Abstract

The current process of designing secure protocols is tantamount to an arms race between attacks and “patches”. We introduce a system theoretic approach to secure protocol design with provable security guarantees against all attacks that fall within the system model. In this approach, we frame the secure protocol design problem as a max-min optimization of a specific payoff function, where the adversarial nodes with a priori knowledge of the protocol choose a strategy that minimizes the payoff, and the protocol seeks to maximize this worst-case payoff. We make the following three contributions. First, we solve the optimization problem. That is, we describe a protocol and show that is max-min optimal. Second, we show that the protocol is actually min-max optimal which is generally higher than max-min optimality. Finally, we show that the adversarial nodes are effectively limited to one of two modes: either jamming or conforming with the protocol.

A Clean Slate Approach to Secure Wireless Networking

The current state-of-art in securing computers and data is tantamount to an arms race between the attackers on one hand and the software developers who have to provide the updates to counteract the attacks on the other.

This monograph introduces a novel technique for designing secure protocols with provable security guarantees for wireless ad-hoc networks infiltrated with adversarial nodes. It does this by developing a system-theoretical approach to secure protocol design that provides provable and comprehensive security guarantees, where any features of a protocol that reduce complexity or maximize the network throughput do not come at the expense of security. The desired approach is one of security first and performance second, which is the reverse of the current approach.

This novel approach uses a game-theoretic framework that addresses the deficiencies in the defense-in-depth approach to securing complex systems. A Clean Slate Approach to Secure Wireless Networking describes in detail the theory of using this approach in both closed synchronized and unsynchronized networks. It will be of interest to all researchers and designers working on securing wireless networks.