An emerging intelligent optimization algorithm based on trust sensing model for wireless sensor networks
- 351 Downloads
Due to the limitation of battery power, processing capacity, and storage, the sensor nodes are easy to be captured, destroyed, or attacked in an open environment. As a result, the security and reliability of data transmission cannot be guaranteed. In order to resist the internal attacks from malicious nodes, an ant colony optimization algorithm for secured routing based on trust sensing model (ACOSR) in wireless sensor networks is proposed. Firstly, a reliable evaluation model of trust perception is presented, which can estimate the node’s trust value derived from its behavior to identify or isolate the malicious nodes effectively. The penalty function and regulator function are applied to reflect the effect of state changes on the trust value according to the node’s behavior in the process of the communication. Secondly, the trust evaluation model is introduced into the ant colony routing algorithm to improve the security for data forwarding. The simulation results show that the proposed algorithm has improved the performance significantly in terms of packet loss rate, end to end delay, throughput, and energy consumption and demonstrates good resistance to black hole attack.
KeywordsEdge-of-things computing Wireless sensor networks Internet of things Energy consumption Emerging intelligent optimization algorithm
Ant colony optimization
Ant colony optimization algorithm for secured routing
Detecting dishonest recommendation
Intrusion-tolerant routing in wireless sensor networks
Quality of service
Security protocols for sensor networks
Secure and robust clustering
Three-tier security scheme
Wireless sensor network
The rapid developments of Internet of things (IoT) give impetus to the construction of cloud computing and constantly promote the popularization of applications for social network service and even the smart city . Depending on different types of distributed smart equipment, smart city can offer wide applications for urban residents in aspects of telemedicine diagnosis, intelligent community, environmental monitoring, and surveillance; owing to the characteristics of low cost, easy-to-deploy and specialized wireless sensor networks (WSNs) have played a crucial role in promoting the various facilities of the smart city. On the one hand, the ubiquitous sensor nodes can be used for collecting the physical information around the environment. On the other hand, they can make use of the widely distributed facilities to realize unattended management. However, the inherent characteristics of an open and distributed system, WSNs are of vulnerability due to the simplicity and limitation of design in the structure of hardware units . In addition, the communication of the wireless channel makes the attack more convenient. Since the traditional secure routing protocols are usually complex and energy-consuming, the existing protocol design is not suitable for resource-constrained WSNs .
At present, the researches on wireless sensor network security have made many achievements, and the security measures mainly include authentication, encryption, information integrity verification, and intrusion detection. Nevertheless, most of the above security mechanisms can only deal with the intrusion or attack from the network outside, and the malicious nodes will enter into the network smoothly and launch the specific attacks inside the network while the occurrence of the internal attacks of wireless sensor networks . The typical malicious attacks include selfish node, malicious forwarding, black hole, rushing, or worm attack [5, 6]. In view of the malicious node attack in WSNs, the trust model-based management mechanism is proven to be a popular and effective method . Besides, trust evaluation model has low computation and communication load, which demonstrate a significant advantage in respect of solving the internal attacks and identifying the malicious nodes .
The rest of this paper is organized as follows: In Section 2, we briefly introduce the related work. In Sections 3 and 4, the assumptions and explanations of the details of our method are described. The experiment methods are shown, and the result is discussed regarding the performance evaluation in Section 5. Finally, Section 6 concludes this paper and discusses the future work.
2 Related works
To protect the security of the link or constrain the malicious attack within a certain range, researchers have proposed different defense strategies, such as TTSS, SPINS, and INSENS [9, 10, 11]. Generally, those protocols usually adopt the technology of encryption, authentication, or key management mechanism to resist the attacks, including selective forwarding, SYBIL attack, wormhole attack, and HELLO flooding [12, 13].
Ganeriwal et al. [14, 15] presented a reputation-based trust management framework (RFSN) and applied to wireless sensor networks, which uses Watchdog system to monitor the neighbor node communication behavior. Besides, the trust value of sensor node is generated by putting the monitoring results into the credit system module, and reputation value is measured by Bayes formula for quantitative analysis of node’s uncertainty. The trust framework is complete and robust, but requires a priori distribution of the subjective assumption of reputation value. Shaikh et al.  presented a trust management scheme for hierarchical WSNs. The trust value is obtained by monitoring the communication behavior between the neighbor nodes by the predefined detective node. The trust mechanism can be established from several aspects to resist attack of malicious nodes, and it can consume less communication overhead as well as efficient memory usage. Feng et al.  put forward a credible trust management scheme for WSNs based on Bayesian theory. Firstly, the trust value can be estimated by RFSN model. Then, they utilized statistical data and Bayesian method to acquire the comprehensive trust value. Moreover, considering the influence of the uncertain factors of third parties on the value of trust in the process of evaluation, time sliding window is applied for trust value’s renovation. Jiang et al.  proposed a distributed trust evaluation model for WSNs, in which the direct trust value and the recommended trust value are selectively evaluated depending on the number of normal packets received by the sensor nodes.
