IP-Based Credit Mobile Ad-hoc Networks

Ad-hoc wireless network is a collection of wireless mobile nodes forming a temporary networkwithout any centralized administration. Here, each node operates not only as a host but alsoas a router. Mobile ad-hoc networking has been an active research area for several years.However, the different routing problems and the cooperation among selfish mobile nodes arenot well addressed yet by the existing protocols. This paper proposes a new credit-basedsecure system for mobile ad-hoc network, where every node has a unique IP address. It alsoprovides incentives for mobile nodes to cooperate on the network services. Moreover comparedto previous approaches, this system does not require any tamper-poof hardware at any nodeto validate its authenticity.

Abstract

Vinu V Das, Member, IEEEDepartment of Computer Science and Engineering

MES College of Engineering, Kuttippuram, Kerala, IndiaEmail: prof.vinuvdas@gmail.com

1. Introduction

In recent years, mobile ad-hoc networks have received much attention due to their potentialapplications and the proliferation of mobile devices [5]. In order to make an mobile ad-hocnetwork functional, the nodes are assumed to follow a self-organizing protocol, and theintermediate nodes are expected to relay messages between two distant nodes. Recentevaluations have shown that ad-hoc networks are not only flexible and robust, but can alsohave good performance in terms of throughput, delay and power efficiency [6]. Mobiles nodes in a network are organized under a base station called cell[8]. Whenevera mobile node wants to communicate, it will reach the destination node through the basestation. Base station will route the packets to the destination and will be charged. Much ofthe bandwidth available for a base station could be used for some other services if thecommunication is done without routing through base station. This paper proposes a new restructured IP-based tree structure mobile ad-hoc networkto save the bandwidth of the base station. The nodes in the same level have the ability tocommunicate each other in a level and child nodes. Each mobile node will route and forwardthe packet independently to the destination node within the network. The mobile nodes in a network are under the authority of different users. So thecooperation behaviors such as forwarding each other’s messages cannot be directly assumed.There are two types of uncooperative nodes: faulty/malicious nodes and selfish nodes. Thefaulty/malicious nodes, refer to the broad class of nodes that are either faulty and thereforecannot follow a protocol, or are intentionally malicious and try to attack the system. Theproblems of faulty/malicious nodes need to be addressed from many layers, for example,using spread-spectrum encoding to avoid interference over the communication channel; usinga reputation system to identify the faulty/malicious nodes and subsequently avoid or penalizesuch nodes [7]; and applying the techniques from fault-tolerant computing to perform

International Conference on Computational Intelligence and Multimedia Applications 2007

0-7695-3050-8/07 $25.00 © 2007 IEEEDOI 10.1109/ICCIMA.2007.132

453

International Conference on Computational Intelligence and Multimedia Applications 2007

0-7695-3050-8/07 $25.00 © 2007 IEEEDOI 10.1109/ICCIMA.2007.132

453

computation correctly even in the presence of faulty/malicious nodes. Although the problemsof faulty/malicious nodes areq important in military applications, this paper focus on selfishnodes, which will be the dominant type of nodes in a civilian ad-hoc network. Specifically, aselfish node is an economically rational node whose objective is to maximize its own welfare,which is defined as the benefit of its actions minus the cost of its actions. Since forwarding amessage will incur a cost (of energy and other resources) to a node, a selfish node will needincentive in order to forward others’ messages.

2. Related work

One possibility to provide incentive is to use a reputation system [7, 1, 2]. For example,in [7], Marti et al. proposed a reputation system for ad hoc networks. In their system, a nodemonitors the transmission of a neighbor ensure that the neighbor forwards others’ traffic. Toidentify the uncooperative nodes they proposed two tools: a watchdog, which identifiesmisbehaving nodes, and a pathrater, which selects routes that avoid the identified nodes. In[1], [2], Buchegger and Le Boudec proposed and evaluated their CONFIDENT protocol,which detects and isolates misbehaving nodes. Another possibility to provide incentive is to use credit (or virtual currency). Buttyan andHubaux proposed a nice solution of this type in [3], and then presented an improved resultbased on credit counters in [4]. For both proposals, a node receives one unit of credit forforwarding a message of another node, and such credits are deducted from the sender (or thedestination). Besides other potential issues, both proposals require a tamper-proof hardwareat each node so that the correct amount of credit is added or deducted from the node. In [4],Buttyan and Hubaux proposed a stimulation approach that is based on a virtual currency,called nuglets, which are used as payments for packet forwarding. Using nuglets, the authorsproposed two payment models: the Packet Purse Model and the Packet Trade Model. The proposed credit-based system for mobile ad-hoc networks with selfish nodes, whichis similar to [3] and [4], uses credit to provide incentive to selfish nodes. However, one of thenovel and distinguishing features is that this system does not need any tamper-proof hardwareat any node. In connection oriented transmission, the sender establishes the connection byidentifying the route before starting the packet transfer. After receiving all the packets, thereceiver sends a summary report to the base station, where Credit Clearance Server (CCS)will credit the incentives to the nodes who have participated in the message transfer. This paper mainly deals with two systems: first, restructuring the present Mobile Ad-hocNetwork (MAN) to an IP-based tree structure and second, proposing a new credit-basedMobile Ad-hoc Network (MAN).

