DISH: Distributed Self-Healing in Unattended Sensor Networks
Unattended wireless sensor networks (UWSNs) operating in hostile environments face the risk of compromise. % by a mobile adversary. Unable to off-load collected data to a sink or some other trusted external entity, sensors must protect themselves by attempting to mitigate potential compromise and safeguarding their data. In this paper, we focus on techniques that allow unattended sensors to recover from intrusions by soliciting help from peer sensors. We define a realistic adversarial model and show how certain simple defense methods can result in sensors re-gaining secrecy and authenticity of collected data, despite adversary's efforts to the contrary. We present an extensive analysis and a set of simulation results that support our observations and demonstrate the effectiveness of proposed techniques.
A New Approach to Secure Logging
The need for secure logging is well-understood by the security professionals, including both researchers and practitioners. The ability to efficiently verify all (or some) log entries is important to any application employing secure logging techniques. In this paper, we begin by examining state-of-the-art in secure logging and identify some problems inherent to systems based on trusted third-party servers. We then propose a different approach to secure logging based upon recently developed Forward-Secure Sequential Aggregate (FssAgg) authentication techniques. Our approach offers both space-efficiency and provable security. We illustrate two concrete schemes -- one private-verifiable and one public-verifiable -- that offer practical secure logging without any reliance on on-line trusted third parties or secure hardware. We also investigate the concept of immutability in the context of forward secure sequential aggregate authentication to provide finer grained verification. Finally, we report on some experience with a prototype built upon a popular code version control system.
Forward-Secure Sequential Aggregate Authentication
Wireless sensors are employed in a wide range of applications. One common feature of most sensor settings is the need to communicate sensed data to some collection point or sink. This communication can be direct (to a mobile collector) or indirect -- via other sensors towards a remote sink. In either case, a sensor might not be able to communicate to a sink at will. Instead it collects data and waits (for a potentially long time) for a signal to upload accumulated data directly. In a hostile setting, a sensor may be compromised and its post-compromise data can be manipulated. One important issue is Forward Security -- how to ensure that pre-compromise data cannot be manipulated? Since a typical sensor is limited in storage and communication facilities, another issue is how to minimize resource consumption due to accumulated data. It turns out that current techniques are insufficient to address both challenges. To this end, we explore the notion of Forward-Secure Sequential Aggregate (FssAgg) Authentication Schemes. We consider FssAgg authentication schemes in the contexts of both conventional and public key cryptography and construct a FssAgg MAC scheme and a FssAgg signature scheme, each suitable under different assumptions. This work represents the initial investigation of Forward-Secure Aggregation and, although the proposed schemes are not optimal, it opens a new direction for follow-on research.