METAAGENTS: A FRAMEWORK FOR INTELLIGENT DISTRIBUTED SYSTEMS

AN INTELLIGENT DISTRIBUTED SYSTEM PRESENTS SIGNIFICANT INTEGRATION AND CONCURRENCY CONTROL PROBLEMS. HOWEVER, THE TOOLS THAT ARE CURRENTLY AVAILABLE FOR DISTRIBUTED SYSTEMS DEVELOPMENT IN THE AREAS OF HETEROGENEOUS SYSTEM INTEGRATION AND MULTIPLE PROCESS CONTROL ARE INADEQUATE. AN APPLICATION THAT REQUIRES MULTIPLE PROCESS COORDINATION AND CONTROL REQUIRES INTRACTABLY COMPLEX MANAGEMENT OF PAIRED PROCESS SESSIONS USING THESE METHODS. THIS PROJECT FOCUSES ON THE DEVELOPMENT OF A FRAMEWORK THAT PROVIDES BOTH HIGH-LEVEL REPRESENTATION AND ALTERNATIVE DISTRIBUTED CONTROL MODELS BY WHICH FAULT-TOLERANT INTELLIGENT SYSTEMS CAN ACCESS INFORMATION AND RESOURCES IN A DISTRIBUTED HETEROGENEOUS ENVIRONMENT. THE BASIS OF THE FRAMEWORK INTERFACE WILL BE AN EXPLICIT, OBJECT-ORIENTED REPRESENTATION OF BOTH THE DATA TYPES AND OPERATIONS OF THE DISTRIBUTED SYSTEM MODELS, SERVICES, AND RESOURCES. THIS KEY INNOVATION IS REQUIRED FOR HIGHLY ADAPTIVE, INTELLIGENT INFORMATION SYSTEMS AS THE DISTRIBUTED ENVIRONMENT CAN BE REPRESENTED IN TERMS OF EXPLICIT, CHANGEABLE DESCRIPTIONS. ADDITIONAL KEY COMPONENTS OF THIS FRAMEWORK ARE IN THE AREA OF DISTRIBUTED CONTROL MODELS: REDUCTION OF RESOURCE CONTENTION THROUGH GROUP-BASEDREPLICATED SERVICES; DISTRIBUTED, RELIABLE, GROUP-BASED RECONFIGURATION, AND RESOURCE ALLOCATION MECHANISMS; AND UNIFORM INTERFACE ACROSS HETEROGENEOUS ENVIRONMENTS. AN INTELLIGENT DISTRIBUTED SYSTEM PRESENTS SIGNIFICANT INTEGRATION AND CONCURRENCY CONTROL PROBLEMS. HOWEVER, THE TOOLS THAT ARE CURRENTLY AVAILABLE FOR DISTRIBUTED SYSTEMS DEVELOPMENT IN THE AREAS OF HETEROGENEOUS SYSTEM INTEGRATION AND MULTIPLE PROCESS CONTROL ARE INADEQUATE. AN APPLICATION THAT REQUIRES MULTIPLE PROCESS COORDINATION AND CONTROL REQUIRES INTRACTABLY COMPLEX MANAGEMENT OF PAIRED PROCESS SESSIONS USING THESE METHODS. THIS PROJECT FOCUSES ON THE DEVELOPMENT OF A FRAMEWORK THAT PROVIDES BOTH HIGH-LEVEL REPRESENTATION AND ALTERNATIVE DISTRIBUTED CONTROL MODELS BY WHICH FAULT-TOLERANT INTELLIGENT SYSTEMS CAN ACCESS INFORMATION AND RESOURCES IN A DISTRIBUTED HETEROGENEOUS ENVIRONMENT. THE BASIS OF THE FRAMEWORK INTERFACE WILL BE AN EXPLICIT, OBJECT-ORIENTED REPRESENTATION OF BOTH THE DATA TYPES AND OPERATIONS OF THE DISTRIBUTED SYSTEM MODELS, SERVICES, AND RESOURCES. THIS KEY INNOVATION IS REQUIRED FOR HIGHLY ADAPTIVE, INTELLIGENT INFORMATION SYSTEMS AS THE DISTRIBUTED ENVIRONMENT CAN BE REPRESENTED IN TERMS OF EXPLICIT, CHANGEABLE DESCRIPTIONS. ADDITIONAL KEY COMPONENTS OF THIS FRAMEWORK ARE IN THE AREA OF DISTRIBUTED CONTROL MODELS: REDUCTION OF RESOURCE CONTENTION THROUGH GROUP-BASEDREPLICATED SERVICES; DISTRIBUTED, RELIABLE, GROUP-BASED RECONFIGURATION, AND RESOURCE ALLOCATION MECHANISMS; AND UNIFORM INTERFACE ACROSS HETEROGENEOUS ENVIRONMENTS.