SECURITY ENHANCEMENTS IN DYNAMIC TELECAST ROUTING

Transcription

1 SECURITY ENHANCEMENTS IN DYNAMIC TELECAST ROUTING Dr. GVN Prasad [1,3] M.Senthilkumar [2,3] 1 Professor, 2 Associate Professor msk.nvp@gmail.com 3 Annai Mathammal Sheela Engineering College, Erumapatty, Namakkal ABSTRACT Wireless network has no stationary infrastructure. It provides several advantages such as fast and easydeployment, little or no reliance on a pre-existing infrastructure and cost-effectiveness. One of the major issues for data communication over wired and wireless networks is security.different from the past work on the designs of network security algorithms and system infrastructures, we will propose a dynamic telecast routing algorithmthat could randomize delivery paths for data transmission. The algorithm is easy to implement and compatible with popular routing protocols, such as the Routing Information Protocol in wired networks and Destination-Sequenced Distance Vector protocol in wirelessnetworks, without introducing extra control messages. The simulation results have been verified from the proposed algorithm and it shows the capability of the proposed algorithm. Keywords: Security-enhanced data broadcasting, dynamic telecast routing, RIP, DSDV Introduction With the advent of low cost, low power, short range radio technologies, wired and wirelessnetworking has potential applications in many personal and local area networking scenarios.in the last few years, various security measures have been introduced to improve the security of data transmission over public networks.significant amount of current research has been directed to designing efficient dynamic routing protocols for wired and wireless networks. The challenge here is to reduce routing overheads in spite of the changing topology. Several new protocols focused on the issue of overhead reduction without compromising on application-visible performance metrics. Notable among them are a class of protocols called on-demand protocols, e.g., Dynamic Source Routing (DSR) [18] Existing work on security-enhanced data transmission includes the de- signs of network security algorithms and system infrastructures and security-enhanced routing methods. Their common objectivesare often to defeat various threats over the Internet, including eavesdropping, spoofing, session hijacking, etc. Among many well-known designs for network security based systems, the IP Security(IPSec) [2,3] and the Secure Socket Layer(SSL) [21] are popularly supported and implemented in many systems and platforms.althoughipsec and SSL do greatly improvethe security level for data transmission, theyunavoidably introduce substantial overheads[1], [7], [13], especially on gateway/host perrouting mechanism to improve routing security. The proposed algorithm is adopted performance and effective data transmission overhead is 5 cycles/byte over an Intel Pentium II with the Linux IP stack alone, and there with a 94

2 dynamic changing configuration. There overhead increases to 58 cycles/byte whenfore, we will propose a dynamic broadcastadvanced Encryption Standard (AES) [10] isrouting algorithm to provide securityen-adopted for encryption/decryption for IPSenhanced data delivery without introducing any[7].extra control messages. Another alternative for security-enhanced data broadcast transmission is to dynamically route packets between each source and its destination so that the chance for system break-in, due to successful interception of consecutive packets for a session, is slim. The intention of securityenhanced routing is different from the adopting of multiple paths between a source and a destination to increase the throughput of data transmission (see, e.g., [8] and [9]). In particular, Lou et al. [14], [15] pro-posed a secure routing protocol to improve the security of end-to-end data transmission based on multiplepath deliveries. The set of multiple paths between each source and its destination is determined in an online fashion, and extra control message exchanging is needed. Bo-hacek et al. [2] proposed a secure stochastic routing mechanism to improve routing security. Similar to the work proposed by Lou et al. [14], [15], a set of paths is discovered for each source and its destination in an online fashion based on message flooding. Thus, a mass of control messages is needed. Yang and Papavassiliou [25] explored the trading of the security level and the traffic dispersion.they proposed a traffic dispersion scheme to re-duce the probability of eavesdropped information along the used paths provided that the set of data delivery paths is discovered in advance. Although excellent research results have been proposed for security-enhanced dynamic broad cast routing, many of them rely on the discovery of multiple paths either in an online or offline fashion.for those online pathsearching approaches, the discovery of multiple paths involves a significant number of control signals over the Internet. On the other hand, the discovery of paths in an offline fashion might not be suitable to networkswith a dynamic changing configuration. There-fore, we will propose a dynamic broadcast routing algorithm to provide securityenhanced data delivery without introducing any extra control messages. The objective of this work is to explore a securityenhanced dynamic broadcast routing algorithm based on distributed routing in-formation widely supported in existing wired and wireless networks. We aim at the randomization of delivery paths for data transmission to provide considerably small path similarity (i.e., the number of common links between two delivery paths) of two consecutive transmitted packets. The proposed algorithm should be easy to implement and compatible with popular routing protocols, such as the Routing Information Protocol (RIP) for wired networks [16] and Destination- Sequenced Distance Vector (DSDV) protocol for wireless networks [20], over existing infrastructures. These protocols shall not increase the number of control messages if the proposed algorithm is adopted. 2. Routingprotocolsforadhocnetworks 2.1.Linkstateprotocols In link state protocols, each node maintains its own view of the network topology, including link costs of all its outgoing links. To keep view sup-to-date, each node broad- casts the link costs of all its neighbors 2 to all other nodes in the network using flooding. Thisisdonewheneverthereis achange in link costs. As a node receives this information, 95

