Maximizing lifetime of event-unobservable wireless sensor networks Kemal Bicakci a, ⁎, Hakan Gultekin b,1 , Bulent Tavli c,2 , Ibrahim Ethem Bagci a,3 a TOBB University of Economics and Technology, Computer Engineering Department, Ankara, Turkey b TOBB University of Economics and Technology, Industrial Engineering Department, Ankara, Turkey c TOBB University of Economics and Technology, Electrical and Electronics Engineering Department, Ankara, Turkey abstract article info Article history: Received 30 April 2010 Received in revised form 18 January 2011 Accepted 18 January 2011 Available online 26 January 2011 Keywords: Wireless sensor networks Contextual privacy Transactional confidentiality Privacy Location privacy Event-unobservability Linear Programming In wireless sensor networks (WSNs) contextual information such as the information regarding whether, when, and where the data is collected cannot be protected using only traditional measures (e.g., encryption). Contextual information can be protected against global eavesdroppers by periodic packet transmission combined with dummy traffic filtering at proxy nodes. In this paper, through a Linear Programming (LP) framework, we analyze lifetime limits of WSNs preserving event-unobservability with different proxy assignment methodologies. We show that to maximize the network lifetime data flow should pass through multiple proxies that are organized as a general directed graph rather than as a tree. © 2011 Elsevier B.V. All rights reserved. 1. Introduction In WSNs, sensor nodes convey their data to a base station possibly relaying through multiple intermediate sensor nodes [1]. Security and privacy in WSNs deployed in harsh and hostile environments are of paramount importance. Protecting integrity and confidentiality of sensor data content—a well-studied problem in the literature [2,3]—is generally achieved using cryptographic tools. On the other hand, contextual information such as whether, when, and where the data is collected cannot be secured only by traditional methods. This so called “event-unobservability” problem is usually addressed by inserting redundant packets into the network. By this way, attackers cannot distinguish real packets and, thus, obtain no useful information by traffic monitoring. The amount of overhead to be carried in the network to safeguard against traffic analysis depends on the threat model (i.e., what capabilities attackers have). The possibility of the existence of a global eavesdropper who can monitor the entire network traffic brings a greater challenge for resource-constraint sensor nodes because a high-assurance solution should depend on a periodic collection method in which every sensor node periodically send encrypted packets regardless of whether there is real data to send or not [4]. Periodic transmission together with encryption hides the source of real packets against external attackers who do not hold the decryption key. Battery power is a limited resource and communication is the dominant factor of energy dissipation in WSNs. Therefore, to increase network lifetime, there is a need for a more communication-efficient technique which reduces the overhead of the periodic collection scheme mentioned above. The basic idea of more advanced solutions referred in the literature as proxy-filtering techniques is that all sensor nodes again periodically generate packets but some of nodes assigned as proxies aggregate all incoming packets into a single packet [5]. Several proxy-filtering techniques were proposed in the past but their impact on the network lifetime has yet to be explored. In this study, through an LP framework, we make the first attempt in the literature to investigate the practical limits on the network lifetime of WSNs using proxy filtering techniques in idealized conditions. We investigate various different proxy assignment strategies and different deployment scenarios. Our investigation leads us to put forward a new filtering idea called OFS (Optimal Filtering Scheme) and enables us to maximize the network lifetime of wireless sensor networks which preserve event-unobservability against global eavesdroppers. The rest of our paper is organized as follows: Section 2 provides the preliminaries for our paper. Section 3 introduces the LP formulation to investigate performance bounds of different proxy filtering techniques. Section 4 covers the experimental analysis we conduct to investigate various aspects of the problem. Section 5 discusses the results of this Computer Standards & Interfaces 33 (2011) 401–410 ⁎ Corresponding author. Tel.: +90 312 292 4262; fax: +90 312 292 4180. E-mail addresses: bicakci@etu.edu.tr (K. Bicakci), hgultekin@etu.edu.tr (H. Gultekin), btavli@etu.edu.tr (B. Tavli), iebagci@etu.edu.tr (I.E. Bagci). 1 Tel.: +90 312 292 4275; fax: +90 312 292 4091. 2 Tel.: +90 312 292 4074; fax: +90 312 292 4180. 3 Tel.: +90 312 292 4000x5045; fax: +90 312 292 4180. 0920-5489/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.csi.2011.01.001 Contents lists available at ScienceDirect Computer Standards & Interfaces journal homepage: www.elsevier.com/locate/csi