Efficacious Distributed Arbitrary Node Duplication Attacks in ...

International Journal of Computer Trends and Technology- volume3Issue1- 2012

Efficacious Distributed Arbitrary Node Duplication Attacks in Wireless Networks Sudheesha Cheepi#1, Venkata Sumanth Mareedu*2, Venkata Durga Kiran Kasula#3 12

M.Tech 2nd Year, 3 Asst.Professor, Department of C.S.E, K L University, Vaddeswaram.

Abstract: Due to the off-the-shelf hardware components in unshielded network, nodes leave them vulnerable to compromise with little effort, an adversary may capture nodes, analyze and replicate them, and surreptitiously insert these replicas at strategic locations within the network it is generally assumed that an adversary can capture and compromise a small number of nodes in the network to corrupt network data or even disconnect significant parts of the network. Without an effective and efficient detection mechanism, these replicas can be used to launch a variety of attacks that undermine many network applications and protocols. In this paper, we present a novel distributed approach called Efficacious distributed arbitrary technique for detecting node replication attacks. Keywords- Node Replication, Replica Detection, Deterministic Multicast, Localized Randomised Multicast.

I.

INTRODUCTION

A. Centralized Detection The most straightforward detection design requires each node to launch a record of its neighbors and their claimed locations to the base station. The base station can then examine every neighbor record to look for replicated nodes. If it discovers one or more replicas, it can revoke the replicated nodes by flooding the network with an authenticated revocation message. While conceptually uncomplicated, this move toward suffers starting numerous drawbacks inherent in a centralized system. First, the base station becomes a single point of failure. Any compromise of the base station or the Communication channel around the base station will render this protocol ineffective. Furthermore, the nodes adjoining to the base station will obtain the impact of the routing load and will be converted into attractive targets for the adversary.. In terms of security, this protocol achieves 100% detection of all replicated nodes, assuming all messages fruitfully realize the base station. As far as efficiency, if we assume that the average path length I to the base station is √ and each node has an average degree d (for d 0 and > 0. Resilience against Node Compromise In SDC, witness nodes are chosen indiscriminately commencing the nodes of a specified cell instead of the intact network as in the prior algorithm. Therefore, pretentious that the adversary’s competence of compromising nodes is inadequate, spontaneously in SDC the likelihood that an adversary can compromise all the witness nodes storing the location claim of a prearranged identity, i.e. ., is privileged than that of the line selected Multicast algorithm. Assuming that the adversary has compromised t nodes in cell D, . Can be calculated as follows: =

=

(

)! !

!.(

)!

eq1 where ( (

=

=(

=(

)(

)…..

)(

)….

! )! !

)

Where t ≥ ω. 2) Parallel Multiple Probabilistic Cells Like SDC, in the P-MPC system, a geographic hash function [10] is employed to map node L’s uniqueness to the target cells. Let c= {c1,c2,c3,……. } signify the set of cells to which a place claim (actually, the identity of the sender) is mapped. Let indicate the probability that the position claim is forwarded to cell . The subsequent two circumstances must be contented while decisive ′ : () ( )

=1 ≥

When i > j for i, j∈ {1,2, … . , } When l broadcasts its location claim, each neighbor autonomously decides whether to promote the claim in the similar way as in the SDC scheme. The neighbors that promote the claim can conclude the destination cell based on a geographic hash function and the predetermined probabilistic distribution of ′ . More exclusively, the neighbors first analyze the deposit of cells (C) to which the distinctiveness of the sender are mapped, based on a geographic hash function with the input of . Then, every neighbor that forwards the claim autonomously generates a

http://www.internationaljournalssrg.org

Page 60

International Journal of Computer Trends and Technology- volume3Issue1- 2012 arbitrary number z ∈ [0, 1). presume that j is the negligible amount that satisfies < ∑ ( < ( ∈ {1,2, … . , }), this neighbor chooses the j th cell (i.e., ) as the objective cell for the location claim. Once the location claim arrives at cell , the sensor receiving it first verifies whether a member of C which can be intended is based on the geographic hash function and the distinctiveness programmed in the claim message Detecting Replicas Let designate the deposit of every combinations of choosing 1 to v-1 elements commencing C, i.e., the deposit of the cells to which is mapped. If the node replication attack is not detected when the adversary adds replica l2 to the network, this implies that the location claims for l2 were forwarded to a deposit of cells that do not have any nodes that accumulate the prior location claims of . Let , represent the prospect that the location claim of l1 is forwarded to every cells in C except the cells in , which is an element of . Let , refer to the probability that the location claim of is forwarded to any cell(s) in . ∑| | , . , . For a given ∈ , let refer to all the combination of choosing 1 to | | -1 elements from . We denote as the set of cells that hoard the location claim from but not , and . Let denote the probability that the location claim of is forwarded to all the cells in C except the cells in Ce1, which is an element of , . Let , denote the probability that the location claim of is forwarded only to all the cells in Ce2. Let is , denote the probability that the location claim of forwarded to any cell(s) in Ce1 except those in . Thus, we have: |

=

||

|

.

,

.

,

Resilience against Node Compromise Let ( ) and (t) denote the functions that output the pts of the SDC scheme and the P-MPC scheme, respectively, when the numeral of the compromised nodes is t. pretentious that the adversary’s capability of compromising nodes is delimited by ∆ , we have ∑ = ∆ , where is the add up to nodes compromised in cell . Let denote the deposit of all the combinations of choosing 1 to v elements from C. For any element in denoted As , the probability that the adversary controls all the witnesses of a given uniqueness, when such a set of cells in C (i.e., ) are selected as the intention cell(s), is the result of all the individual probabilities ’s of the cells. Let pi refer to the probability that accurately the cells in are chosen as the destination cells by the r neighbors that forward the location claim. Let ( ) refer to the of the j th cell of when the numeral of nodes

ISSN: 2231-2803

compromised in this cell is tj . Thus calculated as follows: |

( )=

|

|

(t), can be

|

( .

