User Preference Aware Task Coordination and Proactive Bandwidth Allocation in a FiWi-Based Human–Agent–Robot Teamwork Ecosystem

Cooperative human–agent–robot teamwork (HART) provides enormous opportunities for present-day human users to orchestrate their real-time tasks in a coordinated fashion. However, given human users’ different preferences for real-time HART task execution, e.g., lower delay and monetary cost, the selection of proper task coordination services has emerged as an important research problem by taking dynamically changing cloud agent/robot resources, network bandwidth utilization, as well as delay-sensitive and delay-tolerant HART task properties into account. To cope with these challenges, in this paper, we explore the synergy between caching, computation, and communications for achieving cost-effective HART task execution. To exploit the locality of different HART-centric tasks and local/non-local cloud agent/robot resources for different HART-centric task execution, we consider integrated fiber-wireless (FiWi) enhanced networks with computation task offloading as well as fiber backhaul sharing and WiFi offloading capabilities. More precisely, to minimize task execution delay and monetary cost, we propose a user preference aware HART task coordination framework that selects the appropriate dedicated or non-dedicated robot and cloud agent for given caching and computing HART task execution requirements. Further, to cope with varying bandwidth resources, we propose a proactive bandwidth allocation policy for the execution of both delay-sensitive and delay-tolerant HART tasks execution across FiWi enhanced network infrastructures. We evaluate the performance of our proposed preference aware task offloading scheme and compare it to various baseline schemes in terms of different key performance indicators, including the task execution time and monetary cost saving ratio, communication to computation ratio, and offloading gain overhead ratio. Our findings indicate that the proposed delay cost saving policy exhibits a 27% higher task execution time saving ratio and a 48% lower monetary cost saving ratio than the proposed monetary cost saving policy in a typical scenario.