorim, M. Abdalla, and Y. Viniotis. Securing wireless sensor networks against aggregator compromises. IEEE Communications Magazine, pages 134 141, Apr. 2008. [5] D. B. Johnson, D. A. Maltz, and J. Broch. DSR: the dynamic source routing protocol for multihop wireless ad hoc networks. In C. E. Perkins, editor, Ad Hoc Networking, pages 139172. AddisonWesley, 2001. [6] P. C. Lee, V. Misra, and D. Rubenstein. Distributed algorithms for secure multipath routing. In Proceedings of the IEEE INFOCOM Conference, pages 19521963, Mar. 2005. [7] P. C. Lee, V. Misra, and D. Rubenstein. Distributed algorithms for secure multipath routing in attack-resistant networks. IEEE/ACM Transactions on Networking, 15(6):14901501, Dec. 2007. [8] S. J. Lee and M. Gerla. Split multipath routing with maximally disjoint paths in ad hoc networks. In Proceedings of the IEEE ICC Conference, pages 32013205, 2001.