Wireless Pers Commun DOI 10.

1007/s11277-012-0688-1

A Novel Heuristic Approach Based Trust Worthy Architecture for Wireless Sensor Networks
V. R. Sarma Dhulipala N. Karthik RM. Chandrasekaran

Springer Science+Business Media, LLC. 2012

Abstract A Wireless Sensor Networks (WSNs) consist of spatially distributed autonomous sensors to cooperatively monitor physical or environmental conditions, such as temperature, sound, vibration, pressure, motion or pollutants. WSNs are more vulnerable to attacks and failures due to the involvement of many numbers of tiny sensor nodes. As the technology is tremendously increasing in the recent past, the implementation of this for various time critical applications is quite interesting and challenging. Moreover, WSNs have no specic hierarchical structures, leads to security and maintenance problems. Trust in WSN is dened as the degree of belief or condence about the nodes based on the past interactions and observations has which become a mandatory requirement for reliable communication in WSN under security constraints. In this paper, we propose a Heuristic Approach based Trust Worthy Architecture for WSN that considers the challenges of the system and focus on the collaborative mechanism for trust evaluation and maintenance. Our proposed Architecture could also be capable of fullling critical security, reliability, mobility and performance requirements for reliable communication while being readily adaptable to different applications. The

V. R. S. Dhulipala (B ) Department of Physics and Research Member Centre for Convergence of Technologies, Anna University of Technology, Tiruchirappalli, Tamil Nadu, India e-mail: dvrsarma@gmail.com N. Karthik Department of Information Technology, Adiparashakthi College of Engineering, Kalavai, Tamil Nadu, India e-mail: nkarthikapce@gmail.com RM. Chandrasekaran Department of Computer Science and Engineering, Annamalai University, Chidambaram, Tamil Nadu, India e-mail: aurmc@hotmail.com

123

V. R. S. Dhulipala et al.

simulation results of the proposed architecture outperformed the recent trust worthy architecture using the analysis of the performance requirements such as communication overhead, memory requirements and energy consumption. Keywords WSN Trust Heuristic Security Reliability Mobility

1 Introduction Sensor nodes in WSNs are deployed randomly in the network environment to detect the events and send the information back to the base station or the cluster head. They are limited by resources, computation and communication capabilities for monitoring the physical and environmental conditions [1]. Trust evaluation and its management is a challenging issue in any networked environment and can solve problems such as communication overhead, memory requirements and energy consumption issues from the power of the old-fashioned cryptographic security techniques. Trust worthy architectures considers initiation of trust and its establishment as the primary building block for reliable communication in a wireless environment. Trust, in general is the level of assurance or condence in a node or network of nodes. It solves the problem of access control, selection of reliable routing scheme and anonymous path for routing packets and also ensures authentication, authorization and security mechanisms in WSN [2]. It also may further be given as the degree of belief or condence about the future behaviour of the entities, which is based on the past interactions. It becomes quite important in self-congurable and autonomous sensor networks and could also help its nodes to decide whether another member of the network is being non-cooperative or malicious. Nowadays, the trust management plays a vital role in every network based applications as security is highly needed when the sensor nodes are deployed randomly in hostile and military environment. In such environments, the sensed data from the sensor nodes need to be well protected or hided from the attackers/hackers to ensure the condentiality and integrity. 1.1 Significance of Trust in Ubiquitous Applications Sensor networks are anticipated to play a vital role in the fast growing era of ubiquitous computing. It is thus, more than obvious that such diversied environments such as those ranging from bank transactions to military applications necessitate the employment of security mechanisms at different levels [3]. Trust in ubiquitous application such as military environment, health monitoring and bank transaction applications is very important due to the untetheredness and condentiality. Figure 1 shows the importance to trust in different elds.

1.1.1 Military Applications Wireless sensor networks are the backbone of modern military based application for sharing information with a distrust source could lead a catastrophic condition of the soldiers. Military based communication networks are mainly adhoc in nature and trusted authentication and sharing is must.

