Sarmad Sadik

Supply chain optimization towards personalizing web services

Personalization, which has the ultimate goal of satisfying userpsilas requests, can be perceived in terms of QoS measurement. As one of the means for the success of semantics Web, many techniques have been effectively used in modeling and developing Web service personalization. However, most of these methodologies relied heavily on detailed implicit and explicit information supply by users during initial and subsequent interactions with the systems. We propose in this paper a novel approach using the supply chain management (SCM) technique in personalizing Web services as against the conventional notion of applying SCM only to product manufacturing. Our user-model based framework uses multi-agent system (MAS) components in taking requests from users and working towards their satisfaction including seeking for additional information outside the system as the need arises. Only basic stereotype information furnished by potential users at initial contact is required for personalization during subsequent interactions with the system. The system is adaptive and aimed at high quality autonomous information services where users are successfully presented preferred Web services with minimum information request.

Using Honey Bee Teamwork Strategy in Software Agents

Teamwork among software agents is an active research area in the agent community. Researchers are working to define efficient teamwork architecture but a limited progress has been made towards efficient task execution mechanisms by group of agents in collaboration and coordination with each other. We have developed a teamwork architecture to enhance the performance and task execution efficiency of software agents. This architecture has been inferred from Honey Bee teamwork strategies and a short comparison has been made between the similarities of Honey Bee and Agents teamwork. We have also classified and analyzed two different teamwork approaches depending upon the type of collaboration among agents using case study approach

Towards modeling complex wireless sensor networks using agents and networks: A systematic approach

Wireless Sensor Networks (WSN) are a fast-emerging area of interest in the domain of large-scale communication networks. However, design of novel WSN applications requires first developing models and performing extensive simulation. Modern large-scale WSNs are expected to be both dynamic as well as random and often spread over a large scale. However, this problem has not previously been addressed in literature. In our previous works, we have demonstrated how Agent-based modeling may be used to effectively model various types of Complex Adaptive Systems such as, but not limited to, Self-organizing Communication infrastructures. In this paper, we present first steps towards providing a comprehensive set of guidelines and tutorial for developing deployment models of WSNs besides modeling routing algorithms using agent-based modeling. Our simulation results demonstrate the effectiveness and ease of use coupled with a short learning curve involved in developing agent-based models of complex WSN applications.

Modeling high assurance agent-based Earthquake Management System using formal techniques

Disaster management systems are complex applications due to their distributed and decentralized nature. Various components execute in parallel with high need of coordination with each other. In such applications, interaction and communication issues are difficult to model and implement. In this paper, we have proposed agent-based Earthquake Management System (EMS) which is modeled and analyzed using formal approach. Traditionally, such systems undergo through various transformations starting from requirement models and specification to analysis, design and implementation. A variety of formal approaches are available to specify systems for analyzing their structure and behavior; however, there are certain limitations in using these techniques due to their expressiveness and behavior requirements. We have adopted combination of Pi-calculus and Pi-ADL formal languages to model EMS from analysis to design. The formal approach helps to enhance reliability and flexibility of the system by reducing the redundant information. It reduces chances of errors by explicitly mentioning working flow of information. Additionally, a prototype application is presented as proof of concept in EMS context. We have also evaluated our formal specification by using ArchWare and ABC tools; also, comparison of prototype application with major existing techniques is highlighted.

Formal model of agent-based supply chain framework in a virtual enterprise for reliability

The virtual enterprises are models for distributed manufacturing where activities ranging from product design to actual production are carried out amongst geographically dispersed entities. Communication, collaboration and competition are some of the emerging issues in this form of production. To address these challenges in such distributed systems, we adopt the use of software agents, which are intelligent, collaborative, reactive and mobile entities. Although there is significant literature on implementation of virtual enterprises, there are fewer contributions that address flexibility and reliability of such systems. In this paper, we present formal model of agent-based supply chain management (SCM) in a virtual enterprise involved in product manufacturing. Our model is presented at different levels of abstraction and formalism. A decision tree is used to capture the logic of production planning based on the requirement specification. This is subsequently transformed to a declarative programming language for systempsilas instantiation. The pi-calculus notation is adopted for lower level modeling of the SCM. We propose a model which is useful for the development of flexible software systems and support of virtual enterprises and distributed production. Also, the proposed framework is formalized to enhance the reliability of system specifications.

Honey bee teamwork architecture in multi-agent systems

Teamwork among software agents is an emerging research area in the agent community as researchers are incorporating new architectures in multi-agent systems to facilitate coordination and collaboration among agents. We have proposed an efficient teamwork strategy for software agents which are deployed to work on distributed machines. This architecture is conceptualized from Honey Bee teamwork strategy and named after it as Honey Bee teamwork architecture. Also an abstract mapping is done between the honey bee and software agents working mechanisms. We have classified and discussed the proposed approach in context of two case studies and evaluation is made to exhibit its efficient behavior.

Policy Based Migration of Mobile Agents in Disaster Management Systems

Mobile agents have the potential to change the way distributed applications are developed and deployed. Researchers are actively working to define efficient mobility strategy for mobile agents which are deployed on distributed machines. In this paper, we have presented and analyzed policy based migration approach in the domain of disaster management systems. The operations which are designed for target machines are migrated with mobile agents with more emphasis on reusability concept in code. We have used a case study of deploying mobile agents in disaster management systems using efficient distributed information searching and retrieval technique

A Formal Approach for Design of Agent Based Earthquake Management System (EMS)

Earthquake is a natural phenomenon which has resisted the major efforts to tackle it. The major problems in earthquake management are inefficient communication, complex interaction as well as lack of coordination and prompt service delivery in advent of earthquake. Software agent based systems are more suitable alternative to the available technologies. In this paper, a formal approach is used in the designing and modeling of an agent-based earthquake management system using pi-calculus. Pi-calculus is justified for this work because of its capability to model concurrent and parallel systems. This approach facilitates in removing redundancy in information flow, low chances of errors in system development as well as understanding the execution paths for generating test cases. The services from the system are expected to be more reliable, autonomous, and adaptive in performing timely relief operations.

Policy based Ontology Framework for Mobile Agents

Policy based ontology work is gaining wide attention in semantics related research community. Mobile agents using policy based framework exhibit more adaptive and flexible behavior. Policies which are rule based statements consisting of conditions and actions have some limitations such as interoperability, policy definitions, as well as representation. We have presented the policy based ontology infrastructure for efficient runtime operations of mobile agents and policy related issues. The proposed architecture is illustrated with example scenario in domain of earthquake management system.

Cloud based semantic middleware for data storage on peer to peer network

Cloud computing is a promising business and computing model. The cloud storage is an essential and most used service, which provides stable and massive data storage space to end users. In this research work, new private cloud storage architecture is proposed with decentralized indexing which is based on peer to peer network. The proposed cloud architecture has semantic middleware for data storage. It is designed for a cloud based environment where client interacts with cloud which is comprised of peers. The private cloud will be available to clients through internet. Each peer in the network contributes its resources to form private cloud. The client can store whole file without dividing into chunks. The various modules in middleware performed different tasks to store and retrieve files from cloud. The ontology is created in Protégé-OWL to make it reusable and ease of management. The proposed architecture is being implemented using latest Java NetBeans framework with Jena API for ontology management.