The i2CAT Foundation participated in AGA, a project that applied Machine Learning technology to the problem of congestion management of data networks and brought it to a Telco cloud. The i2CAT Foundation was responsible for the Telco cloud effort.
Bequant, a company based in Madrid, launched the experimental project AGA, a technological transfer initiative supported by the “Centro para el Desarrollo Tecnológico Industrial (CDTI)” through the “Proyectos I+D de Transferencia Cervera” program. The i2CAT Foundation leveraged its expertise in Virtualized Infrastructure Management technologies (VIM) and led the Telco cloud part of the project.
The Telco cloud objective of AGA (Automatic Learning for Bandwidth Management) was to build a VIM for the deployment of the congestion manager fulfilling Telco cloud requirements, in particular NFV. The implemented Cloud is based on the well-known platforms, OpenStack and Kubernetes. Both enable the seamless migration from conventional hardware-based services to virtualized-based services. This can be achieved by enabling hardware virtualization. Indeed, AGA’s innovation originates from the use of novel features and tools such as Software Defined Networking (SDN), Enhanced Platform Awareness Optimization (EPA), and Open Source Mano (OSM), which can guarantee that latency is minimized and computing performance is maximized, as well as ease the deployment and orchestration of virtualized services.
Bequant trusted in the i2CAT Foundation to lead technically the Telco cloud part of the project. The research center evaluated the most suitable solutions regarding virtualization and containerization by carrying out performance tests that validated the feasibility of running Bequant services in a Cloud platform. To this end, i2CAT developed the necessary software to successfully orchestrate the deployment of a prototype of Bequant services in a Cloud environment.
This innovative Cloud platform opens the door to an autonomous and automated virtual infrastructure that reduces the costs related to hardware equipment as well as getting the most out of the required hardware.