3 it updates its view of the network topology and applies a shortest path algorithm (Dijkstra sshortest path algorithm [8] in our simulations) to choose the next hop toadestination. Asynchronouslinkcostupdatesmaygive rise to short-lived routing loops; however, they disappear by the time update messages have propagated throughout thenetwork[19]. In our simulation we used the SPF implementation of link state protocol as described in[32]. 2.2.Distancevectorprotocols In the distance vector approach, for each destination i,everynode j maintains a set of distance sorcosts,dik(j) 1 Except in are sent work atcmu[3]done concurrently with our work. 2In a mobile, adhoc network any node can potentially be a neighbor Where k ranges over the neighbors of i. Nodekistreatedas the next hop node for a data packet destined for i, if dik(j)=min k{dik(j)}. To keep these distances up-to- date, whenever there is any change of this minimum distance because of link cost changes, the new minimum distance is reported to the neighboring nodes. If, as a result, a minimum distance to any neighbor changes, this processisrepeated. This technique is the classical distributed Bellman Ford algorithm[2].routing loops, both short-lived and long-lived, are possible in the distributed Bellman Ford algorithm. There is also a possibility of the counting-toinfinity problem, where it takes a large number of update messages to detect that a node is unreachable [19]. Several protocol shave been pro- posed to avoid the longlived loops and counting-to-infinity problems. They typically work by increasing the amount ofinformationexchangedbetweennodesorpro vidingsome sortofinternodalcoordination. For example, in the Border Gateway Protocol (BGP)the entire path between the source and the destination is sent instead of just the distance[30]. InDUAL(DistributedUpdateAlgorithm)[10]i nter-nodal coordinationisachievedviaatechniqueknown asdiffusing computation. We focus our attention on two distance vector proto- cols. The first, extended Bellman Ford, achieved good performance for stationary networks in earlier simulation studies [32]. The second, DSDV [29], was specifically proposed for mobile, wired and wireless networks. 3.Problem Statement The objective of this work is to explore a security-enhanced dynamic broadcast routing algorithm based on distributed routing in-formation widely supported in existing net-works. In general, routing protocols over networks could be classified roughly into two kinds: distancevector algorithms and link-state algorithms [11]. Distance-vector algorithms rely on the exchanging of distance in-formation among neighboring nodes for the seeking of routing paths. Examples of distance-vector-based routing algorithms include RIP and DSDV. Link-state algorithms used in the Open Shortest Path First protocol [19] are for global routing in which the network topology is known by all nodes. Our goal is to propose a distance-vector-based algorithm for dynamic broadcast routing to improve the security of data transmission. 96

4 Consider the delivery of a packet with the destination t t a node Ni. In order to minimize the probability that packets are eaves dropped over a specific link, a randomization process for packet deliveries shown in Procedure 1 is adopted. In this transmission. The exclusion of hs for the next hop selection avoids transmitting two consecutive packets in the same link, and the randomized pickup prevents attackers from easily predicting routing paths for the coming transmitted packets. Procedure RANDOMIZEDSELECTOR (s, t, pkt) The number of entries in the history record for packet deliveries to destination nodes is N in the worst case. In order to efficiently look up the history record for a destination node, we maintain the history record for each node in a hash table. Before the current packet is sent to its destination node, we must randomly pick up a neighboring node excluding the used node for the previous packet. Once a neigh-boring node is selected, by the hash table, we need O(1) to determine whether the selected neighboring node for the current packet is the same as the one used by the previous packet. Therefore, the time complexity of searching a proper neighboring node is O(1). Routing Table Maintenance Let every node in the network be given a routing table and a link table. We assume that the link table of each node is constructed by an existing link discovery protocol, such as the Hello protocol in [18]. On the other hand, the construction and maintenance of routing tables are revised based on the well- known 97