To promote the accuracy of recommendation trust, multiple indicators, such as communication behavior, node’s residual energy, and the number of normal packets, are taken into consideration. Then, the model defines the trust reliability and familiarity to evaluate node’s trustworthiness. Hossein et al.  proposed a distributed trust management system with fuzzy theory to measure the trust value of nodes in WSNs. Yenumula et al.  examined the behavior of malicious nodes under the attack of selective forwarding and conducted the performance evaluation of the impact quantitatively. Due to the packets from normal nodes be selectively discarded, it will result in network failure or even collapse. To ensure the authenticity and reliability of the aggregated data, Hu et al.  proposed a security model based on trusted data fusion to resist the capacity of the risks derived from data tampering. Amol R et al.  presented a trust evaluation model for intrusion detection to identify malicious nodes or selfish nodes effectively, and the trusted routing is allowed by eliminating malicious nodes. Zhang et al.  proposed a dynamic trust establishment and management framework, and a trust varying function is defined to generate certain weight value for adapting to the dynamic changes of the network. Mejia et al.  proposed a trust evaluation model with game theoretic for online distributed evolution. Considering that the game theory is an auxiliary method for decision making rather than a prediction tool, it may not be suitable for resolving the trust problem of the sensor node. Bao et al.  presented a hierarchical trust management protocol by formulating social trust and QoS (quality of service) trust, which regards intimacy or honesty as a metric of trust evaluation and selects QoS’ trust by energy dissipation and node selfishness.
The existing trust evaluation methods are proposed for the characteristics of the node’s past behavior and different application scenarios. However, it is not enough to consider the trust assessment which combines subjective judgment and objective evaluation. In addition, the trust value is expressed by floating point number instead of integer value with single byte, which will result in excessive energy consumption between the nodes while passing the recommendation trust.
3 Trust sensing model
3.1 Direct trust value
Direct trust value is the immediate evaluation given by node i to node j through direct communication behavior, which can be obtained by Bayesian Statistical theory . In this way, the Watchdog mechanism will be applied to monitor the behavior of the neighbor nodes during interactive communication, and the monitoring results can be used to obtain the trust values of each node. In addition, the uncertainty of a random signal or event can be estimated by entropy theory. Based on information entropy theory, the uncertainty of a random signal or event can be measured.
With the increasing number of successful interactive communication between nodes, the value of the regulator function is more and more close to one with steady growth. It reflects the actual situation that the dynamic changes in the behavior of successful communication between nodes should be a long-term stable process, and the steady increase of node’s trust value can alleviate the possibility of conspiracy attacks on the system.
3.2 Indirect trust value
The indirect trust value refers to the evaluation of the node’s behavior from the recommendation nodes, which are composed of the neighbor nodes of node i and node j. It should be noted that not all recommended nodes are trustworthy, and the false recommendation from untrusted nodes will impair the quality of the true credibility of the node. Therefore, to obtain the right evaluation of the indirect trust value through the recommended nodes, it is necessary to select a set of credible nodes from common neighbors as the recommendation ones.
4 Secured routing method
Ant colony optimization algorithm (ACO) is proposed by Marco Dorigo, which is a heuristic intelligent algorithm for finding optimal paths in graphs [27, 28]. During the process, the collective behavior of ant colony shows a positive feedback of information, and finally, the whole ant colony will obtain the optimal path. It is very suitable for the characteristics of wireless sensor network routing.
To resist the possible attacks from malicious nodes, the node trust evaluation model is introduced into the ant colony algorithm, and the security and energy consumption of nodes should be regarded as the optimization objective. The main idea is to select the next hop neighbor node j of node i according to pheromone, sensor node’s residual energy and evaluated trust value. Usually, the selection of the node in the next hop should be satisfied with higher trust value and more residual energy.
Since the direct trust relationship between nodes is unable to involve all nodes, the estimation of the indirect trust value from the adjacent recommendation node. Besides, the smaller the weight value γ is, the more obvious the effect of the indirect trust demonstrates.
5 Results and discussion
The specific parameters
Initial trust value
Data flow rate
In this scenario, the trust value of the proposed algorithm is always lower than the trust value of DDR. It fully reflects that the trust value of ACOSR is more sensitive to malicious communication behavior and malicious attacks than DDR as the frequent occurrence of unsuccessful interactions. It also evaluates the trust relationship between nodes in ACOSR more accurately and reliably and identifies malicious nodes more quickly and effectively.
