Deceptive Routing Games Quanyan Zhu, Andrew Clark, Radha Poovendran and Tamer Bas ¸ar Abstract— The use of a shared medium leaves wireless net- works, including mobile ad hoc and sensor networks, vulnerable to jamming attacks. In this paper, we introduce a jamming defense mechanism for multiple-path routing networks based on maintaining deceptive flows, consisting of fake packets, between a source and a destination. An adversary observing a deceptive flow will expend energy on disrupting the fake packets, allowing the real data packets to arrive at the destination unharmed. We model this deceptive flow-based defense within a multi-stage stochastic game framework between the network nodes, which choose a routing path and flow rates for the real and fake data, and an adversary, which chooses which fraction of each flow to target at each hop. We develop an efficient, distributed procedure for computing the optimal routing at each hop and the optimal flow allocation at the destination. Furthermore, by studying the equilibria of the game, we quantify the benefit arising from deception, as reflected in an increase in the valid throughput. Our results are demonstrated via a simulation study. I. I NTRODUCTION Multi-hop wireless networks have been deployed or envi- sioned in applications ranging from infrastructure monitoring to battlefield communication [1]. The use of an open wire- less medium, however, leaves wireless networks vulnerable to jamming attacks, in which an adversary broadcasts an interfering signal in the vicinity of a receiving node and thereby prevents packets from being correctly decoded [2]. The jamming attack can severely limit the throughput of a communication session unless defense measures are taken. Current approaches to mitigating jamming attacks use randomization techniques at one or more layers to prevent the adversary from targeting packets. At the physical layer, fre- quency hopping is used to prevent the adversary from identi- fying the frequency band used by the nodes [3]. Generalized mechanism-hopping methods are employed at higher layers, in which the network nodes switch between different com- munication protocols in order to prevent protocol-specific attacks [4]. Jamming has also been mitigated by dividing traffic flows among multiple paths, so that network flows can be shifted away from paths that are being jammed [5], [6]. These defense methods are inherently reactive, in that they are not activated until the adversary has already reduced the network throughput. Furthermore, while their main objective The research was partially supported by the AFOSR MURI Grant FA9550-10-1-0573, and also by an NSA Grant through the Information Trust Institute at the University of Illinois. Q. Zhu and T. Bas ¸ar are with the Coordinated Science Laboratory and Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801 USA. Email: {zhu31, basar1}@illinois.edu Andrew Clark and Radha Poovendran are with the Department of Electrical Engineering, University of Washington, Seattle, WA 98195 USA. Email: {awclark, rp3}@u.washington.edu is to hide information on the frequency channel, communi- cation protocol, or network routing topology, they do not employ deception to actively mislead the adversary. In this paper, we introduce a proactive deception mecha- nism for mitigating wireless jamming, in which the source node introduces a false flow, consisting of randomly gen- erated packets. If all traffic is encrypted, then an adversary will not be able to distinguish between real and false flows. The adversary will then expend its limited resources, such as jamming power, on attacking the false flow, thereby reducing the impact on the real flow. We introduce a game-theoretic framework for modeling and developing deceptive flow-based jamming mitigation between a single source and a destination. Our framework consists of two components. First, at the intermediate nodes between the source and the destination, we formulate a multi- person Stackelberg game, in which the intermediate node moves first and chooses the next hop for both the real and false flows. The adversary then selects how much power to allocate to jamming each flow at that hop. We introduce the concept of path Stackelberg equilibrium, describing the optimal strategies of both the intermediate node and the adversary, and prove the existence of such an equilibrium in behavioral mixed strategies. Second, we consider the rates chosen for the real and deceptive flows at the source. Under this formulation, the source first chooses the flow rates, and then the intermediate nodes and adversary respond by choosing routing and jam- ming strategies, respectively. We introduce a rate Stackelberg equilibrium describing the optimal flow allocation by the source, and prove the existence of such an equilibrium. We provide an efficient procedure for computing both the path and rate Stackelberg equilibria for a given network. We demonstrate this approach by analyzing a network in which the source has a logarithmic utility function and the adversaries pursue independent strategies at each hop. The effectiveness of our approach is demonstrated through numerical examples. We also introduce the notion of value of deception in order to evaluate the benefit of deception in multi-hop routing games. When the value is greater than 1, deception is valuable to the source node for mitigating the attack, and the utility gain of deceptive routing is measured by the difference between the equilibrium utility of the game and the utility under routing strategies without deception. The rest of this paper is organized as follows. In Section III, we present our system model and game formulations. We discuss the existence of equilibrium solutions and a backward induction method to compute the equilibrium. In