Design of a novel dynamic trust model for spectrum management in WRANs of TV white space

https://doi.org/10.1016/j.jnca.2017.09.007Get rights and content

Abstract

The cognitive radio (CR) technique is considered as a better and more efficient answer to the predicament brought by the high demands of wireless communication services and the scarcity or low utilization of frequency spectrum resources. The IEEE 802.22 standard has regulated the wireless regional area networks (WRANs) as one example of the CR networks, which is operating in the television white space (TVWS) spectrum. In the centralized WRANs, the secondary user base station (SUBS) allocates the spectrum holes for the customer premise equipment (CPE) on the basis of the spectrum sensing information from the CPEs around. However, since the CPEs are easy compromised under the varying and uncontrollable environment, the WRANs is subjected to the primary user emulation attacks (PUEA) and the spectrum sensing data falsification (SSDF) attacks, which may lead to a wrong global spectrum decision made by SUBS. A novel dynamic trust model for the SUBS spectrum management based on the subjective logic for the centralized WRANs of TVWS is proposed in this paper. The proposed trust model has a positive impact on reducing the influence of users’ vicious behaviors and encourages the CPEs to participate in truthful actions for communications. The analysis shows that the time complexity of the proposed trust model is O(n). Simulations indicate that the proposed trust model can significantly reduce the probability of the errors by the SUBS in the spectrum decision making.

Introduction

Due to the rapid growth of the multifarious wireless communication services, the conflict between the scarcity of spectrum resources and the low utilization ratio of licensed frequency spectrum is turning to be much sharper (McHenry, 2005). Large portions of assigned spectrum resources have been underused geographically and contingently. Efficient use of those underused bands will enhance the utilization and alleviate the scarcity of the spectrum resources. In order to settle the predicament between the growing requirements on spectrum and the practical underused spectrum resources, the Federal Communications Commission (FCC) (Federal Communications Commission (FCC), 2012) has issued a rule that the unlicensed users can opportunistically use the contingently TV white space (TVWS) without interfering the incumbent users. The opening use of the TVWS to the new unlicensed users activated by the dynamic spectrum access technology will bring certain social and economic profits through capacitating the best use of unused or underused frequency spectrum. By the new cognitive communication epitome, unlicensed users, also known as CPEs in WRANs, can fill the spectrum holes opportunistically under the condition of non-interfering to the licensed users, also called as primary users, PUs. The CPEs need to sense the spectrum holes continuously to avert disturbing the PUs. When the PUs reclaim the spectrum, the CPEs are required to vacate the corresponding spectrum bands.

The first worldwide achievement to specify a standardized air interface for WRANs to share the frequency spectrum resources in TVWS by applying the CR techniques is the IEEE 802.22 standard (IEEE Std. 802. 22, 2011). The IEEE 802.22 WRANs are devised to engage in the TV broadcast bands between 54 and 862 MHz without interference to the communication of the incumbent users. The WRANs employ an orthogonal frequency division multiple access technique at the physical layer and form a centralized, connection orientated network. At the medium access control (MAC) layer, the SUBS manages the spectrum resource allocations within its cell, which allows user terminals located as far as 100 km with a proper scheduling of the traffic when exceptional RF signal propagation conditions appear. The physical and MAC layers consist of new CR functions to guard the PUs and acquire businesslike spectrum utilization. The architecture of a point-to-multipoint WRAN includes a fixed SUBS and multiple fixed or portable CPEs. The SUBS should be able to supply high-speed Internet service for up to 512 fixed or portable CPEs within its coverage area while meeting the regulatory requirements for protection of the PUs. The exploitation of the TVWS in wireless networks attracts some researchers recently (Angela and Marcello, 2016).

One of the dominant issues to deploy the WRANs in TVWS similar to other wireless communication networks is the security issue. Multifarious security and trust mechanisms, e.g., authentication, confidentiality, integrity and cryptography, have been proposed to prevent security menaces such as message replay, eavesdropping and remanufacture of messages. Nevertheless, the security mechanisms are subjected to plentiful security attacks, such as denial-of-service attacks and node capture attacks. The conventional security models can stand up to external attacks, but cannot guard against internal attacks efficaciously which are aroused by the spy users or compromised users. In addition, the WRANs are faced not merely with the conventional security menaces in wireless communication networks but also new challenges involved in the functions of cognitive radio techniques. In the WRANs, the CPEs with cognitive features account for sensing the real-time status of the communication channels including spectrum holes, geo-location, getable services etc. and informing the relevant spectrum information to the SUBS. The unique cognitive characteristics of the WRANs therefore bring more challenges to the security functionality. To set up safe communications, it should be certain that all communication participants are credible. However, as a result of the changeable environment nearby the CPEs, the existence of vicious cyber attackers and the open nature of the WRANs, the WRANs appear all kinds of security indeterminacy to the trustworthiness of data communications. The CPEs not merely set about the work of sensing spectrum holes, but also act as the end users of frequency spectrum resources in the cognitive cycles. For sake of removing the negative influence of malicious attacks from inner nodes breaking the FCC's rules, it's required to manage the CPEs with the trusts.