3. Restructuring the IP based MAN

The days are not so far, when every mobile node in the ad-hoc network will be identifiedby its IP address. Communication will then be much cheaper and faster. This is achieved byrestructuring the existing mobile ad-hoc network. The entire mobile ad-hoc network will actas interconnected local area networks having IP (preferably V 6) address. They can bestructured as a tree having root node as base station of the mobile network and all otherinternal as well as leaves nodes as mobile nodes. Each mobile node is under the control of the network and capable of communicatingindependently with other siblings as well as its child nodes as shown in the Figure 1. Everynode will be having separate routing tables of the neighboring nodes and child nodes. Theyare capable of contacting the neighboring nodes and child nodes without the help of basestation or the parent nodes. The base stations are under the single authority of service providers

454454

and if any node wants to communicate with the node of the other service provider, then itcould be through the service provider only. This helps the service provider to calculate/claimthe incentives from/ for the other service providers. Unlike the present communication system,only the inter-communication between the nodes of other base stations or service providersneed to use the bandwidth of the base station or service provider. That is the base station hasno role in the communication within the cell. Rest of the band width could be used for anyother useful services. Every node in the level has temporary IP address. The levels can be viewed as theindependent local area network with (may be) different class IP address, where each nodemay also act as a gateway. When ever a new mobile node is switched on, it will be recognizedby a base station where routine checking is done to analyse the communication software, andhardware tampering. This routine check up might be repeated at random period to findouttampered handset. The communication is checked accepting report from the sender andreceiver by sending junk packets via base station. Thus we can assume that all the mobilenodes are trust worthy for any communication. Then the IP address is allotted that is vacantfrom left to right and top to bottom and placed (or register) as a leaf node of some mobilebase parent node. When ever a mobile is switched off, it reports to the base station and freesthe IP address. If the node is not a leaf node, it needs to readjust the tree structure so as tomaintain uninterrupted communication. In this case the left most leaf node will be promotedto the parent node position which will now have its predecessors routing table and otherinformation. Gnerally, the left node is chosen because it is the node that is in service for thelongest period. Figure 2c shows the same.

4. Proposed credit based MAN

This section of the paper describes the different issues related to crediting the incentivesfor the services provided by any node. When a packet is forwarded by any selfish node it hasto be paid. A Credit Clearance Server (CCS) is attached along with each base station and theservice providers. Aove it the CCS of each base station will take care of the credit incentivesof the nodes within the cell, and the CCS of the service provider takes care of the incentivesfor the packet forwarded between the service providers. The system of the proposed creditarchitecture is shown in Figure 3. The traffic in the base station is redistributed to the mobile nodes under its control. Theselfish mobile nodes will forward the packets honestly if they provide adequate incentivesfor the resources (such as energy and handset usage) used by the base station. The honesty of the mobile node has to be traced properly to credit the incentives correctly.Before sending a packet to a destination node (under the connection oriented transmission) itwill locate the route to be taken to reach the destination. The sender will generate the request(Request_Open) to the base station for opening the specific network route from sender IDnumber (S_ID) to destination ID number (D_ID).

Request_Open(S_ID, D_ID) The base station finds out the shortest reliable route from S_ID to D_ID. Route is a finiteset of IP address that specifies the path to be taken b y the packets. This route informationthen passed on to the S_ID, D_ID and intermediate nodes.

Route_info(S_ID, D_ID, Route) Now all the routing nodes opens their port for the packet from S_ID to D_ID. The packetin the intermediate node will be forwarded to the prescribed route (by the base station) usingrouting table. The packets are forwarded as in any network protocols with IP address andother header information. Whenever an intermediate node forwards packets to another node,

455455

first node gets an acknowledgment from the second. If the first one does not receive anacknowledgement, then it has to resend the message to second node until it receives theacknowledgement. After trying for quite some time, if the first node is still not getting anacknowledgement for a packet, abort message has to be send to its sender recursively.

Abort(First_ID, Second_ID, Pckt_No) The sender will forward the error message to the base station for any corrective measures.Then the base station will try to find another shortest route from S-ID to D_ID. Eachintermediate node will attach its ID, while forwarding, to trace the route at the end oftransmission by the receiver. After the successful completion of the transmission of the packets,the receiver will send the summary report to the base station to credit the incentive to theintermediate route nodes.

Report(Route, No_Transmission) The proposed protocol for packet forwarding between nodes is explained with a typicalcase as shown in Figure 4. Consider a case where the node N1 wants to communicate with N7 and the link betweenN6 to N5 is somehow unreachable. N1 is able to locate the route (N2, N4, N6, N5) to sendfew packets to destination node N7. Once this connection is established by the sender N1, itwill send the route nodes to the base station to open a route. After opening a route, messagewill be forwarded to the destination and incentives will be credited to the nodes by the basestation upon receiving the summary report from the destination node.