( ))

Note that in Equation (4), | | denotes the add up of all the combinations of choosing 1 to v elements from C, while | | denotes the numeral cells restricted in a selected combination, i.e. . In additional, = 1 when there is no witness in the jth cell of . ( )=

.

( )

The accomplishment rate that adversaries organize all the witnesses of a given uniqueness is condensed by a factor of 1- . III. LOCALISED RANDOMISED MULTICAST To progress the resiliency of the prior multicasts, we put forward a innovative protocol that randomizes the witnesses for a known node’s location claim, so that the adversary cannot anticipate their identities. When a node announces its location, each of its neighbors sends a replica of the location claim to a deposit of indiscriminately chosen witness nodes. If the adversary replicates a node, then two sets of witnesses will be chosen. In a network of n nodes, if each location produces √ witnesses, then the birthday paradox predicts at least one collision with lofty probability, i.e., at least one witness will obtain a couple of contradictory location claims. The two contradictory locations claims figure adequate verification to invalidate the node, so the witness can flood the pair of locations claims from side to side the network, and each node can autonomously prove the revocation verdict. These protocols presume that each node knows its own location. We also presume that the network utilizes an identity-based public key system such that every node α is deployed with a private key, ∝ , and any other node can analyze α’s public key using α’s ID, i.e., Kα=f(α). If essential, we could substitute this classification with a supplementary traditional PKI in which we assume the network authorities use a master public/private-key pair (/(KM, ) to sign α’s public key; however, transmitting this public-key certificate will have a substantial communication overhead. A. Description At a high level, the protocol has each node transmit its location claim, along with a signature authenticating the claim. Each of the node’s neighbors probabilistically forwards the claim to an arbitrarily chosen set of observer nodes. If any observer receives two dissimilar location claims for the similar node ID, it can retract the simulated node. The birthday paradox makes sure that we notice

http://www.internationaljournalssrg.org

Page 61

International Journal of Computer Trends and Technology- volume3Issue1- 2012 replication with high likelihood using a moderately limited number of witnesses. More officially, each node α transmit a location claim to its neighbors, β1,β2,β3,…..βd. The location claim has the ( ∝, ∝, { ( ∝ , ∝ )} ∝ ) arrangement where represents ∝ ’s location (e.g., geographic coordinates( , ). Upon hearing this statement each neighbor βi, confirm α’s signature and the plausibility of 1, (for example, if each node knows its own position and has some knowledge of the maximum propagation radius of the communication layer, then it can loosely bound α’s set of potential locations Each witness that receives a location claim first verifies the signature. Then, it checks the ID next to all of the location claims it has conventional thus far. If it ever receives two dissimilar locations claims for the similar node ID, then it has notice a node replication attack, and these two locations maintain serve as confirmation to revoke the node. It blacklists α from further communication any immediately f loading the network with the pair of conflicting location claims, lα and lα's. Each node in receipt of this pair can independently verify the signatures and agree with the revocation decision. Thus, the sensor network both detects and defeats the node replication attack in a fully distributed manner. Furthermore, the randomization prevents the adversary from predicting which node will detect the replication. B. Security Analysis Let malicious node α maintain to be at L locations, l1,l2,…lL. We would like to decide the probability of a collision using the randomized multicast protocol outlined above, since a collision at a witness corresponds to discovery of α’s duplication. At each location li, p. d nodes randomly select g witnesses. If the neighbors synchronized perfectly, this would store α’s location claim at exactly p.d .g location. However since we prefer to have each neighbor act independently, there may be some amount of overlap between the witnesses each neighbor selects. To decide the impact of this partly cover, we would like to decide the number of nodes, Nreceive, that will receive the location maintain assuming the neighbors choose witnesses independently. If Pclaim is the likelihood that a node hears at smallest amount one claim and Pnone is the likelihood that a node hears no location claims, then we have: [

]= .

= 1− Since each neighbor is tacit to choose g random, unique witness locations, the probability (Pf ) that a node fails to attend to any of the g announcements from one neighbor is: =1− Since each neighbor decides autonomously whether to propel out location claims, the numeral of nodes that

ISSN: 2231-2803

actually launch out location claims is distributed binomially, with mean p.d and variance d · p(1 −p). For a network with d =20 and p = , the variance will be fewer than 0.005, so we will approximate the number of neighbors that send out locations claims as p.d Since the neighbors choose their destinations independently, we have: .

= 1− E[

.

]= . 1− 1−

The Binomial Theorem allows us to approximate (1−x) y as (1 − xy) for small x, so as long as g n, we have Nreceive≈ p·d·g, so over lapping witness locations should not impact the security of the protocol. As an example, in a network with n = 10, 000, g = 100, d = 20, and p = 0.1, perfect coordination would tell 200 nodes, while independent selection would tell 199. Thus, for the remainder of the analysis, we will assume that p · d · g nodes receive each location claim. Standard derivation of the birthday paradox, the probability Pnc1 that the ∙ ∙ recipients of claim l1 do not receive any of the ∙ ∙ copies of claim l2 is given by: . . . . = 1− Similarly, the probability Pnc2 that the ∙ ∙ recipients of claims l1 and l2 do not receive any of the ∙ ∙ copies of claim l3 is given by . . 2. . . = 1− =

1−

. .

. . .

The standard deviation that (1+x)