IJETAE_1112_108

8/13/2019 IJETAE_1112_108

1/5

8/13/2019 IJETAE_1112_108

2/5

International Journal of Emerging Technology and Advanced Engineering

Website: www.ijetae.com (ISSN 2250-2459, Volume 2, Issue 11, November 2012)

689

A mobile node can access data items held by other

connected mobile hosts, and it is more possible to share

different kinds of replica among them.

Fig.1 SAF replica allocation

The SAF method causes low data accessibility when

many mobile hosts have the similar access characteristics

hence some of the data items to be duplicated in many

nodes.

B. The DAFN(Dynamic Access Frequency andNeighbourhood) Method

To overcome the problem of replica duplication in the

SAF method, a new method of replica allocation called

DAFN method [9] was developed. It eliminates thereplica duplication among neighbouring mobile hosts.

The algorithm of DAFN[2] method is as follows:

Each mobile host broadcasts its host id and accessfrequency information at relocation period.

Each mobile node allocates the replica according toSAF method.

If two mobile nodes having the same data item thenthe node having replica changes it to another replica

which having high access frequency.

Fig.2 [1] shows an example of executing the DAFN

method in the diagram given.

Fig.2 DAFN replica allocation

At each relocation period, the mobile nodes exchange

information about replicas allocated in the memory space.

So the overhead and the traffic are high compared with

the SAF method.

C. The DCG(Dynamic Connectivity and Grouping)Method

The DCG method [2][3] shares replica of data items in

many groups of mobile nodes than the DAFN methodthat shares replicas among neighbouring nodes.

The DCG method creates groups of mobile nodes that

are bi-connected components in an ad hoc network. In

spite of grouping mobile nodes as a bi-connected

component, the group is not divided even if one mobile

node is disconnected from the network.

The algorithm of DCG [3] method is as follows:

Each mobile node broadcast its host id andinformation about its access frequency with data

items to other nodes. By using the broadcasting information every node

identifies the bi-connected component nodes.

In each group, an access frequency of the group toeach data item is calculated by adding all the access

frequencies of mobile node in that group.

According to the access frequencies of the group,replicas of data items are allocated until memory of

all mobile nodes in the group becomes full.

After allocating replicas of all data items, if themobile nodes have any free space then replicas are

allocated according to their access frequencies until

the memory space is full.

It causes high traffic due to exchange of information

but it provide high data accessibility and stability over

nodes.

D. The SCF (Self Cantered Friendship) methodThe SCF method [1] consists of three parts: 1)

Detection of selfish nodes, 2) Construct the SCF-tree,and 3) Allocation of replica at a relocation period; each

node executes the following procedure:

Each host detects the selfish nodes depend on creditrisk scores.

CreditRisk(CR) = Expected risk/ Expected value(1)

Each node makes its own topology graph andconstructs its own SCF-tree by excluding selfish

nodes.

According to SCF-tree, each node allocates replicain a distributed manner.

The CR score [4] is updated during the query

processing phase. It considers the notion of credit risk

from economics which measure the degree of

selfishness. In economics, credit risk is the measured

risk of loss due to a debtors nonpayment of a loan. A

bank examines the credit risk of an applicant before

approving the loan.

8/13/2019 IJETAE_1112_108

3/5



690

The measured credit risk value (1) of the applicant

indicates if he/she is worthy or not. A host wants to know

if another node is believable; in that case a replica can be

served upon request to share a memory space in aMANET. With the measured degree of selfishness, a tree

that represents relationships among nodes in a MANET,

for replica allocation termed the SCF-tree. The SCF-tree

resembles human friendship management in the real

world. The key strength of the SCF-tree-based replica

allocation techniques is that it can reduce thecommunication cost, while achieving high data

accessibility.

Each node detects selfishness and makes replica

allocation at its own discretion, without forming any

group so it reduces overhead.