Service Provider 1

Service Provider 2

Base Station 1

Base Station 2

Base Station m

Node Node Node Node Node Node

Node Node Node Node Node Node

Figure 1 – Restructuring the IP-based mobile ad-hoc network

Service Provider 1

Service Provider 2

Base Station 1

Base Station 2

Base Station m

Node Node Node Node Node Node

Node Node Node Node Node Node

Figure 1 – Restructuring the IP-based mobile ad-hoc network

Node Node Node

Node Node Node Node

Newly added node

Node Node Node

Node Node Node Node

Node to be removed

Node Node Node

Node Node Node Node

Node to be removed

Figure 2 a – Node ON Figure 2 b – Leaf node OFF Figure 2 c – Intermediate node OFF

Node Node Node

Node Node Node Node

Newly added node

Node Node Node

Node Node Node Node

Newly added node

Node Node Node

Node Node Node Node

Node to be removed

Node Node Node

Node Node Node Node

Node to be removed

Node Node Node

Node Node Node Node

Node to be removed

Node Node Node

Node Node Node Node

Node to be removed

Figure 2 a – Node ON Figure 2 b – Leaf node OFF Figure 2 c – Intermediate node OFF

456456

ServiceProvider 1

ServiceProvider 2

ServiceProvider n

CCS

Base Station 1

Base Station 2

CCS CCS Base Station m

CCS

Base Station

CCS

N1 N3 N5 N7

N2 N4 N6

Figure 3 – Credit based MAN Figure 4 – Typical case

ServiceProvider 1

ServiceProvider 2

ServiceProvider n

CCS

Base Station 1

Base Station 2

CCS CCS Base Station m

CCS

ServiceProvider 1

ServiceProvider 2

ServiceProvider n

CCS

Base Station 1

Base Station 2

CCSCCS CCSCCS Base Station m

CCSCCS

Base Station

CCS

N1 N3 N5 N7

N2 N4 N6

Base Station

CCSCCS

N1N1 N3N3 N5N5 N7N7

N2N2 N4N4 N6N6

Figure 3 – Credit based MAN Figure 4 – Typical case

This paper restricts itself to other security issues pertaining to communication. Anycryptographic algorithms can be used depending upon the available minimum resources inthe nodes.

5. Conclusions

In summary, the contributions of this paper are the following. First, the paper hasrestructured the mobile ad-hoc network system to free the bandwidth available for a basestation by communicating within the network. Second, the proposed system provides incentiveto selfish mobile nodes to cooperate, and motivates each node to report its actions honestly,even when a collection of the selfish nodes collude. Third, in this model the essentialcomponent of the credit-based system is the selection of the route by the sender throughwhich the message will be forwarded. This can be the first pure-software solution that hasformal proofs of credit security. The message-forwarding in unicast, can be extended toroute discovery as well as multicast.

6. References

[1] Sonja Buchegger and Jean-Yves Le Boudec, “Nodes bearing grudges: Towards routing security, fairness,and robustness in mobile ad hoc networks,” In 10th Euromicro Workshop on Parallel, Distributed and Network-based Processing, 2002.[2] Sonja Buchegger and Jean-Yves Le Boudec, “Performance analysis of the CONFIDANT protocol: Cooperationof nodes – fairness in dynamic ad-hoc networks,” In Proceedings of IEEE/ACM Workshop on Mobile Ad HocNetworking and Computing (Mobi-HOC). IEEE, June 2002.[3] L. Buttyan and J. P. Hubaux, “Enforcing service availability in mobile ad-hoc WANs,” In IEEE/ACM Workshopon Mobile Ad Hoc Networking and Computing (MobiHOC), Boston, MA, August 2000.[4] L. Buttyan and J. P. Hubaux, “Stimulating cooperation in self-organizing mobile ad hoc networks,” ACMJournal for Mobile Networks (MONET), special issue on Mobile Ad Hoc Networks, summer 2002.[5] Charles Perkins, Ad Hoc Networking, Addison-Wesley, 2000.[6] Hung-Yun Hsieh and Raghupathy Sivakumar, “Performance comparison of cellular and multi-hop wirelessnetworks: A quantitative study,” In Joint International Conference on Measurement and Modeling of ComputerSystems (SIGMETRICS) 2001, Cambridge, MA, June 2001.[7] Sergio Marti, Thomas Giuli, Kevin Lai, and Mary Baker, “Mitigating routing misbehavior in mobile ad hocnetworks,” In Proceedings of The Sixth International Conference on Mobile Computing and Networking 2000,Boston, MA, August 2000.[8] Andrew S Tanenbaum, Computer Network, Prentice-Hall India, 2001.[9] A. Shajin Nargunam and M.P Sebastian, “Fully Distributed Cluster Based Routing Architecture for MobileAd Hoc Networks,” IEEE Comm Mag, 39(1), April 2005, pp. 383-390

457457