In the research works of wireless networks (Xiong and Liu, 2004, Govindan and Mohapatra, 2012a), the trust of a wireless node is an important aspect of security functionality. The idea to employ a trust management model in CRNs has also appealed to the researchers’ interests (Guo et al., 2011), in which a multicast trust model with optimizing energy use is proposed but with polynomial time complexity. Chen et al. have named 3 instances to explicate various requirements for security and trust in CRNs (Chen et al., 2008) but without introducing the mathematical efforts to facilitate trust and its applications to trusted routing of CRN as the core feature. Researchers have incorporated the trust and reputation to alleviate the menace of the spectrum sensing data falsification (SSDF) attacks on CRNs (Feng et al., 2016, Tevfik and Huseyin, 2009, Talay and Altilar, 2013, Thilina et al., 2013), in SSDF the secondary users intentionally manipulate the sensing reports, thus leading the data fusion algorithms to make the wrong decision. However, the data fusion algorithms of (Feng et al., 2016, Tevfik and Huseyin, 2009, Talay and Altilar, 2013, Thilina et al., 2013) do not consider the trust and reputation. A dependable trust based authentication method for CRNs is proposed by Parvin S. et al. Parvin et al. (2010a), where the trust value of a node is derived from all kinds of the communication attributes. It is determined by whether the node can join in the CRNs relying on the trust value. Parvin S. et al. has also integrated the certificate with the behavior into a trust model which will profit by expressing certificates as trust, in the meantime by supplying a behavior-based valuation of trust (Parvin et al., 2011). But the paper (Parvin et al., 2010a, Parvin et al., 2011) do not perform the simulation depending on the mathematical terms. Parvin S. et al. even has put forward a trust based model to authenticate the SUs in CRNs and a trust throughput based mechanism to evaluate the throughput in CRNs (Parvin et al., 2010b), However, (Parvin et al., 2010b) do not deal with the biasing between the Certificate Authority and other nodes, so the some specific node's trust value will always be higher. Parvin S. et al. has proposed a contiguous trust appraisal way in a CRN to settle the security threats caused by unreliable users to guarantee safe spectrum sharing in CRNs (Parvin et al., 2013), which however violates the FCC rules that primary network must not be changed. Pei Q.Q. et al. have proposed a trust frame for centralized CRNs (Pei et al., 2011) and an adaptive trust management model based on cognitive cycle (Pei et al., 2012), but with the limitation of the precondition of adaptability which is that the surrounding environment changes slowly. Pei Q.Q. et al. have also proposed a newfangled trust management model (SRTrust) for CRNs by which the trust value of SUs in sensing and allocation is measured along with spectrum sensing reputation and spectrum etiquette reputation respectively based on 2 parts of SUs in the cognitive cycle (Pei et al., 2013). Chen H et al. have proposed a reputation based cooperative spectrum sensing method in CRNs to work out the indeterminacy brought by the shadowing impact and multipath fading (Chen et al., 2009). However, the schemes in (Pei et al., 2013, Chen et al., 2009) do not mitigate the influence of inside attack. So there is a pressing requirement for the trust-based methodology to overcome the security threats in the IEEE 802.22 WRANs as one of the instances of the CRNs. Nevertheless, considering computational trust as a way to acquire security has not been addressed much in the research so far for WRANs. We, therefore, propose a novel dynamic trust model (DTM) for the spectrum management by the SUBS based on the previous behaviors of the CPEs to enhance the security functionality in the WRANs, which can make the communication more robust and secure. The SUBS only believes the sensing data from the trusted CPEs for the spectrum decision and allocates the spectrum resources to the trusted CPEs. The model has a significant impact on reducing the influence of users’ vicious behaviors and encourages the CPEs to perform truthful and positive actions in the communications. It is assumed that the SUBSs in the WRANs are trustworthy while the CPEs controlled by the SUBS may not. The untrustworthiness of the CPEs mainly comes from the fact that CPEs may not report the right sensing data before the spectrum resource allocation and that they may not vacate the spectrum when the PUs come after spectrum resource allocation. By the proposed DTM scheme, whenever a CPE would like to gain access to the spectrum holes in a WRAN, if the trust value of the CPE evaluated on its previous behaviors by the subjective logic is less than the given critical value, the CPE cannot access the resource. The DTM works based on the behavior pattern of the CPEs and can quickly find out the users who have reported false sensing data. In addition, the DTM scheme follows an equitable trust-based spectrum allocation method by which the CPE's with higher trust values will have higher priorities to gain access the free spectrum and encourages the CPEs to send the right sensing information and obey the opportunistic access rules.

The paper is organized as follows: in Section 2, we briefly present a review of the system background of the WRANs. In Section 3, we provide trust metrics and the proposed DTM model in detail, followed by the evaluation the proposed DTM model in Section 4. We conclude the paper in Section 5.

