A5EdgeInfra

A5EdgeInfra

The A5GEdgeInfra project aims to investigate the technologies that will enable the coverage of scenarios for which current 5G and Edge Computing deployments are not yet prepared. The communication needs are growing and highly variable in both range and duration. From plants that require high-quality video retrieval 24 hours a day, to remote environments where computing capacity is needed temporarily for a specific activity. Therefore, the creation of a portable 5G infrastructure with Edge Computing capabilities has been considered, which will provide a solution to multiple use cases. Due to the portability characteristics of this solution, it is essential that it can be deployed very quickly, offering an agile deployment planning system that is as autonomous as possible in terms of network and electrical resources.

In the most latency-sensitive use cases, it will be necessary to deploy the edge nodes at the 5G base stations themselves. The main issue with these locations is the space and electrical consumption limitations when installing a large number of servers. This is why the use of new lightweight virtualization technologies, which allow the deployment of more applications using the same physical resources, will be a key factor.

To offer a unified solution for all these environments, the use of lightweight virtualization technologies will be investigated with the following characteristics:

Allowing the use of serverless architectures, thereby reducing the resources consumed by applications in standby, allowing for a greater number of applications ready to be executed.

• Enabling applications to run on edge nodes with different CPU architectures without requiring code modifications or recompiling the application for each architecture.
• Reducing the size of the applications that need to be transmitted to the edge locations, thus reducing the total download time and deployment time of new applications.
• Reducing the execution time of edge applications.
• Allowing the use of serverless architectures, thereby reducing the resources consumed by applications in standby, allowing for a greater number of applications ready to be executed.

---