123

A Novel Heuristic Approach Based Trust Worthy Architecture

Fig. 1 Trust in ubiquitous applications

1.1.2 Health Monitoring Applications Trust plays a major role in the health monitoring using WSN. In health monitoring applications physiological sensors are being attached to the patients body and the condential data from the Body Sensor Network (BSN) is being transmitted to a trusted monitoring source. Diagnosis process will be taken care based on the data received from the BSN.

123

V. R. S. Dhulipala et al.

1.1.3 Bank Applications In bank applications trust worthiness of bank transaction is essential in order to make a secure banking among the customers with the bank. The transactions from the customers are not always done from the trusted computer or trusted resource. Hence proper authentication and trust mechanism should be used by the customer. Based on the above application scenarios, in this paper, we propose a novel and Heuristic Approach based Trust Worthy Architecture (HATWA) for WSN. HATWA consist of following unique features: It evaluates the trust of the node and the trust of the group (cluster), to counterpart the existing approaches concentrating only on the trust evaluation of the single node. It works well for both clustered and layered architectures of WSN. Moreover the usage of this method reduces the communication overhead and cost associated with the trust evaluation of the node and to the group. It is the rst of its kind for trust evaluation on the node with mobility concerns. The rest of the paper is organized as follows: Sect. 2 reports various related works of trust management in WSN. Trust definition, properties, representation of trust values and assumptions for HATWA are explained in Sect. 3. The network monitoring node and HATWA model are described in the Sect. 4. Section 5 describes the theoretical requirements and simulated results of the proposed model followed by the comparison with recent model. Finally Sect. 6 concludes our work.

2 Related Works Quite a good number of trust models for trust management and trust based schemes are available in the literature [411]. But none of these models or techniques provided the exact requirement of trust management for WSN with mobility concerns. Only few trust management schemes such as GTMS [2], RFSN [12], ATRM [13], PLUS [14] are discussed in detail as far as the concept of the mobility is concerned. In Group based Trust Management Scheme (GTMS), the authors proposed a new light weight trust management scheme for WSN. It works with two different topologies: intragroup and intergroup, where distributed trust management and centralized trust management is adopted respectively. For the intragroup network, each node in the network calculates the individual trust values for all group nodes in the cluster. Based on the trust values, a node will be assigned any one of the possible states to the in the cluster. The possible states are represented as Trusted, Untrusted and Uncertain respectively. Based on the trust states of the node in the cluster, the cluster head will assign the overall trust state for the group and also detects the malicious and faulty nodes. The clustering scheme and group deployments enable the sensor node to work in a cooperative manner. So the selection of cluster head and the detection of faulty or malicious nodes is easy. The advantage of the scheme is that, it evaluates the trust for the group of nodes rather than a single node in the cluster. In Reputation based Framework for Sensor Networks (RFSN) [12], each node in the network maintains the reputation for neighbouring nodes. The values of the trust are evaluated on the basis of that reputation and Bayesian formulation was used for representing the values of reputation. According to this scheme, each node in the sensor network develops a

123

A Novel Heuristic Approach Based Trust Worthy Architecture

reputation for other node in the network by making direct observations. This reputation is used to calculate the trustworthiness of other nodes and make decisions within the network. It assumes that the node has enough interactions with the neighbour node so that the reputation can reach the stationary state. The significant factor of this model is that, higher the mobility of the node indicates that there is no enough interaction with the nodes and hence the reputation information will not be stabilized. An Agent Based Trust and Reputation Management scheme (ATRM) [13] introduced by the authors is based on a clustered WSN with the mobile agent system. This scheme requires every node to hold the mobile agent which is administrating the trust and reputation of the hosting node. The Trusted Authority (TA) is responsible for launching the mobile nodes on every node to administrate the trust and mobile agents are resilient against unauthorized analysis and modication. It calculates trust in a fully distributed manner. It works on specic agent based platform. The importance of the scheme is that it minimizes the communication overhead in terms of extra messages and time delay. Finally in Parameterized and Localized trUst management Scheme for sensor network security (PLUS) [14] is a localized distributed approach where highly abstracted parameters give the scheme exibility to adapt to different operational environments and different applications. The trust is calculated based on either direct or indirect observations. This scheme works on top of the own dened routing scheme called PLUR_R. It maintains a parameter database to describe the operational environments, application types, and network status and node information.

