ECCE: Enhanced cooperative channel establishment for secure pair-wise communication in wireless sensor networks

Abstract

This paper presents the ECCE protocol, a new distributed, probabilistic, cooperative protocol to establish a secure pair-wise communication channel between any pair of sensors in a wireless sensor network (WSN). The main contributions of the ECCE protocol are: to allow the set-up of a secure channel between two sensors (principals) that do not share any pre-deployed key. This feature is obtained involving a set of sensors (cooperators) in the channel establishment protocol; to provide probabilistic authentication of the principals as well as the cooperators. In particular, the probability for the attacker to break authentication check decreases exponentially with the number of cooperators involved; to trade off the memory space required to store the pre-deployed encryption keys with the number of cooperators involved in the protocol. Hence, memory storage can be used to store keys built with the ECCE protocol, which helps amortizing the (limited) overhead incurred in the ECCE key set-up; to be adaptive to the level of threat the WSN is subject to. We provide analytical analysis and extensive simulations of ECCE, which show that the proposed solution increases both the probability of a secure channel set-up and the probability of channel resilience with respect to other protocols.

Introduction

A wireless sensor network (WSN) is a collection of sensors whose size can range from a few hundred sensors to a few hundred thousand or possibly more. Sensors do not rely on any pre-deployed network architecture, thus they communicate via an ad-hoc wireless network. The power supply of each individual sensor is provided by a battery, hence both communication and computation activities must be optimized. Distributed in irregular patterns and left unattended, sensors should autonomously aggregate into collaborative, peer-to-peer networks. Sensors networks must be robust and survivable in order to overcome individual sensor failure due to either malicious (e.g. destruction) or non-malicious (e.g. battery depletion) events. A WSN can be deployed in both military and civil scenarios [1]. For instance, it could be used: to provide a relay network for tactical communications in the battlefield; to collect data from a field in order to reveal the presence of toxic gases; to facilitate rescue operations in wide open hostile areas; to fulfill perimeter surveillance duties; to operate in harsh environments for commercial purposes.

WSNs are expected to be the basic building block of pervasive computing environments, hence establishing secure pair-wise communications could be useful for many applications. In particular, it is a pre-requisite for the implementation of secure routing, and can be useful for secure group communications. Further, secure pair-wise communication allow in-network processing [25], or facilitate the establishment of a cluster key, hence enabling passive participation, in which a sensor can take certain actions based on overheard messages. It was pointed out in [21], [1], [5], that asymmetric cryptography such as RSA or elliptic curve cryptography (ECC) is unsuitable for most sensor architectures due to high energy consumption and increased code storage requirements. However, it is worth noticing that evolution in technology allows to sparingly use asymmetric cryptography for a certain class of WSN [24]. For instance, in [20] the authors devise a protocol that, with the seldom use of ECC, thwarts the replication attack [19], [13]. However, it seems reasonable that there will be always some classes of WSNs in which asymmetric cryptography would rise an unfeasible cost due to either energy consumption or memory constraints. Indeed, as for energy consumption, in a mobile WSN if we have a secure key establishment protocol based on ECC whenever two sensors wants to agree on a shared key for the first time, this would put high requirements on battery consumption. As for memory, it seems unfeasible that any node could host the public keys of all the other nodes in the network (for instance, the Mica mote is equipped with a 8 bit 4 MHz processor and has 4 kB of RAM and 128 kB of flash RAM only). Note that the constraint on memory stands even in a static WSN. Hence, while solutions that intend to address specific problems can directly benefit of the sparingly use of EEC [20], building communication channels on symmetric algorithms, which are three order of magnitude more efficient than ECC [24], is still an attractive research field [5], [22].

This paper presents the ECCE protocol, a new protocol to establish a secure pair-wise communication channel between any pair of sensors in the WSN. The ECCE protocol can be classified as probabilistic and cooperative. Unlike other protocols for channel establishment, ECCE allows to establish a secure channel between sensors that do not share any key, involving a set of cooperating sensors (cooperators) which are not required to share a key with both principals. The same feature is not present in actual protocols such as Multipath Key Reinforcement [6] and Cooperative [10]. The overhead required is limited and it is sustained just once during the sensor life-time. ECCE shows better performance in channel existence and channel resilience than existing protocols. The Protocol also guarantees implicit and probabilistic mutual authentication of principals and cooperators without any additional overhead and without the presence of a base station. Further, the proposed protocol could be used also between sensors that already share some secret keys to increase the resilience of these shared keys. The proposed protocol is also adaptive to the required security level: to achieve an higher level of security, it suffices to involve an higher number of cooperators in the channel set-up. Finally, the protocol allows to trade off the memory required to store pre-deployed keys with cooperators. In particular, it is possible to set the number of cooperators in order to have a reduced key ring that provides the same level of security and the same probability of channel existence of solutions that involve no cooperators but a larger key ring size. For example, choosing a pool of size 1.000, a key ring of size 12, and involving 8 cooperators, provides the same probability of channel existence of a scenario in which every sensor has 20 pre-deployed keys but there are no cooperators. As for resiliency, with a pool of size 10.000, a key ring of size 100 and 8 cooperating sensors, the attacker is required to capture 110 sensors to corrupt a channel, while with the same parameters, but with no cooperators, the attacker has to corrupt only 75 sensors to corrupt the channel. Note that reducing the key ring size provides the possibility for sensors to store the cooperative keys set-up with the ECCE protocol. Analytical and experimental results show that the ECCE protocol has better performance than the other protocols as for channel existence and channel resiliency to the attacker.

The remainder of this paper is organized as follows. In Section 2, we review the current contributions in the field. In Section 3, we report some preliminaries and define our system assumptions. In Section 4, we describe the ECCE protocol, while in Section 5, we analyze the probability to establish a secure channel and the resilience of the established channel.