Collaborative Computing for Advanced Tactile Internet Human-to-Robot (H2R) Communications in Integrated FiWi Multirobot Infrastructures

TL;DR: An adaptive resource allocation model is introduced and an analytical framework is developed to evaluate the task allocation delay, energy consumption, and task response time for noncollaborative and collaborative task execution scheme across integrated fiber-wireless multirobot networks.

Abstract: With the emergence of the Tactile Internet and advent of remote-controlled robots, proper task allocation among robots has attracted significant attention to enable robotic applications and services based on the human-to-robot communications paradigm. However, limited computing, energy, and storage resources of robots may hinder the successful launch of such applications. Task offloading to collaborative nodes is a promising approach to improve the task execution time and energy efficiency of robots. In this paper, we investigate a proper task allocation strategy by combining suitable host robot selection and computation task offloading onto collaborative nodes. We exploit conventional cloud, decentralized cloudlets, and neighboring robots as collaborative nodes for computation offloading in support of a host robot’s task execution. More specifically, our proposed task allocation policy selects a suitable robot based on several key parameters, including robot availability, remaining energy, and task execution time. Furthermore, our proposed computation offloading strategy examines the suitability of collaborative nodes in terms of task response time and energy consumption and then chooses an appropriate collaborative node to conduct the requested computation. We introduce an adaptive resource allocation model and develop an analytical framework to evaluate the task allocation delay, energy consumption, and task response time for noncollaborative and collaborative task execution scheme across integrated fiber-wireless multirobot networks. The results show that the proposed collaborative task execution scheme outperforms the noncollaborative scheme in terms of task response time and energy consumption efficiency.