3 Trust Definition, Properties, Representation and Assumptions In this section we described the definition of trust and its properties and representations. We also have described the assumptions used for the implementation of the model. 3.1 Definition Trust in networks is the degree of belief or condence about the other nodes in the network based on the past interactions and observations [2]. Further with more clarity and randomness of node deployment, trust in WSN may be dened as, a combined characteristic model in WSN for providing reliability, security with respect to mobility. The significant properties of the trust are cooperativeness, subjectivity, semi-transitivity, temperance, dynamicity, and non-monotonicity which was considered in our heuristic algorithm for security, mobility and reliability and node calculations. 3.2 Representation The trust value of the node in the WSN can be represented as a continuous variable and its value range from 1 to +1 [15]. In general the trust value in the network can be represented as a numeric value. But in the proposed architecture (HATWA), we represented the trust value as a decimal value for security model, mobility model and reliability model and the overall trust of the node is represented as trust state which can be further understood as trusted, untrusted and uncertain states. The trust values and their respective trust states are shown in the Fig. 2.

123

V. R. S. Dhulipala et al.

Fig. 2 Trust values and states

3.3 Assumptions The sensor network consists of a number of tiny sensor nodes deployed randomly in the region for monitoring the physical and environmental conditions. For the the executions of the trust mechanism in WSN environment, we used the following assumptions: We assume that every node in the sensor network has unique ID and there exists a network monitoring node outside every cluster for storing the past interactions for trust evaluation of nodes in the network. Furthermore, the network monitoring node has no resource constraint problem and is free from the attacks. It stores all the interactions of the nodes in the cluster or group. It will also maintain the history of the node for a particular period of time. The memory requirements for maintaining the past history is set initially by simulation specications. This time varies from one the application to another. It uses a secure channel and routing protocol for communicating the trust value among the nodes in the network.

4 Evaluation of Trust Using a Network Monitoring Node The proposed trust model HATWA works in two stages for the trust evaluation in the network. 1. Trust calculation at the node level. 2. Trust calculation at the cluster/group level. 4.1 Trust Value Calculation of a Node With Network Monitoring Node Consider a network of 10 nodes in a clustered architecture. CH is the cluster head whose selection is made based upon the optimum energy level. All the nodes in the network are randomly distributed in the terrain. Each node is assigned a node ID and the network communication is managed by a network monitoring node for the issues such as security and mobility. At the node level, the trust value can be calculated by past interaction of the node which is stored in the network monitoring node. If the node 2 wants to have a trusted communication with node 5, it will send the ID of node 5 to cluster head. The cluster head will forward the node ID to network monitoring node for the evaluation of the trust value of the node. The network monitoring node follows the the proposed model and heuristic algorithm for estimating the trust value of the node. If the node does not have the past interactions, the trust can be calculated by peer recommendations. However, if there is no peer recommendation for the node, pre-trust the node is taken into consideration. The network monitoring node resides outside the perimeter of the cluster as shown in the Fig. 3, which has more energy

123

A Novel Heuristic Approach Based Trust Worthy Architecture

Fig. 3 Trust value calculation of the node using a network monitoring node

Fig. 4 Trust value calculation of the cluster using a network monitoring node

than the other nodes and cluster head as it does not involve in events and is meant exclusively for access and trust value calculation. 4.2 Trust Calculation for a Group/Cluster The process of trust calculation of the group or the cluster is shown in the Fig. 4. In order to calculate the trust for the group or cluster, the network monitoring node requests the nodes in the group or in clusters for their node ID. The network monitoring node will evaluate the trust state of every node based on the trust state of each node in the group/cluster, the trust state for the cluster can be assigned. If more than 80 % of the node are in trusted state, then the network monitoring node will assign the trusted state to the group/cluster and hence network nodes ensures reliable communication. 4.3 Network Monitoring Node Architecture for HATWA A network monitoring node in the proposed architecture is more significant. Its architecture involves four stages namely stage 1 through stage 4. These stages are crucial due to their importance in the selection of layer, routing protocol, mobility model and energy consumption calculations. As shown in Fig. 5 the four stages of HATWA are as follows: Stage 1 called as layer stage includes various layers such as Physical through application in which trust can be calculated based upon the fault tolerant management mechanism [16].