5 Bellman-Ford algorithm [4] and described as follows: Initially, the routing table of each node (e.g., the node Ni) consists of entriesbetween neighboring nodes, the routing table of Ni is accordingly updated. Note that the ex-changing for distance vectors among neighboring nodes can be based on a predefined inter-val. The exchanging can also be triggered by the change of link cost or the failure of the link/node. In this paper, we consider cases when Ni receives a distance vector from a neighboring node Nj. Each element of a distance vector received from a neighboring node Nj includes a destination node t and a delivery cost WN j,t from the nodenj to the destination node. their delivery costs are the same as that of the minimal cost path. Performance Evaluation The purpose of this section is to evaluate the performance of the proposed algorithm, referred to as the DDBRA. We compare the performance of DDBRA with the popular Shortest-Path Routing Algorithm (SPRA) and the Equal-Cost Routing Algorithm (ECRA) used in RIP. In SPRA, only one path with the minimal cost is derived for each source destination pair. On the other hand, more than one path can be accommodated in ECRA if The primary performance metric is the aver-age value [ESimPS]of path similarity for all source- destination pairs in PS. The values of [ESimPS] is calculated by the following procedure: For each sourcedestination pair with the length of the minimal-cost path equal to l, a considerable number of packets are transmitted from the source node to the corresponding destination node. The average path similarity of the sourcedestination pair is calculated by summing the path similarity of each two consecutive packets divided by the packet number minus 1. The same operation is done for the rest of source destination pairs. Finally, the value of [ESimPS] can be obtained by averaging the path similarity of all source-destination pairs with the length l of minimal-cost paths. In the AT&T US topology, the cost of each link is set as 1 or 4, depending on the bandwidth of each link. The bold lines in Fig. 2 represent the links with 9.6-Gbps bandwidth. The bandwidth of any other 98

6 links is equal to 2.4 Gbps. In The simulated traffic is constant bit rate (CBR) over User Datagram Protocol (UDP). The interval time ofcbr is 10msand the packet size is 1,000 bytes. The simulation time is set to 100 seconds. In addition to path similarity, the per-formance of the proposed algorithm.ddra_withoutrandomizedsel ector, ECRA, and SPRA under the AT&T topology. From this figure, we observe that our DDRA-based methodologies greatly outperform SPRA and ECRA for all l values under investigation,1 which indicates that our DDRA provides larger path variation and, thus, more secure packet routing. Also, the E[simps]values for SPRA, ECRA, and DDRA_without_Randomized Se-lector increases. The increasing rates for SPRA and ECRA are much larger than those for DDRA_without_Randomized Selector especial-ly when l is large. Specifically, the E[simps]value for SPRA is the same as the length of minimal-cost path because all packets always go through the minimal-cost path be-tween sourcedestination pairs.onthe other hand,whenl increases,[simps]forddra_with_random izedselector increases and then decreases. For all l values, the performance of DDRA_with_Randomized Selector is better than that of DDRA_without_Randomized Selector. The RandomizedSelector can prevent from selecting the previous nexthop for the current packet delivery and therefore avoids that consecutive packets are transmitted to the same nexthop. Fig.3 shows the experimental results of E[Simps]for DDRA_with_RandomizedSelector, Conclusion This paper has proposed a security-enhanced dynamic routing algorithm based on distributed routing information widely supported in existing networks. The proposed algorithm is easy to implement and compatible with popular routing protocols, such as RIP and DSDV, over existing infrastructures. An analytic study was developed for the proposed algorithm and was verified against the experimental results. A series of simulation experiments were conducted to show the capability of the proposed algorithm, for which we have very encouraging results. We must point out that the proposed algorithm is completely orthogonal to the work based on the designs of cryptography algorithms and system infrastructures. Our security enhanced dynamic broadcast routing could be used with cryptographybased system designs to further improve the security of data transmission over networks. 99