In order to resist internal attacks from malicious nodes, this paper proposes an ant colony optimization algorithm for secured routing based on trust sensing model in WSNs. Firstly, aiming at the problems of internal attacks such as black hole attacks, the ACOSR adopts the trust evaluation model to isolate malicious nodes effectively according to the behavior of nodes, which can reduce the packet loss rate and establish secure routing. In addition, by adopting the residual energy of the node as the key factor of the selection probability and taking into account of the node’s average energy when updating pheromone, it can effectively balance the energy expenditure among all nodes and reduce the average energy consumption of the whole network. In the future work, we will expand the model for node’s behavior evaluation under more complex attack mode as well as reducing the energy consumption of exchange of recommendation trust.
The authors thank the reviewers for giving the helpful comments and suggestions, which help to improve the quality of the manuscript.
This work was supported by the National Natural Science Foundation of China under grant no. 61603420.
YW and WS contributed to the conception and algorithm design of the study. YW and MZ contributed to the acquisition of simulation. YW, MZ, and WS contributed to the analysis of simulation data and approved the final manuscript.
Yongmei Wang received a M.S. degree from Anhui University, in 2011. She is an experimentalist in School of Computer Science and Technology, Hefei Normal University, Hefei, China. His research interests lie in computer application, intelligent arithmetic, and Internet of things.
Min Zhang received a M.S. degree from Anhui University, in 2011. She is an experimentalist in College of Information and Computer, Anhui Agricultural University, Hefei, China. His research interests lie in network of computer and Internet of things.
Wanneng Shu received a M.S. and Ph.D. degree from Central China Normal University and Wuhan University, in 2007 and 2013. He is an assistant professor in the College of Computer Science, South-Central University for Nationalities, Wuhan 430074, China. His research interests lie in Green cloud computing, parallel computing, and Internet of things.
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 1.MR Eslaminejad, M Sookhak, SA Razak, M Haghparast, A review of routing mechanisms in wireless sensor networks. Int. J Comput. Sci. Telecommun. 10(2), 1–9 (2011)Google Scholar
- 3.A Chakrabarti, V Parekh, A Ruia, A trust based routing scheme for wireless sensor networks. Lect Notes Inst Comput Sci Soc Inf Telecommun Eng 1(84), 159–168 (2012)Google Scholar
- 4.J Duan, D Yang, H Zhu, et al., TSRF: a trust-aware secure routing framework in wireless sensor networks. Int. J. Distrib. Sens Netw. 2014(1), 1–14 (2014)Google Scholar
- 5.F Ishmanov, AS Malik, SW Kim, et al., Trust management system in wireless sensor networks: design considerations and research challenges. Trans. EmergiTelecommun. Technol. 26(2), 107–130 (2015)Google Scholar
- 12.MS Syed, HB Safdar, et al., Neighbor node trust based intrusion detection system for WSN. Procedia. Comput. Sci. 8(3), 183 (2015)Google Scholar
- 14.L Lintao, H Lei, Z Na, Hierarchical routing trust model for wireless sensor networks. Comput. Eng. 23(6), 101–103 (2010)Google Scholar
- 15.S Ganeriwal, LK Balzano, MB Srivastava, Reputation-based framework for high integrity sensor networks. ACM. Trans. Sens. Netw. 4(3), 66–77 (2004)Google Scholar
- 19.J Hossein, RA Mohammad, et al., A fuzzy fully distributed trust management system in wireless sensor networks. Int. J. Electron. Commun.. 9(17), 1–10 (2015)Google Scholar
- 20.Y Reddy, J Durand, S Kafle, in Proc. of IEEE 7th International Conference on Information Technology, IEEE Press. Detection of packet dropping in wireless sensor networks (2010), pp. 879–884Google Scholar
- 21.HU Xiangdong, W Qinfang, T Hui, Model and simulation of creditability-based data aggregation for the internet of things. Chin. J. Sci. Instrum. 31(11), 2636–2640 (2010)Google Scholar
- 22.AR Dhakne, PN Chatur, TCNPR: trust calculation based on nodes properties and recommendations for intrusion detection in wireless sensor network. Int. J. Comput. Sci. Netw. Secur. 16(12), 1–10 (2016)Google Scholar
- 27.M Mendonca, JS Aguilar, N Perozo, An Emergent Ontology for Ambient Intelligence Based on an Ant Colony Optimization Algorithm, Computing Conference (CLEI), September 15–19, 2014 (IEEE Press, Montevideo, 2014), pp. 1–11Google Scholar
- 29.N Iltaf, A Ghafoor, U Zia, A mechanism for detecting dishonest recommendation in indirect trust computation. EURASIP J. Wirel. Commun. Netw. 2013(1), 1–13 (2013)Google Scholar
- 31.P Padh, RK Dash, K Martinez, A utility-based adaptive sensing and multihop communication protocol for wireless sensor networks. ACM. Trans. Sens. Netw 6(3), 27–39 (2010)Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.