Multilevel agent negotiation on service bindings for efficient multi-tasking in IoT environments

Abstract

In Internet-of-Things (IoT) environments, various IoT resources are utilized to provide the services necessary to accomplish users’ tasks. Therefore, IoT resources must be effectively and efficiently allocated so that service providers can successfully generate their service capabilities utilizing IoT resources. Additionally, conflicting resource allocation must be avoided for services supporting multiple user tasks. Furthermore, resource allocation and conflict resolution must be performed dynamically because users perform their tasks spontaneously. To address these issues, we propose an agent-based service-binding approach for efficient resource allocation and conflict resolution through negotiations among resources, service providers, and task agents. Negotiation occurs when a task agent attempts to use a service that is currently occupied by another task agent. The task agent that uses the service is determined in the negotiation process by considering the transfer fee based on the usage price of the service and the availability of a substitute service. Negotiations can occur at the task-service or service-resource levels to support flexible service binding. To evaluate the effectiveness of our approach, we compared its success rate with that of traditional resource allocation methods in terms of allocating resources successfully for multiple services. We also analyze the execution time to measure the overhead incurred by multilevel negotiation, which is an essential component of our approach. The proposed approach showed an average success rate of 79% which is 1.17%p lower than multi-objective optimization algorithms (MOOA), but the running time of the proposed approach was 98.3 s which was 56% of the running time of MOOA.