Edge Computing Evolution

The work brings virtualization and disaggregation in different segments (home, edge, datacentre, NTN ground and on-board flying nodes), under the scope of multi-access edge computing concepts. The scope is to address functional placement and optimised computing distribution capabilities as a function of requirements emerging from ultra-low latency, ultra-high-capacity immersive applications. Additionally, it may include:

• AI and distributed security to protect transmitted data at the edge computing servers based on ledger or other technologies may be considered.
• The development of the data and control plane techniques required to realize such a multi-edge architecture.
• New IoT device management techniques as needed to operate over distributed architectures for IoT systems based on an open device management ecosystem.
• Novel programming models and engineering practices (e.g. split-computing), preferably applicable to open software environments, enabling the flexible distribution and migration of computation tasks, both horizontally among peer devices, and vertically along the IoT-edge-cloud continuum to enable economic sectors exploiting at the best the potential of the edge computing.
• Optimal deployment of the required data plane paths and control plane elements across a set of distributed physical edge nodes.
• Flexible hardware platforms, and/or programming abstractions (covering individual and aggregate resources) to achieve the benefits for agile, simplified yet automated composition and management of resources, possibly separated in “islands”.
• Models for devices interacting with the physical world (sensors and actuators).
• Demonstrate the provision of high-quality services (including reliable Operations Support System (OSS) mechanisms) while executing a very precise-latency or capacity control over massively distributed resources.
• Provide the necessary openness for making edge computing a “service innovation platform” running their own vertical-centric functions with optimised slice distribution and management.

Related work includes data exploitation for easy generation of big data pipelines, supporting increasingly complex/intelligent data processing techniques. To explore the wealth of data expected in new services, the project will need to provide a common, standard dataspace, to be openly offered for future research activities and paving the way towards intelligent infrastructure and service management, including sustainability aspects.