123

V. R. S. Dhulipala et al.

Fig. 5 Network monitoring node architecture for HATWA

Fig. 6 Process diagram for HATWA

Routing protocol of Stage 2 is meant for selection of secure routing protocol which is considered from the existing models. Stage 3 is rst of its kind where trust calculation is performed with mobility concerns. Finally stage 4 is called as energy/ power management which is included for power and energy requirements for HATWA. 4.4 Proposed Architecture: HATWA and Its Mechanism As explained in the previous section, HATWA architecture involves three heuristic algorithms such as security, mobility and reliability which involve the functioning of network monitoring node architecture. These models are introduced to redesign the issues and challenges in the trust management for WSN. It considers the predened parameters and attributes taken into account to provide the trustworthiness of the network. Further, HATWA for WSN in terms of process diagram comprises of security, reliability and mobility models as in [15]. Initial trust value can be calculated by the direct observation and recommendations. The initial trust value is given as input to the HATWA model. The steps involved in HATWA and output of overall trust value are shown in Fig. 6. The trust value in the network can be represented as a continuous variable runs over a specic range of 1 to 1 [17]. The individual models of HATWA process diagram are explained as follows:

123

A Novel Heuristic Approach Based Trust Worthy Architecture Algorithm 1 Algorithm for security model Input: node ID, Signal strength. Output: Trusted, Secure communication using HATWA Initial Condition: Select CH, V. N entering into terrain. Begin: CH checks authentication. V.N Exchange key with CH if(V.N = = authorized)then allow communication else access abnegated exit endif. Select secure routing protocol. Encrypt the data using encryption algorithm and route the packets. if(V.N= =mobile) then select mobility model else V.N= static Calculate hop count. endif. End Notations: TWS-Trust worthy System, CH- Cluster Head, V.N-Visitor Node.

4.4.1 Security Model Security is a difcult task to be achieved in sensor networks. But it is essential for the success of WSN applications, particularly in unattended environments. Providing satisfactory security protection in WSNs is a challenging task due to various network constraints and malicious attacks. In security model, we addressed the node authentication to provide the access control. Encryption of packets covers the data security in sensor nodes. The secure and anonymous routing provides communication security in the network. The trust value of the security model of the node is high provided the node has secure routing protocol, access control and encryption of the routed packets. The following algorithm was proved good in our simulation experiments for the high trust value in the security model of the node in the network. The security model is executed with the heuristic algorithm shown in Algorithm 1. 4.4.2 Mobility Model Mobility is dened as movement of nodes in WSN. The mobility model describes the movement pattern and the location of the node. In mobility model, the localization issue helps us to locate the nodes in the terrain. We can estimate the mobility of the node based on the destination and source node location. Mobility [18] helps to propagate the trust between the nodes. The trust aggregation can be improved by the mobility of the node. The mobility of nodes affects the trust measurement accuracy in the sense that, the more sources of information a node has in order to gather recommendations from implies that the more accuracy in its perception about the trust level of the other nodes. So, if a node moves to a region where there are a few reachable nodes, it will have too little information to properly build an accurate trust score for other nodes. Various mobility models in WSN are Random Waypoint model, Reference Point Group Mobility, Freeway (FW) Mobility Model and Manhattan Mobility (MH) Model. Secured mobility model and optimum energy consumption ensure trustworthiness in the ad hoc network environment. The trust value in the mobility model of the

