WP1 : Tactile Internet use cases requirements and testbed
The Tactile Internet addresses a wide range of use cases ranging from industrial control, skills training, remote medical diagnosis and surgery to entertainment and social communication. The remote environment can be real or virtual. The QoS requirements regarding latency, reliability, and throughput depend on the specific use case. The envisioned joint testbed is crucial to promote interactions between partners and to achieve meaningful crossdisciplinary advances and solutions from partners with different background and expertise. This WP serves as the basis for identifying requirements for the use cases and coordinating implementation, prototyping, and demonstration actions at the core of this training network. The testbed will be designed to support instructors around the world in running TI experiments and integrating the experiments and results into their lectures or laboratories in a plug-andplay manner. The testbed will provide well-defined interfaces for TI researchers, facilitating future extensions and improvements. The testbed will be available following EU rules and recommendations for open data.
Haptic communication focuses on the local processing and controlling of haptic data at the master and slave devices, which optimizes the transmission strategy for networked teleoperation while ensuring system stability and high fidelity. Due to the limited network resources, it is essential to investigate signal representations and develop corresponding compression schemes that exploit the properties and conceptual limitations of haptic information. Teleoperation performance is highly affected by the haptic communication schemes and the network QoS. In this WP, we aim to develop kinaesthetic and tactile codecs by investigating adaptive haptic processing and control frameworks. This not only delivers real-time haptic communication over the TI, but also studies a wider range of latency and reliability characteristics, e.g. mimicking long-distance TI8 .
This WP aims to provide agile communication and networking solutions for TI services in the Network Domain. The communication between the operator and teleoperator can cross radio access network (RAN), transport network and core network in 5G and next generation communication system. Therefore, an E2E multidomain orchestration with optimal 5G/B5G network configuration is vital to establish TI services offering the required QoS/QoE/QoT. In addition, as TI services require communication of multi-modal sensor data (visual/auditory/haptic), which have different requirements regarding sampling rate, transmission rate, latency, reliability and others10, it requires the network to support at least uRLLC traffic (e.g. haptic data) and eMBB traffic (e.g. 3D video stream). From a radio resource management and information theory perspective, multiplexing of large video packets (low update rate) with small haptic packets (high update rate) requires careful consideration of their characteristics. Furthermore, the different data use different error control schemes and congestion control methods. Hence, the significant communication and networking challenges are how to support highly demanding and diverse information flows and to meet their processing and control needs, while keeping the network operating at optimal points in the latency, resilience and throughput trade-off hyperspace. Fortunately, TI services often provide a certain amount of flexibility on how the service functions can be configured and operated11. For example, TI services can trade different haptic operation modes for dynamics or precision against robustness towards network latency. That opens the opportunity for a collaborative network and service configurations for TI services, which can be enabled by 5G and beyond, being a flexible joint networking and computing platform.
This WP aims to design and develop advanced functionalities that support the TI services, referred to as Support Engine (SE) in the IEEE P1918.1 TI architecture, by leveraging edge computing and artificial intelligence (AI) (i.e., Edge Intelligence). Edge computing provides both computing and storage resources close to the edge of the network or, more precisely, in the proximity of the operator and teleoperator. The Edge Intelligence Engine promises to overcome critical challenges in TI, thereby improving the performance of the operator and teleoperator and meeting the latency and reliability requirements of E2E communication.