III.STRATEGIES FOR HANDLING SELFISH BEHAVIOURINNODES

A.Reputation based techniqueIn reputation based technique [5] a node receives one

unit of credit for forwarding a message of another node

and such credits are deducted from the sender or thedestination. In reputation based technique, a node

monitors the transmission of a neighbour to make sure

that the neighbour forwards others traffic. If the

neighbour does not forward others traffic, it is considered

as selfish node and this uncooperative reputation ispropagated throughout the network. Each node in thenetwork runs the Confidant protocol. It observes the

behaviour of neighbour nodes to detect misbehaviour

such as packet dropping. This requires nodes to run in

promiscuous mode. When the monitor finds

misbehaviour, it notifies the reputation system, which

manages a table containing nodes and their ratings. If thenumber of times a node misbehaves exceeds a threshold,

the reputation system updates the nodes rating. If a

nodes rating falls below a threshold, the system

considers it a malicious node. The reputation system

maintains a list containing the selfish nodes. When

forwarding packets, nodes avoid next nodes on the list.

When the reputation system detects a selfish node, it

notifies the trust manager to broadcast an alarm message

in the network. Trust managers also receive alarms from

other trust managers. The path manager ranks the path

according to the ratings of the nodes on the path. Itdeletes all paths containing malicious nodes and drops

route requests received from selfish nodes.

1.Watchdog mechanismThe watchdog [6] is one of the mechanisms which

detect selfish nodes by running a misbehaving node

locator on every host that maintains a buffer of recently

sent packets.

It overhearing packets transmitted and compares it

with the packets in the buffer to found if there is a match

between the packets sent. If the packet has been sent

from the buffer then watchdog removes the packet fromthe buffer. If there is any mismatch occurs and certain

packets occupy the buffer for more than particular time,

the watchdog increases a failure count for the node

responsible for forwarding the packet. If the count

exceeds a threshold value, the watchdog considers that

host as a misbehaving node.

2.Pathrater methodA pathrater [6] is a mechanism which maintains a

rating for every other host in the network. To choose a

route that is considered to be reliable, it calculates a pathmetric by averaging the rating of the nodes on the paths

and chooses the path with the highest metric.

If any node gets very low rating, it should be

considered as a selfish node and thus excludes them from

routing.

It increases throughput by 17% in a network with

moderate mobility and increases network throughput by

27%, with extreme mobility. Pathrater also having some

of the draw backs such as increases overhead in the

transmissions from 9% to 17% with moderate mobility.

Without using watchdog path rater is inefficient. Using

watch dog is necessary in all the detection systems.

B.Credit-payment techniqueAd hoc-VCG [7] is one of the reactive routing

protocols, which starts discovering routing paths when a

network node initiates a session. Ad hoc-VCG uses a

DSR like route discovery protocol that provides all

information about shortest paths to the destination node.

The destination node calculates the shortest path and theVCG payments and sends this information back to the

source. In the data transmission phase, the source sends

packets combined with electronic payments to the

destination along the shortest path. Ad hoc-VCG is

reliable against a single cheating node but it may fail in

the presence of coalitions of nodes (coalition forming)which try to maximize their total payments. It provides

truthfulness and assures cost efficiency but it having

some disadvantages such as excessive overhead and

Coalition-forming.

C.Game theory based techniqueSelfish nodes are sometimes called as freeloaders [8]

getting resources from the network and did not upload

any resources to the network. Minimising the effects of

freeloaders require the services of some external

centralized authority. The inclusion of third party

produces overhead in tracking, storing and processing the

behaviour of other nodes.

8/13/2019 IJETAE_1112_108

4/5



691

The algorithm called SLAC algorithm used to

eliminate the use of external third party and minimises

the overhead.

1.SLAC algorithmIt assumes that nodes want to use their abilities

selfishly to increase their own utility in a greedy way.

