OPTIMIZING COMMIT RATE IN ADVANCED PERVASIVE SYSTEMS WITH TRANSACTIONAL SEMANTICS IN CONTEXT-AWARE SERVICE COMPOSITION

Abstract

Pervasive computing, characterized by dynamic environments, ad hoc networks, and resource-limited devices, poses challenges for the execution context of transactional services. Existing approaches overlook contextual information and user requirements, leading to suboptimal transactional properties. Context-aware computing, emerging in the 90s, aims to adjust system behavior based on perceived conditions, providing relevant information and services to users. However, current definitions of context-aware systems fail to encompass systems that only collect context for application use. To address these issues, this study proposes context-aware management of transactional services, considering user preferences and contextual requirements. For instance, a hotel booking transaction should consider the user's center of interests and the compensability requirement. Systems operating in context-aware environments must adapt to flexible and dynamic conditions, facing contextual requirements that affect transaction execution. Transactional properties, considered extra-functional properties, are described using languages like WS-Policy in Web Services. Preliminary studies have explored the determination of transactional behavior for composite services. Integration of transactional aspects in composite services necessitates specific characteristics from transactional processing and web services domains. This paper provides an overview of the challenges and requirements for context-aware management of transactional services, highlighting the need for context-awareness in transactional systems. The proposed approach aims to enhance transaction execution cost and successful commit rates by dynamically adjusting services based on contextual information and user preferences. Experimental evaluations and case studies will be conducted to validate the effectiveness and efficiency of the proposed approach