Security and Robustness for Collaborative Monitors
Decentralized monitors can be subject to robustness and security risks. Robustness risks include attacks on the monitor’s infrastructure in order to disable parts of its functionality. Security risks include attacks that try to extract information from th
- PDF / 495,396 Bytes
- 20 Pages / 439.37 x 666.142 pts Page_size
- 84 Downloads / 214 Views
2
Utrecht University, Utrecht, Netherlands {B.J.G.Testerink,M.M.Dastani}@uu.nl Delft University of Technology, Delft, Netherlands [email protected]
Abstract. Decentralized monitors can be subject to robustness and security risks. Robustness risks include attacks on the monitor’s infrastructure in order to disable parts of its functionality. Security risks include attacks that try to extract information from the monitor and thereby possibly leak sensitive information. Formal methods to analyze the design of a monitor with respect to these issues can help to create more secure designs and/or identify critical parts. In this paper we specify a model for analyzing robustness and security risks for collaborative monitors constructed from a network of local monitors. Keywords: Monitoring
1
· Runtime verification · Security
Introduction
Normative systems help to make sure that agents behave according to preset guidelines/norms in multi-agent systems [5]. One approach is provided by exogenous normative systems where norms are explicit. The normative aspect of the multi-agent system is captured by an exogenous—to the agents—organization or institution. With this approach it must be verified whether any norm violation occurs in the multi-agent system’s execution. Monitoring large distributed multi-agent systems such as traffic, smart grids and economic markets requires decentralized approaches. Monolithic centralized monitors can impose a bottleneck due to the distributed nature of multi-agent systems and a single point of failure in case of break downs. A major concern of many decentralized verification applications is their robustness and security. The data that is gathered from a multi-agent system can severely compromise the agents’ privacy if leaked. Adversaries can also try to take down parts of the network to impede its functioning. Formal models of decentralized monitors allow for the analysis of critical parts in monitors in terms of robustness and security. Such an analysis allows the developers of decentralized monitors to invest more resources in critical parts. In this paper we present a formal model for decentralized monitors that supports their formal analysis to face the aspects of robustness and security when designing a monitor. As an c Springer International Publishing Switzerland 2016 V. Dignum et al. (Eds.): COIN 2015, LNAI 9628, pp. 376–395, 2016. DOI: 10.1007/978-3-319-42691-4 21
Security and Robustness for Collaborative Monitors
377
Location 2 Location 1
Location 4
Location 3
Fig. 1. Example scenario. Black dots indicate locations, arrows indicate traffic flow and double lines indicate roads.
example we shall use an abstract traffic monitoring scenario (Example 1). Traffic monitoring faces many challenges, including physical attacks on the monitor infrastructure and the privacy of individuals (cf. [7]). In our approach we assume that monitors observe the execution trace of a system in order to detect specific properties of its behavior. These properties are expressed in linear-time temporal log
Data Loading...