123

V. R. S. Dhulipala et al. Algorithm 2 Algorithm for mobility model Input: V.N Output: hop count, spot of nodes, energy consumption. Initial condition: authorized V.N undergoing secure communication. Begin: If (V.N= =mobile) Select mobility model, Estimate mobility: calculate hop count. Estimate localization: calculate spot of nodes and energy consumption. Else (V.N= static) Find the spot of the node. endif

Algorithm 3 Algorithm for reliability model Input: V.N Output: reliable communication, latency, acquisition delay, energy consumption. Initial condition: CH permits V.N, Authorized V.N undergoing communication Begin: If (V.N undergone communication) then Calculate latency, bandwidth, and energy consumption endif. If (V.N receives packets) then Calculate acquisition delaypacket loss, energy consumption during reception of the packets. endif. If (V.N sends packets) then Calculate energy consumption during the transmission of the packets. endif.

node is calculated based on the optimum energy consumption during mobility of the node and the selection of the mobility models is based on the specic application requirement. The frequent change of velocity for the mobile nodes makes the intruders hard to trace its region. Algorithm 2 guarantees the ideal trust value for the node if it has optimum energy consumption checked for its correctness during the simulation experiment.

4.4.3 Reliability Model Reliable communication is one of the key aspects in WSNs. Reliability model must provide reliable or error free delivery of packets to ensure trustworthiness for communication mechanism. In reliability model, the process of data fusion, negligible packet loss, less delay and the optimum energy consumption during transmission and reception of packets leads to reliable communication. The following algorithm ensures enough trust value for the node in the reliability model, if it has less acquisition delay, latency and energy consumption, as proved by the simulation experiments. Algorithm 3 shows the reliability model algorithm.

123

A Novel Heuristic Approach Based Trust Worthy Architecture

4.4.4 Node Level Trust Calculation The node level trust calculation is always the prime factors for reliable communication with WSN environment. Algorithm 4 [19] explains in detail about the node level trust for WSN applications. If the node receives the communication request, the node evaluates the requested node trust value. The mobility of the node is considered in the trust value calculation. Because the mobility of the node [20] may weaken the trust value as it moves away from the terrain. The initial trust value of the node depends on the recommendations by the neighbours. Usually the value will be in negative or zero because the neighbours having negative recommendations or no recommendations since the node is mobile in nature. For example, the trust value in the security model is calculated as follows: If the node has access control, securely routed the encrypted packets in the history, then the trust value will be 0.33 in the security model. In reliability model, the trust value will be 0.33 if the node has a less acquisition delay, less packet loss, minimum latency and optimum energy consumption during data fusion. In mobility model, the trust value of the node will be 0.33, if the node has minimum energy consumption during mobility, the velocity of the node will be unknown to the attackers so it is difcult to nd the node in the terrain. If the trust value of the node is above 0.50 then the node is considered as trusted and it starts communication. 5 Results and Discussions This section presents the experimental setup, analyses of significant requirements such as memory consumption, communication overhead and energy consumption and their comparison with more recent model GTMS. The simulation specications and analyses are performed taking heuristic models of security, mobility and reliability into consideration. 5.1 Simulation Specications We veried the proposed models on mobile nodes, that is, nodes that do change its position according to the application requirements in a given terrain. Network Simulator 2 (NS2) has been used for validating the proposed models of HATWA in terms of trust metrics. We evaluated the performance of HATWA in three phases key issues such as memory consumption, communication overhead and energy consumption. During the simulation experiment, we have used rectangular, at terrain dimension of 1, 000 1, 000 m2 area in which the sensor nodes are randomly distributed (Table 1). 5.2 Memory Consumption Analysis for Trust Value Calculation The sensor node is comprised of a sensor, processing element which is the combination of the memory and processor and communication stages. The main stages of the sensor node that plays a vital role in communication is processed. Processing stage involves a memory element whose specications are a deciding factor for the reliable communication. This is due the allocation of the memory of the sensor node for various activities such as the capability of storing trust values of the mobile nodes, static nodes and for memorizing the peer recommendations. The typical MICA sensor node has 128 kbyte of program ash memory, 4 kbytes of EEPROM and 512 bytes of measurement ash [21]. The networking monitoring