Section snippets

PRELIMINARY — Subjective logic

Subjective logic theory offers a helpful mathematical framework to model and analyze trust networks containing indeterminacy and partial knowledge. Subjective logic is a logic which acts on the subjective faiths about the facts, and utilizes the term opinion to signify the delegate of a subjective faith (Josang, 2001). Subjective logic offers an image between the evidence space and the opinion space. The trust value of an entity's is calculated on the basis of the opinion. The opinion is

Proposed trust model

The proposed DTM uses the subjective logic to evaluate the trust value of a CPE based on its previous behaviors. The trust value is dynamic, which satisfies that trust is difficult to gain, easy to lose. The DTM includes direct trust value, indirect trust value, incentive trust value, the integrated trust value and the update of trust value.

Analysis

In this section, we first analyze the time complexity and then perform simulation experiments to evaluate the performance of the proposed DTM scheme by using Python. Python is an interpreted and object-oriented high-level programming language with the dynamic data types. It is easy to learn and widely used with a good portability and a high degree of scalability. We use Python 2.7.5 in Window 7 to evaluate the performance of the DTM.

Conclusion

In this paper, we have proposed a subjective trust model for the spectrum management based on the previous behaviors for the centralized WRANs in TVWS. The trust value of a CPE is calculated by its previous behavior including whether sensing the right spectrum hole and whether vacating the spectrum when a PU comes, etc. The trust value is used as the foundation of the spectrum decision by the SUBS. The trust-based spectrum sensing data can greatly enhance the validity of the spectrum decision.

Acknowledgments

This research was supported in part by the National Natural Science Foundation of China (NSFC) under grant No. 61702367 and No. 61172063.

Cong Wang was born on February 10, 1988 in Hunan, China. She is now a teacher of Colleage of Computer Science and Information Engineering, Tianjin University of Science and Technology, China. She received the Ph.D. degree in 2017 from School of Computer Science and Technology, Tianjin University (TJU) and the B.E. degree in 2012 from TJU, majoring in wireless network. Her research interest includes TV White Space Security and Authentication Protocol design

References (29)

  • Federal Communications Commission (FCC), 2012. ET Docket 10-174: Second Memorandum Opinion and Order in the Matter of...
  • Gao W., Zhang G., Chen G., Li Y., 2009. A Trust Model Based on Subjective Logic. In: Proceedings of the 4th...
  • K. Govindan et al.

    Trust computations and Trust dynamics in mobile Adhoc networks: a survey

    IEEE Commun. Surv. Tutor.

    (2012)
  • K. Govindan et al.

    Trust computations and Trust dynamics in mobile Adhoc networks: a survey

    IEEE Commun. Surv. Tutor.

    (2012)
  • Cited by (3)

    • Performance analysis of radio spectrum for cognitive radio wireless networks using discrete time Markov chain

      2019, Journal of Systems and Software
      Citation Excerpt :

      According to 802.22-2011 (2011), the unlicensed secondary devices can use highly underutilized TV band ranging from 54 MHz to 862 MHz provided that they will not interfere with incumbents (licensed user for TV band). The purpose of this deployment is basically to provide broadband wireless access to rural areas (Wang et al., 2017; Mufti et al., 2017). Its coverage area is between 33 km to 100 km, which is greater than IEEE 802.16 WiMAX and IEEE 802.11 WLAN, with minimum throughput of 384 kbps in upstream (uplink) direction and 1.5 Mbps in downstream (downlink) direction.

    Cong Wang was born on February 10, 1988 in Hunan, China. She is now a teacher of Colleage of Computer Science and Information Engineering, Tianjin University of Science and Technology, China. She received the Ph.D. degree in 2017 from School of Computer Science and Technology, Tianjin University (TJU) and the B.E. degree in 2012 from TJU, majoring in wireless network. Her research interest includes TV White Space Security and Authentication Protocol design

    Maode Ma (SM’09) received the B.E. degree from Tsinghua University, Beijing, China, in 1982, the M.E. degree from Tianjin University, Tianjin, China, in 1991, and the Ph.D. degree from the Hong Kong University of Science and Technology, Hong Kong, China, in 1999. He is now an Associate Professor at the School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. He has about 200 international academic publications including more than 80 journal papers, more than 140 conference papers and/or book chapters, and three academic books. His research interests are wireless networking and wireless network security. Dr. Ma is a member of a few technical committees in the IEEE Communication Society. He has been a member of the technical program committees for more than 100 international conferences. He has been a general chair, technical symposium chair, tutorial chair, publication chair, publicity chair and session chair for more than 50 international conferences. He serves as an Editor-in-Chief/Associate Editor for six International journals.

    Zenghua Zhao is an Associate Professor of Computer Science at Tianjin University, China. She received her M.S. and Ph.D. in Computer Science, Tianjin University, China in 2000 and 2006 respectively. She is an IEEE/ACM member. Her current interests are focused on Wireless and mobile computing, long-distance wireless mesh networks, wireless sensor networks and network system design.

    View full text