Steffen Kern

MobiSoft: an agent-based middleware for social-mobile applications

We present an agent-based middleware for social-mobile applications, which has been developed as part of an ongoing linkage project One aspect of the MobiSoft project is the vision of facilitating, augmenting, and promoting human social interaction by electronic personal assistants during face-to-face encounters Possible areas of social mobile applications include the establishment of groups or communities based on shared interests or goals, the exchange of information such as personal profiles, news, private sales, or any kind of recommendations, and the preselection of possible communication partners in social networks We outline the decentralized peer-to-peer based architecture and present techniques for information representation using semantically rich languages based on existing standards We describe, how mobile agents are facilitated as user representatives and intelligent information carriers in mobile ad-hoc networks and present a first prototype of a social-mobile application.

Tracy: An Extensible Plugin-Oriented Software Architecture for Mobile Agent Toolkits

In this chapter we propose a software architecture for mobile agent toolkits and describe our Tracy toolkit as a reference implementation of this architecture. Agent toolkits mainly consist of a software system that forms an agency, which is responsible to host mobile and stationary software agents. In contrast to most architectures developed so far, which already define a large set of services for agent migration, communication, and security, we propose to employ a kernel-based approach. The kernel only provides fundamental concepts common to all agent toolkits and abstracts from any of these services. In particular, although Tracy was developed as a mobile agent toolkit, its kernel abstracts from all issues related to agent mobility, delegating this to an optional service implementation. This makes it possible to replace Tracy’s migration service with another implementation and even to have two different migration services in parallel. Service implementations are developed as plugins that can be started and stopped during run-time. We have already developed almost a dozen plugins for agent migration, communication, authentication and authorization, and security solutions, only to name a few. We believe that this architecture is a useful foundation for research on agentrelated topics as it allows research groups to implement their own results as a service which can be used by other groups running an agent system based on the same architecture.

Class Splitting as a Method to Reduce Migration Overhead of Mobile Agents

Mobile agents were introduced as a new design paradigm for distributed systems to reduce network traffic as compared to client-server based approaches simply by moving code close to the data instead of moving large amount of data to the client. Although this thesis has been proved in many application scenarios, it was also shown that the performance of mobile agents suffers from too simple migration strategies in many other scenarios. This has lead to the development of a new migration protocol, named Kalong, which provides fine-grained transmission of code and data instead of viewing a mobile agent as a single transmission unit. In this paper we report on first results of the application of Kalong to improve the performance of mobile agents by splitting the code of mobile agents. First results show that by using this technique the number of bytes which have to be transferred can be reduced significantly.

MobiSoft: Networked Personal Assistants for Mobile Users in Everyday Life

This paper provides an overview of the MobiSoft project, its ideas and aims as well as the achieved results. In MobiSoft, we applied mobile software agents to support humans in their mobile everyday life. We developed a generic application framework that can be customized to fit into completely different scenarios ranging from industry use cases to social human interactions during leisure time. We describe this framework as well as several prototypes that demonstrate its general applicability. This paper also delivers first results of a survey at the university campus, that tried to capture user interest in personal assistants and mobile applications in general.

Assistant-based mobile supply chain management

This paper describes our approach for changing the way supply chain management is performed today. We aim to support human interactors with software assistants which will perform most of the tedious tasks like negotiating on new contracts, handling and altering production workflows or managing the stock. Additionally, our assistants will have the ability to move from one machine to another allowing a more flexible transfer of information and new ways for handling specific tasks. The owner of those assistants can be informed at any time by allowing its representative to move back to his notebook, PDA or mobile phone and presenting some results. Or the assistant may ask for advice in case of an error