Towards Immersive Tactile Internet Experiences: Low-Latency FiWi Enhanced Mobile Networks With Edge Intelligence [Invited]
Much of the research on latency in edge optical networks is focused on transport and the underlying infrastructure. In contrast, this paper is a detailed discourse of a burgeoning application that requires low-latency at the network edge, namely the Tactile Internet. It first defines the Tactile Internet, emphasizing the human-to-machine interactions that distinguish it from the Internet of Things. This is followed by three stand-alone sections that examine latency issues from different perspectives. First, a traffic model is proposed and verified for data generated by teleoperation (e.g., remote operation of machinery), a prime use-case of the Tactile Internet. Next, various heterogeneous fiber/wireless access-network architectural options and their impact on latency are discussed, including a qualitative analysis of bandwidth allocation strategies. Finally, an intelligent, and relevant, application of machine learning (ML) is proposed and tested: if data is delayed at the edge beyond acceptable latency thresholds, ML is used to generate a logical estimate of the missing data. Overall, this is a very complete paper, concluding with some thoughts on the relationship of the Tactile Internet to blockchain-based applications.