[Paper Review] Towards 5G Enabled Tactile Robotic Telesurgery

Qi Zhang, Jianhui Liu+1 more|arXiv (Cornell University)|Mar 9, 2018

Wireless Body Area NetworksEngineering10 references87 citations

TL;DR

The paper analyzes QoS requirements for multi-modal data in robotic telesurgery and proposes a 5G-enabled edge-cloud architecture with MEC, network slicing, and AI to meet ultra-low latency and reliability needs for haptic feedback and 3D video.

ABSTRACT

Robotic telesurgery has a potential to provide extreme and urgent health care services and bring unprecedented opportunities to deliver highly specialized skills globally. It has a significant societal impact and is regarded as one of the appealing use cases of Tactile Internet and 5G applications. However, the performance of robotic telesurgery largely depends on the network performance in terms of latency, jitter and packet loss, especially when telesurgical system is equipped with haptic feedback. This imposes significant challenges to design a reliable and secure but cost-effective communication solution. This article aims to give a better understanding of the characteristics of robotic telesurgical system, and the limiting factors, the possible telesurgery services and the communication quality of service (QoS) requirements of the multi-modal sensory data. Based on this, a viable network architecture enabled by the converged edge and core cloud is presented and the relevant research challenges, open issues and enabling technologies in the 5G communication system are discussed.

Motivation & Objective

• Identify limiting factors of telesurgical systems and the QoS needs of multi-modal sensory data.
• Categorize telesurgery services and outline a development roadmap.
• Propose a network architecture that converges edge and core cloud to support required QoS.
• Highlight enabling 5G technologies (MEC, network slicing, SDN, AI) to address challenges.
• Discuss open research issues and potential cost-effective, secure solutions.

Proposed method

• Survey telesurgery system architecture and QoS requirements for multi-modal data.
• Present a converged edge-cloud network architecture with D-RAN/C-RAN, SDN, and MEC.
• Outline data compression, scalable coding, and multiplexing approaches for sensory streams.
• Discuss enabling 5G technologies (PHY/air interface, MEC, network slicing, AI) to meet QoS.

Experimental results

Research questions

• RQ1 What are the QoS requirements (latency, jitter, packet loss, data rate) for the multi-modal sensory data in robotic telesurgery?
• RQ2 What network architecture can meet these QoS requirements in a 5G-enabled environment?
• RQ3 How can 5G technologies (MEC, network slicing, SDN, AI) address the challenges of tactile telesurgery?
• RQ4 What are the open research issues and potential security/cost barriers to wide deployment?

Key findings

• Multi-modal data require stringent latency, jitter, and reliability; typical targets include <150 ms latency for 2D/3D video and <10^-3 packet loss for video streams.
• Haptic data demand very low latency (<10 ms) and ultra-low loss (<10^-4) with data rates of 128-400 Kbps.
• A converged edge-cloud architecture with MEC and SDN-enabled network slicing can meet diverse QoS for different modalities.
• 5G PHY improvements (shorter TTIs, fragmented subframes) and MEC can reduce end-to-end latency and support real-time AI-driven assistance.
• Network slicing and SDN enable isolation, dynamic resource allocation, and potential cost reductions compared to private networks.