The algorithm [9] depends on Selfish Link and behaviour

Adaptation to produce Cooperation (SLAC).According to

the activities of each node SLAC generates some

measure of utility U (The number of files downloaded orjobs processed). Periodically each node (i) compares its

performance against another node (j), randomly selected

from the network. If the utility Ui

8/13/2019 IJETAE_1112_108

5/5



692

REF NO DETECTION

METHODS

ROUTING

OVERHAED

THROUGHPUT FALSE

POSITIVES

DETECTION

[6] Watchdog andPathrater

Low 75.2%-88.6% High 100% at 1 m/s80% at 15 m/s

REF NO SELFISHNESS HANDLING

METHODS

SCALABILITY NUMBER OF

HITS

NORMALIZED

ACCEPTANCE RATE

(NAR)

[9] SLAC High Scalability High Low NAR

[10][11] GTFT High Scalability High Low NAR

[12] TSLS Low Scalability Low High NAR

V. CONCLUSIONThe survey paper considers a selfish node detection

method and SCF tree replica allocation techniques tohandle the selfish replica allocation. It refer traditional

selfish node detection techniques such as watch dog and

pathrater and replica allocation techniques such as DCG,

SAF and DAFN are failed to consider selfish nodes in

terms of replica allocation .The new SCF tree allocationmethod shows that the considered techniques outperform

existing cooperative replica allocation techniques in

terms of data accessibility, communication cost, and

query delay.

Acknowledgement

I would like to thank reference authors and also like tothank the anonymous reviewers, whose comments and

suggestions have helped them to improve the quality of

the original manuscript.

REFERENCES

[1 ] Jae-Ho Choi, Kyu-Sun Shim, Handling Selfishness in ReplicaAllocation over a Mobile Ad Hoc Network SangKeun Lee, andKun-Lung Wu, Fellow, IEEE.2012.

[2 ] T. Hara, Effective Replica Allocation in Ad Hoc Networks forImproving Data Accessibility, Proc. IEEE INFOCOM, pp. 1568-1576, 2001.

[3 ] T. Hara and S.K. Madria, Data Replication for Improving DataAccessibility in Ad Hoc Networks, IEEE Trans. Mobile

Computing, vol. 5, no. 11, pp. 1515-1532, Nov. 2006.

[4 ] L.J. Mester, Whats the Point of Credit Scoring? Business Rev.,pp. 3-16, Sept. 1997.

[5 ] Y. Liu and Y. Yang, Reputation Propagation andAgreement inMobile Ad-Hoc Networks, Proc. IEEE Wireless Comm. And

Networking Conf., pp. 1510-1515,

[6 ] S. Marti, T. Giuli, K. Lai, and M. Baker, Mitigating RoutingMisbehavior in Mobile Ad hoc Networks, Proc. ACM MobiCom,pp. 255-265, 2000.

[7 ] L. Anderegg and S. Eidenbenz, Ad Hoc-VCG: A Truthful andCost-Efficient Routing Protocol for V. Srinivasan, P. Nuggehalli,C. Chiasserini, and R. Rao, Cooperation in Wireless Ad Hoc

Networks, Proc. IEEEINFOCOM, pp. 808-817, 2003.

[8 ] Hales, From Selfish Nodes to Cooperative Networks- EmergentLink-Based Incentives in Peer-to-Peer Networks, Proc. IEEE

Intl Conf. Peer-to-Peer Computing, pp. 151-158, 2004.

[9 ] S.U. Khan and I. Ahmad, A Pure Nash Equilibrium-Based GameTheoretical Method for Data Replication across Multiple

Servers, IEEE Trans. Knowledge and Data Eng., vol. 21, no. 4,pp. 537-553, Apr. 2009.

[10 ]M.J. Osborne, An Introduction to Game Theory. Oxford Univ.,2003.

[11 ]B.-G. Chun, K. Chaudhuri, H. Wee, M. Barreno, C.H.Papadimitriou, and J. Kubiatowicz, Selfish Caching inDistributed Systems: A Game-Theoretic Analysis, Proc.

[12 ]B.-G. Chun, K. Chaudhuri, H. Wee, M. Barreno, C.H.Papadimitriou, and J. Kubiatowicz, Selfish Caching in

Distributed Systems: A Game-Theoretic Analysis, Proc. ACM

Symp. Principles of Distributed Computing.

Documents

IJETAE_1112_108