123

V. R. S. Dhulipala et al. Algorithm 4 Algorithm for node level trust calculation Algorithm: For Calculating the Trust of the Node. Initial condition: Node wants to communicate with other node in the network. Input: Node id, Security, Reliability, Mobility Model of the node. Output: Trust value calculation and communication. Begin: Initial trust calculation of the node: Tinitial=((S+U)/(Ti+S)) or Pr; If (Tinitial is adequate for communication) Allow communication with the node. Else calculate the trust for the security model for the node. endif. Ts=A+E+R; If (Ts is adequate for communication) Then allow communication with the node. Else calculate the trust value for the mobility model for the node. endif. If node is static Then assume the trust value of the node in the mobility model is zero. Else calculate the trust value of the node is in mobility. Tm=Me+Em; endif. If (Tm is adequate for the communication) Then allow the communication with node. Else calculate the trust for the node in the reliability model. Tr=D+Ed; endif. If (Tr is adequate for communication) Then allow communication; Else calculate the overall trust for the node. Overall trust=Tinitial+Ts+Tm+Tr; endif. If Overall trust is adequate for communication Then allow the communication with the node. Else Deny communication with the node. endif. End. Notations: Tinitial=Initial Trust value calculation. S=Successful interaction. U=Unsuccessful interaction. Ti=Total Interactions Pr=Peer Recommendations. Ts=Trust value of the node in the security model. A=Access control. E=Encryption of the packets. R=Secure routing of the packets. Tm=Trust value of the node in the mobility model Me=Mobility estimation. Em=Energy consumption during the mobility model of the node.

123

A Novel Heuristic Approach Based Trust Worthy Architecture Algorithm 4 continued Tr=Trust value of the node in reliability model. D=Data fusion. Ed=Energy consumption during the data fusion.

node in HATWA model maintains a database past interactions of each node for a particular time. Here, we consider the security, mobility and reliability models of the node for storing the past experience of the nodes communication. The memory requirement for a mobile node, the memory requirement for the storing the trust value for static node and the memory requirement for storing the peer recommendations are shown in Tables 2, 3 and 4, respectively. 5.3 Communication Overhead Analysis When a node wants to interact with any neighbouring node, it needs to have a trust value of the node for secure and reliable communication. A total of three packets is needed for the calculation of the trust value of the single node in the cluster. If a node wants to interact with n other nodes in the network, the communication overhead will be 3 (no-1) packets, i.e. The network monitoring node needs two packets for the calculation of the trust value of a node in the network. If there are n number of nodes requires to know the trust value, the communication overhead at network monitoring node will be 2n. 5.4 Energy Consumption Analysis According to the memory requirement analysis, the initial battery level of the typical node will be 1106 J as in [2]. For transmitting the packet, a node consumes 1.6 W and the node takes 1.2 W for reception of the packet. In idle state, the node gives away 1.15 W. For the calculation of the trust value, a node needs three packets and it consumes 8.4 W. 5.5 Comparison of HATWA with GTMS For the purpose of comparison, we have calculated the theoretical results and simulated with the NS2. During simulation, we considered that each cluster in the network will have an equal
Table 1 Simulation specications Simulation tool Terrain dimension Simulation time Node security Routing protocol Reliability Mobility Network Simulator 2 (NS2) 1, 000 m 1, 000 m2 500 ms Encryption, authentication SAODV Packet loss, latency, energy consumption Spot of the nodes, energy consumption

Table 2 Memory requirement for storing the trust values of mobile nodes

Node ID 2 Bytes

Security model 2 Bytes

Reliability model 2 Bytes

Total trust value 2 Bytes

123

V. R. S. Dhulipala et al. Table 3 Memory requirement for storing the trust values of static nodes Node ID Security model 2 Bytes Reliability model 2 Bytes Mobility model 2 Bytes Total trust value 2 Bytes

