An Overview of FORCES: An INRIA Project on Declarative Formalisms for Emergent Systems

The FORCES project aims at providing robust and declarative formalisms for analyzing systems in the emerging areas of Security Protocols, Biological Systems and Multimedia Semantic Interaction. This short paper describes FORCES’s motivations, results and future research directions. Introduction FORCES (FORmalisms from Concurrency for Emergent Systems) is an ongoing project funded by the Equipes Associées programme of INRIA. It is carried by the INRIA team COMETE (France), the IRCAM Music Representation Team (France) and the team AVISPA (Colombia). The main goal of FORCES is to provide robust declarative formalisms for modeling systems from emergent application areas of computer science in which our teams have been working during recent years: Namely, Security Protocols, Biological Systems and Multimedia Semantic Interaction. Process calculi are formalisms that treat communicating processes much like the λ-calculus treats computable functions: The structure of terms reflects the structure of processes and process evolution is represented by term reduction. Concurrent Constraint Programming (CCP) based calculi [1] are computational models that combine the operational view of process calculi with a declarative one based upon logic. Some of the members of FORCES developed and used ntcc [2], a timed CCP calculus, to predict the behavior of systems from Security Protocols [3], Systems Biology [4] and Multimedia Semantic Interaction [5]. Although these areas differ significantly from one another, there is a crucial commonality in the analysis we wanted to perform in them: Reachability i.e., whether a system reaches a particular state. The ntcc calculus provides several reasoning tools for reachability analysis. These include a temporal logic, a proof system, verification techniques, and a denotational semantics. Nevertheless, we have learned from our modeling experience and theoretical studies that ntcc is not sufficiently robust for these applications. E.g., some security protocols use a mechanism to allow communication of nonces (i.e., uniquely generated random number). The ntcc calculus can at best express this mechanism indirectly [6]. Also, ntcc lacks constructs for quantitative information, which are essential for biological systems. Furthermore, we have identified musical settings exhibiting complex non-regular timed behavior that cannot be expressed in ntcc. Our research strategy in FORCES has been, with the benefit of hindsight, to develop declarative formalisms for modeling systems from the above-mentioned areas as suitable extensions or specializations of ntcc. Our expertise in ntcc as well as our modeling experience have been fundamental for guiding our research. This short paper provides an overview of FORCES. Further information can be found at http://www.lix.polytechnique.fr/comete/Forces. Declarative Models of Security Protocols A fundamental ability for security protocols is that of generating and communicating private nonces; process calculi for security therefore include mechanisms for creating and communicating local names. Neither ntcc nor its predecessor tcc [7] features such mechanisms. As a remedy to this, in [3] we introduced the Universal Timed CCP process calculus (utcc): a generalization of tcc that allows for the communication of local names (or links). This additional expressiveness paves the way for the declarative modeling of a wider class of systems, most notably dynamic ones. We have endowed utcc with a number of reasoning techniques for reachability analysis. A symbolic semantics was defined to deal with problematic operational aspects involving infinitely many substitutions which often arise when modeling security protocols. The semantics uses temporal constraints to finitely represent infinitely-many substitutions; it has been used to exhibit secrecy flaws in some security protocols [3]. The utcc calculus also enjoys a declarative view of processes as First-Order Temporal Logic (FLTL) formulae [8]. This allows for reachability analysis of utcc processes using FLTL techniques. For instance, in [3] we used the FLTL formulae representing the model of a protocol to know if it reaches a state where the attacker knows a secret. We also defined a denotational semantics for utcc [9]. This way, processes can be represented as partial closure operators. As an application of the semantics, we identified a language for security protocols that can be represented as closure operators over a cryptographic constraint system. We showed that the least fixed point of such an operator may then be used to check a secrecy property in a protocol. To our knowledge, this is the first denotational account in the context of calculi for security protocols. This way, our work has brought new semantic insights into the verification of security protocols, and is related to the research in security protocols from areas closely related to CCP. Namely, Constraint Programming (e.g. [10]) and Logic Programming (e.g. [11,12]). To our knowledge there is no work on Security Protocols that takes advantage of the reasoning techniques of CCP. Declarative Models of Biological Systems Quantitative information is fundamental for biological systems. For example, behavior in most biochemical reactions is highly dependent on the presence of a certain amount of the substances involved. Very often, information is partial as obtaining exact values for parameterizing models is difficult. Unpredictable behavior is thus an inherent condition of the biologic phenomena, and one counts with partial behavioral information for describing system interactions. This partial information not only ignores elements on how reactions occur (e.g. what components actually interact), but also on when such reactions commonly happen (e.g. the relative speeds of the interacting components). While the notion of partial quantitative information is central to CCP via constraints, partial behavioral information is actually the novelty of ntcc via non-deterministic and