Goals and Resource Constraints in CARMA

Paul Piho,Anastasis Georgoulas,Jane Hillston

doi:10.1016/j.entcs.2018.03.039

Abstract

Carma is a recently developed, high-level quantitative modelling language developed for the design and analysis of collective adaptive systems. In the current Carma language, agents within a system consist of a behaviour, captured as a process, and knowledge, represented as a store of attributes. In this paper we present the first steps to equipping agent specifications with more sophisticated forms of knowledge in terms of goals and targets, and demonstrate how these may be integrated into the modelling and analysis process. We illustrate the ideas with a simple example taken from the domain of swarm robotics.

Highlights

Examples of collective adaptive systems (CAS) are widespread in nature, ranging from the swarming behaviour of insects to patterns of epidemic spread in humans
Georgoulas, Hillston interact with their neighbours. Through these interactions complex emergent behaviour may be formed at the system level, often of a form which is difficult to predict from the simple behaviour of the individual agents
Modelling becomes a valuable tool in the design of collective adaptive systems, allowing possible configurations of agents to be evaluated with respect to global goals

Summary

Introduction

Examples of collective adaptive systems (CAS) are widespread in nature, ranging from the swarming behaviour of insects to patterns of epidemic spread in humans. Collective Adaptive Resource-sharing Markovian Agents, is a recently defined, high-level stochastic process algebra-based modelling language [10]. As a first step towards this long-term objective, in this paper we present an extension of the Carma modelling language which supports the specification of resource constraints and system-level functional goals. This extension provides basic monitoring capabilities for Carma, which can subsequently be used for reasoning about control policies. A Carma component captures an agent operating in the system It consists of a process, that describes the agent’s behaviour, and of a store, that models its knowledge. The plug-in implements an appropriate high-level language, named the Carma Specification Language, that simplifies the creation of Carma models by providing rich syntactic constructs inspired by main stream programming languages

Capturing requirements

Individual constraints

Global goals

Case study

Translation tool

Conclusion