2 Bytes

Table 4 Memory requirement for storing the peer recommendations

Node ID 2 Bytes

Peer recommendations 2 Bytes

Fig. 7 Communication overhead

number of nodes. We simulated individual model of HATWA and implemented the proposed model for static nodes. Further results are compared only with GTMS. Because the other existing models like ATRM, PLUS needs special platform and routing scheme respectively. We simulated the experiment with a simulation time of 500 ms for 5 clusters, each of the clusters has ten nodes. The theoretical and simulation of HATWA is compared with GTMS and are depicted in Figs. 7, 8 and 9, respectively for communication overhead, memory requirements and energy consumption. Results showed that HATWA consumes lesser-energy, less communication overhead and less memory requirement for the storing and keep track of communication history by the network monitoring node for the evaluation of node trust.

6 Conclusion and Future Work In this paper, we have introduced the architecture for trust management with heuristic algorithms to the key issues in WSNs such as security, mobility and reliability. The proposed architecture of the trust management in sensor networks may be integrated into many real time applications. The design of the trust management in WSN must be taken into account with diverse needs including reliability and security. The theoretical and simulated results

123

A Novel Heuristic Approach Based Trust Worthy Architecture

Fig. 8 memory requirement

Fig. 9 Energy consumption

of HATWA outplays the existing trust management techniques for WSN. In the future, we are planning to deal with more attributes such as accuracy, scalability, fault tolerance of the network and likely to have more algorithms for the key issues.

References
1. Akyildiz, I. F., Su, W., Sankarasubrarnanian, Y., & Cayirci, E. (2002). Wireless sensor networks: A survey. Computer Networks, Elsevier Science, 38(4), 393422. 2. Riaz Ahmed, S., Jameel, H., dAuriol, B. J., Sungyoung Lee, H., & Song, Y.-J. (2009). Group-based trust management scheme for clustered wireless sensor networks. IEEE Transactions on Parallel and Distributed Systems, 20(11). 3. Prasad, N. R., & Alam, M. (2006). Security framework for wireless sensor networks. Wireless Personal Communications, 37(34), 455469.

123

V. R. S. Dhulipala et al. 4. Ming-Zheng, Z., Zhang, Y., Wang J., & Zhao, S.-Y. A Reputation Model Based on Behavior Trust in Wireless Sensor Networks. In Eighth IEEE international conference on scalable computing and communications. 5. Tae Kyung, K., & Suk Seo, H. (2008). A Trust Model using Fuzzy Logic in Wireless Sensor Network. In World Academy of Science and Engineering and Technology (p. 42). 6. Lei, X., Xu, Y., Li, X., & Zhu, Y. (2006). A lightweight scheme for trust relationship establishment in ubiquitous sensor networks. In Proceedings of the sixth international conference on computer and information technology (CIT 06). 7. Zhang, W., Das, S., & Liu, Y. (2006). A trust based framework for secure data aggregation on wireless sensor networks. In The proceedings of the IEEE SECON, September 2006. 8. Efthimia, A., & Gritzalis, S. (2009). Trust-based data disclosure in sensor networks. In Proceedings of IEEE ICC 2009. 9. Aivaloglou, E., Gritzalis, S., & Skianis, C. (2007). Towards a exible trust establishment framework for sensor networks. Telecommunication Systems, 35(3), 207213. 10. Fei, S., & Zhao, B. (2008). Trust-based LEACH protocol for wireless sensor networks. In Second international conference on future generation communication and networking, FGCN 08, 2008. 11. Haiguang, C. (2009). Task-based trust management for wireless sensor networks. International Journal of Security and Its Applications, 3(2), April, 2009. 12. Ganeriwal, S., & Srivastava, M. B. (2004). Reputation-Based Framework for High Integrity Sensor Networks. In Proceedings of ACM workshop security of ad hoc and sensor networks (SASN 04), pp. 6667, Oct. 2004. 13. Boukerche, A., Li, X., & EL-Khatib, K. (2007). Trust-based security for wireless ad hoc and sensor networks. Computer Communications, 30, 24132427. 14. Yao, Z., Kim, D., & Doh, Y. (2006). PLUS: Parameterized and localized trust management scheme for sensor networks security. In Proceedings of third IEEE international conference on mobile ad-hoc and sensor systems (MASS 06), pp. 437446. 15. Sarma Dhulipala, V. R., Vishnu Prabha, B., & Chandrasekaran, R. M. (2010). Trust Worthy Architecture for Mobile Adhoc Networks. CCIS 70 (pp. 557560). Springer, Berlin. 16. Paradis, L., & Han, Q. (2007). A survey of fault management in wireless sensor networks. Journal of Network and Systems Management, 15(2), 171190. 17. Marsh, S. (1994). Formalising trust as a computational concept. In Department of Computer Science and Mathematics, University of Stirling, pp. 184. 18. Kannan, G., Mohapatra, P., & Abdelzaher, T. F. (2010). Trustworthy wireless networks: issues and applications. Invited paper in International Symposium on Electronic System Design, 2010. 19. Karthik, N. & Sarma Dhulipala, V. R. (2011). Trust value calculation in wireless sensor networks. In Proceedings of the IEEE international conference on electronics computer technology (Vol. 4, pp. 376380). 20. Gomez Marmol, F. (2010). Trust and Reputation Management in Distributed and Heterogeneous Systems, Ph.D Thesis, University of Murcia, April, 2010. 21. Xbow, MICA2 series, http://www.xbow.com.

