An adaptive user-centric IoT service composition framework

Rawan Sanyour, Manal Abdullah, Salha Abdullah

Users need to utilize several services to efficiently fulfill their requirements. These required services can be integrated to form service compositions by which value-added services can be created to meet the diverse users' requirements. However, in the dynamic IoT environment, creating on demand services driven by end users is a challengeable task. Personalizing service delivery through an adaptive, on- demand integration of available services requires the support of user-centric service composition approaches to spontaneously deliver the required functionalities. This research aspires to personalize IoT service delivery according to the evolve changing of end users' requirements. Motivated by this aspiration, an adaptive user-centric IoT service composition framework is proposed. It allows the user to discover, select and interconnect services on demand at the runtime. The framework provides a flexible and multi-step interaction between the user and the system. It consists of four modules: service discovery, service selection, service composition and service execution. The suggested services list will be filtered based on some selection criteria such as user's profile and Quality of Experience (QoE), Quality of Service (QoS) parameters and environmental context. Filtering services according to these factors can be considered as a Multi objective Optimization Problem MOP with constraints. Optimization techniques such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) can be used to find a solution for this type of optimization problems. Several scenarios in different application domains would be conducted to evaluate the applicability and performance of the proposed framework. The applicability will be tested through evaluating the ratio of successful compositions as the context and requirements changing. The performance will be evaluated by measuring the changes in computation time as the number of the candidate services increase. Several simulation environment platforms such as SimpleSoft, NetSim and IoTIFY could be utilized to test and evaluate the performance of the proposed framework. 

Advanced Studies: Euro-Tbilisi Mathematical Journal, Vol. 16,  supplement issue 2 (2023), pp. 135-158