INTRODUCTION Mobile computation, where independent agents roam widely distributed networks in search of resources and information, is fast becoming a reality. A number of programming languages, APIs and protocols have recently emerged which seek to provide high-level support for mobile agents. These include Java [30], Odyssey [15], Aglets [19], Voyager [24] and the latest revisions of the Internet protocol [25, 2]. In addition to these commercial efforts, many prototype languages have been developed and implemented within the programming language research community --- examples include Linda [8, 9], Facile [16], Obliq [7], Infospheres [11], the join calculus [13], and Nomadic Pict [33]. In this paper we address the issue of resource access control for such languages. Central to the paradigm of mobile computation are the notions of agent, resource and location. Agents are effective entities that perform computation and interact with other First publis

We study the distributed infrastructures required for location-independent communication between migrating agents. These infrastructures are problematic: different applications may have very different patterns of migration and communication, and require different performance and robustness properties; algorithms must be designed with these in mind. To study this problem we introduce an agent programming language - Nomadic Pict. It is designed to allow infrastructure algorithms to be expressed as clearly as possible, as translations from a high-level language to a low level. The levels are based on rigorously-defined process calculi, they provide sharp levels of abstraction. In this paper we describe the language and use it to develop an infrastructure for an example application. The language and examples have been implemented; we conclude with a description of the compiler and runtime.

Abstract. This paper presents an abstract machine for a variant of the stochastic pi-calculus, in order to correctly model the stochastic simulation of biological processes. The abstract machine is proved sound and complete with respect to the calculus, and then used as the basis for implementing a stochastic simulator. The correctness of the machine helps ensure that the simulator is correctly implemented, giving greater confidence in the simulation results. A graphical representation for the pi-calculus is also presented, as a potential front-end to the simulator. 1

We study communication primitives for interaction between mobile agents. They can be classified into two groups. At a low level there are location dependent primitives that require a programmer to know the current site of a mobile agent in order to communicate with it. At a high level there are location independent primitives that allow communication with a mobile agent irrespective of its current site and of any migrations. Implementation of these requires delicate distributed infrastructure. We propose a simple calculus of agents that allows implementations of such distributed infrastructure algorithms to be expressed as encodings, or compilations, of the whole calculus into the fragment with only location dependent communication. These encodings give executable descriptions of the algorithms, providing a clean implementation strategy for prototype languages. The calculus is equipped with a precise semantics, providing a solid basis for understanding the algorithms and for reasoning about their correctness and robustness. Two sample infrastructure algorithms are presented as encodings.

We introduce a calculus which is a direct extension of both the and the π calculi. We give a simple type system for it, that encompasses both Curry's type inference for the -calculus, and Milner's sorting for the π-calculus as particular cases of typing. We observe that the various continuation passing style transformations for -terms, written in our calculus, actually correspond to encodings already given by Milner and others for evaluation strategies of -terms into the π-calculus. Furthermore, the associated sortings correspond to well-known double negation translations on types. Finally we provide an adequate cps transform from our calculus to the π-calculus. This shows that the latter may be regarded as an "assembly language", while our calculus seems to provide a better programming notation for higher-order concurrency.

This paper considers how locality restrictions on the use of capabilities can be enforced by a static type system. A distributed π-calculus with a simple reduction semantics is introduced, integrating location and migration primitives from the Distributed Join Calculus with asynchronous π communication. It is given a type system in which the input and output capabilities of channels may be either global, local or absent. This allows compile-time optimization where possible but retains the expressiveness of channel communication. Subtyping allows all communications to be invoked uniformly. We show that the most local possible capabilities for internal channels can be inferred automatically.

Abstract. We define an extension of the π-calculus with a static type system which supports high-level specifications of extended patterns of communication, such as client-server protocols. Subtyping allows protocol specifications to be extended in order to describe richer behaviour; an implemented server can then be replaced by a refined implementation, without invalidating type-correctness of the overall system. We use the POP3 protocol as a concrete example of this technique. 1

An interpretation of Abadi and Cardelli's first-order Imperative Object Calculus into a typed pi-calculus is presented. The interpretation validates the subtyping relation and the typing judgements of the Object Calculus, and is computationally adequate. The proof of computational adequacy makes use of (a pi-calculus version) of ready simulation, and of a factorisation of the interpretation into a functional part and a very simple imperative part. The interpretation can be used to compare and contrast the Imperative and the Functional Object Calculi, and to prove properties about them, within a unified framework.

The theory of relational parametricity and its logical relations proof technique are powerful tools for reasoning about information hiding in the polymorphic -calculus. We investigate the application of these tools in the security domain by defining a cryptographic -calculus---an extension of the standard simply typed -calculus with primitives for encryption, decryption, and key generation--- and introducing syntactic logical relations (in the style of Pitts and Birkedal-Harper) for this calculus that can be used to prove behavioral equivalences between programs that use encryption. We illustrate

www.cs.bu.edu/fac/kfoury The Ambient Calculus was developed by Cardelli and Gordon as a formal framework to study issues of mobility and migrant code [CG98]. We consider an Ambient Calculus where ambients transport and exchange programs rather that just inert data. We propose different senses in which such a calculus can be said to be polymorphically typed, and design accordingly a polymorphic type system for it. Our type system assigns types to embedded programs and what we call behaviors to processes; a denotational semantics of behaviors is then proposed, here called trace semantics, underlying much of the remaining analysis. We state and prove a Subject Reduction property for our polymorphically-typed calculus. Based on techniques borrowed from finite automata theory, type-checking of fully type-annotated processes is shown to be decidable. Our polymorphically-typed calculus is a conservative extension of the typed Ambient Calculus originally proposed by Cardelli and Gordon [CG99].