Author Biographies
V. R. Sarma Dhulipala Assistant Professor in the department of Physics and Research Member in Center for Convergence of Technologies (CCT) is presently working for his Doctoral degree in the area of Wireless Sensor Networks. He completed his M.Phil. from Jawaharlal Nehru Technological University (JNTU), Hyderabad in the year 2006 and M.Sc. (Electronics) from Andhra University, Waltair, Visakhapatnam in the year 2000. His expertises in electronics lead him to teach various subjects in electronics, Pervasive Computing areas. He is an associated faculty of the departments of Nano Science and Technology (NST) and Pervasive computing Technology (PCT). His areas of research interest include Wireless and mobile ad hoc networks, Nano electronics simulation and modeling. He published 20 research articles in the international journals, 6 book chapters and 30 papers in national and international conferences. So far, he guided 40 M.E/M.Tech/M.C.A projects. He is a reviewer for referred journals like Journal of Systems and Software by Elsevier and Ad hoc Sensor Wireless Networks,

123

A Novel Heuristic Approach Based Trust Worthy Architecture Springer Journal of Wireless Personal Communication and International Journal of Ad hoc Sensor Wireless Networks. He is having a teaching experience over 11 years.

N. Karthik is currently working as Lecturer in Department of Information Technology, Adhiparasakthi College of Engineering, Kalavai, Tamilnadu, India. He completed his Graduation in Computer science and Engineering, from Anna University, Chennai and Post Graduation in Pervasive Computing Technology from Anna University of Technology, Tiruchirappalli. His research interest includes Pervasive Computing, Wearable Computing, and Embedded Systems.

RM. Chandrasekaran received the B.E. Degree in Electrical and Electronics Engineering from Madurai Kamaraj University in 1982 and the MBA (Systems) in 1995 from Annamalai University, M.E. in Computer Science and Engineering from Anna University and Ph.D. Degree in Computer Science and Engineering from Annamalai University, Tamil Nadu, India in 1998 and 2006 respectively. He is currently working as a Professor at the Department of Computer Science and Engineering, Annamalai University, Annamalai Nagar, Tamil Nadu, India. From 1999 to 2001 he worked as a software consultant in Etiam, Inc, California, USA. He has conducted Workshops and Conferences in the Areas of Multimedia, Business Intelligence and Analysis of algorithms, Data Mining. He has presented and published more than 50 papers in conferences and journals and is the author of the book Numerical Methods with C++ Program (PHI, 2005). His Research interests include Data Mining, Algorithms, Networks, Software Engineering, Network Security, and Text Mining. He is Life member of Computer Society of India, Indian Society for Technical Education, Institute of Engineers and Indian Science Congress Association.

123