7 References [1] C. Alaettinoglu. A.U. Shankar. K. Dussa- Zieger and I. Matta. Des ignand implementation of MaRS: A routing testbed. Journal of Internetworking: Research and Experience 5(1) (1994) [2] IiBertsekas and R Gallager. Data Networks (Prentice-HaILEnglewood Cliffs.NJ. 1987). [3] J. Broth. D.A. Maltz. D.B. Johnson. 1.-C. Hu and 3.Jetcheva.A perforinance comparison of multi-hop wireless adhoc network routing protocols.in: Proc. of the 4th Internal. Conf on Mobile Computing and Networhng (ACM MOBICOM98) (October 1998) pp [4] C. Cheng. K Riley and S.P.R Kumar. A loop-free extended Bellman Ford routing protocol without bouncing effect. in: Plvc. of AcMSIGCOMM conf(1989) pp [5] MS. Corson. S. Batsell and 3. Macker. Architectural cons iderations for mobile mesh networking (May 1996) Request for comments draft: http: //tonnant.itd. nrl.navy.mil/ mmnet/mnetrfc.txt. [6] M.S. Corson and A. Ephremides.A distributed routing algorithm for mobile wireless networks, Wireless Network 1(1) (1995) 61 SI. [7] S. Corson..S. Papadenietriou. P. Papadopoulos. V. Park and A. Qayvum. An Internet MANET EncapsuLationProtocol (tmep) specification.internet Draft (1998), internetdraft8/draft- jet f-manet-imep-spec-o1.txt. [8] E.W. Dijkstra. A note on two problems in connection with graphs.numerical Mathematics 1 (1959) [9] E. Gafni and D. Beitsekas.Distributed algorithms for generating loop-free routes in networks with frequently changing opology.ieee Transactions on Communication 29(1) (1981) [10] J.J. Garcia-Luna-Aceves. A unified approach to ioop free routing using distance vectors or link states, in: Proc. ofacafslgcom1f 89 Conf (September 1989) pp [11] M. Gerla and it- C.Tsai.Multicluster.mobile. multimedia radio networks. Wireless Network 1(3) (1995) [12] Z. Hass, A new routing protocol for the reconfigurable wireless networks, in : Proc. Of the IEEE Internet. Conf. on Universal Personal Communications (October 1997). 13] Z.J. Hass and M.R. Pearlman, The Zone Routing Protocol (ZRP) for ad hoc networks, IETF Internet Draft (1997), [14] Z.J. Hass and M.R. Pearlman, The performance of query control schemes for the zone routing protocol, in : Proc. Of ACM SIG- COMM 98 Conf. Vancouver (September 1998) pp [15] D. Sidhu, R. Nair, and S. Abdallah. Finding Disjoint Paths in Networks. In Proceedings of the ACM SIGCOMM, pages 43 51, [16] B. Tuch. Development of WaveLAN, an ISM band wireless LAN. AT&T Technical Journal, 72(4):27 33, July/Aug [17] S. Vutukury and J. J. Garcia-Luna-Aceves. MDVA: A Distance-Vector Multipath Routing Protocol. In Proceedings of the IEEE INFOCOM, to appear. [18] W.T. Zaumen and J. J. Garcia-Luna- Aceves. Shortest multipath routing using generalized diffusing computations. 100

8 [19] S. J. Lee and M. Gerla. Split Multipath Routing with Maximally Disjoint Paths in Ad Hoc Networks. In Proceedings of the IEEE ICC, 2001.to appear. [20] M. K. Marina and S. R. Das. Performance of Route Caching Strategies in Dynamic Source Routing. In Proceedings of the Int l Workshop on Wireless Networks and Mobile Computing (WNMC) in conjunction with Int l Conf. on Distributed Computing Systems (ICDCS), pages , Apr [21] N. F. Maxemchuk. Dispersity Routing.In Proceedings of the IEEE ICC, pages , [22] J. Moy. OSPF version 2. RFC 1247, July [23] S. Murthy and J. J. Garcia-Luna- Aceves.Congestion-Oriented Shortest Multipath Routing. In Proceedings of the IEEE INFOCOM, March [24] A. Nasipuri, R. Castaneda, and S. R. Das. Performance of Multipath Routing for Ondemand Protocols in Mobile Ad Hoc Networks.ACM/Baltzer MONET Journal, 2001.to appear. [27] VD. Park and MS. Corson, A highly adaptive distributed routing algorithm for mobile wireless networks, in: Proc. of IEEE INFOC OM 97 Conf (April 1997). [28] C. Perkins. Ad hoc Ondemand Distance Vector (AODV) routing. IETF Internet draft (1997). internet-drafts/draftietf-manet-aodv-0o.txt. [29] C.E. Perkins and P. Bhagwat.Highly dvnazncdestination- Sequenced Distance-Vector routing (DSDV) for mobile computers. in: Proc. of the ACM SIGCOMM 94 Conf (August 1994) pp [30] Y. Rekhter and T. Li.A border gateway protocol-4.rfc 827 (March 1995). [31] N. Schacham and 3.Westcott.Future directions in packet radio architecturesand protocols. Proceedings of the IEEE 75(1) (1987) [32] AU. Shankar. C. Alaettinoglu. K Dussa- Ziegerand I. Matta. Trans ientand steady-state performance of routing protocols: Distancevector versus link-state, Journal of lnternetworking: Research and Experience 6(1995) (Preliminary version appeared in: Proc. ofacmsigmetrics/performance Conf(1992) pp ) [33] Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications. IEEE Standard IEEE Standards Department (1997). [25] V. Park and S. Corsou. A performance comparison of TORA and ideal link state routing.in: Proc. of IEEE S iinposlum of Computers and Communication (June 1998). [26] V. Park and S. Corson..Temporally Ordered Routing Algorithm (TORA) Version 1. Functional specification IETF Internet draft (1998). ietf.org/internet-drafts/ draft-ietf-manet-tora-spec-o1.txt. 101