Trace-Driven Haptic Traffic Characterization for Tactile Internet Performance Evaluation

The Tactile Internet is expected to be the central part of next-generation mobile networks. The future Tactile Internet will not only allow for the transmission of traditional triple-play traffic (i.e., audio, video, and data), but it will also enable the transmission of sense of touch. Substantially improved human-to-machine/robot interaction will enable humans to touch and physically interact with remote objects in real-time. Despite the growing interest in the Tactile Internet, the characteristics of real haptic traffic are still unknown, especially at the packet level. In this paper, we model Tactile Internet traffic by means of haptic traces, taking different teleoperator robots into account, and determine the best fitting packet interarrival time distributions while exploiting the inherent human-in-the-loop nature of haptic interaction to reduce the packet rate via perceptual coding.