Abstract— This paper proposes a multi-coordination approach for the design of mobile agent interactions. The approach is founded on the multi-coordination concept, which is a synergic exploitation of multiple coordination models which best fit interaction requirements. In particular, the proposed approach is based on two steps: (i) candidate design solutions are defined through a procedure which allows to identify the most effective coordination models for a given mobile agent interaction scenario; (ii) the defined candidate design solutions are quantitatively evaluated through a discrete-event simulation framework which allows for an easy evaluation of mobile agent interaction scenarios in terms of ad-hoc defined performance indices. Index Terms— Agent Interaction Design, Mobile Agents, Multi- Coordination, Performance Evaluation. I. INTRODUCTION ode mobility paradigms have been introduced to support the design and the implementation of flexible, dynamic and reconfigurable distributed applications in terms of software components which are not confined in a single run- time context for their entire lifecycle but can migrate autonomously or on-demand across different contexts [1]. Among them, the most fascinating paradigm is represented by the mobile agents, executing software components capable of autonomous migration by retaining code, data and execution state. Although it is advocated that the exploitation of mobile agents can provide many benefits [2], they have introduced specific and not yet fully addressed issues that actually limit their advertised wide-spread use [3]. An interesting issue concerning with the design of mobile agent interactions regards how to clearly identify which agents will be interacting and how their interactions can be modeled. To deal G. Fortino is with the Department of Electronics, Informatics and Systems (DEIS), University of Calabria, Rende (CS), 87036 Italy. (corresponding author; phone: +39.0984.494063; fax: +39.0984.494713; e-mail: g.fortino@unical.it). A. Garro is with the Department of Electronics, Informatics and Systems (DEIS), University of Calabria, Rende (CS), 87036 Italy. (e-mail: garro @unical.it). S. Mascillaro is with the Department of Electronics, Informatics and Systems (DEIS), University of Calabria, Rende (CS), 87036 Italy. (e-mail: samuele.mascillaro@deis.unical.it). W. Russo is with the Department of Electronics, Informatics and Systems (DEIS), University of Calabria, Rende (CS), 87036 Italy. (e-mail: w.russo @unical.it). with mobile agent interactions, communication paradigms and mechanisms as well as coordination models and architectures for non mobile software components have been enhanced to be mobility-aware (message-passing, tuple space, publish/subscribe, etc) and new ones have been purposely defined for logical and physical mobility (meeting, blackboard, shared transiently tuple spaces, reactive tuples) [4, 5, 6, 7, 8, 9]. Although some mobile agent frameworks already offer several mechanisms based on the aforementioned communication/coordination paradigms and architectures, in the current practice mobile agent interactions are designed on the basis of a single paradigm which is mainly based on message passing or, in some application domains, on tuple spaces [4]. As single model based communication/coordination might not be effective for satisfying all needs of mobile agent interactions in all possible application scenarios, the exploitation of multiple communication/coordination paradigms, namely Multi- Coordination, can enhance design effectiveness, improve efficiency, and enable adaptability in dynamic and heterogeneous computing environments [10]. In particular, Multi-Coordination allows agents to choose among a variety of different communication/coordination paradigms which best fit mobile agent interaction needs. Moreover, although several design patterns have been proposed for driving the design of mobile agent interactions [11, 12] and programmable coordination models and related frameworks (e.g. TuCSoN [13]) are now available, systematic methods for supporting the development of mobile agent interactions which specifically take into account an integrated exploitation of multiple coordination models are surprisingly still lacking. To overcome this lack, this paper proposes a multi- coordination approach for the design and evaluation of mobile agent interactions. The design is based on a procedure which uses suitable agent interaction patterns to fulfill agent coordination requirements. In particular, interaction patterns are first characterized by appositely defined parameters and associated to specific coordination models according to such parameters; then, the most appropriate coordination model is selected for implementing a given interaction pattern so providing a design solution for the related coordination requirement. The evaluation is based on a discrete-event simulation framework which allows to evaluate the designed solutions in Using multi-coordination for the design of mobile agent interactions Giancarlo Fortino, Alfredo Garro, Samuele Mascillaro, and Wilma Russo C