Carlos Parada works at the R&D branch of Portugal Telecom producing a working prototype that will deliver edge networking in tomorrows mobile environment.
Working for the research division Altice Lab for seventeen years’ you would expect Carlos to be addicted to the thrill of innovation and change of working environments. In reality, he says the small Portuguese town of ‘Aveiro is a small city, and people do not tend to move company very often!’
The prototype is being developed internally, under the umbrella H2020 Superfluidity project, a super-fluid, cloud-native, converged edge system being developed by operators and universities alike.
We caught up with Carlos to grab a glimpse at their fully functioning computing prototype hidden in the depths of Alice Labs.
Can you tell us about Altice Labs fully functioning edge computing solution?
It includes the MEC Host with a traffic offloading function to divert the traffic towards the applications instantiated at the edge. So far, we are working on the services (LOC, RNIS) and the management and orchestration layer.
How does Altice Labs approach ME management?
We believe that NFV MANO tools have similar functions. For this reason, we are using OSM for it (although it is in a very early stage of development). Other tools like Cloudify have more features and are more mature, but they are not NFV/MEC aligned.
What is an MEC orchestration layer? Why is it important?
In our view, the orchestration layer is extremely important in order to be able to easily manage and orchestrate the Apps in the edge. Other advancements in the future will be the automation of those decisions (scaling, instantiation, migration, healing, etc. ) based on monitoring, historical data (big data), etc.
What are some interesting results of MEC Proof of Concepts?
There are a few:
- Easy to integrate with existing mobile networks with minor changes, making the MEC introduction feasible and costless;
- High performance, assuring hundreds of Mb even using basic COTS hardware;
- Significant reduction of the delay, key for some Apps;
- Reduction of the core/edge traffic.
What challenges have there been during Implementation of the ME platform?
As we are developing as the standardization is progressing, sometimes it is hard to understand how to implement some features not yet clearly defined.
On the other side, there are some impacts in 3GPP (and now other) networks that are not clear yet.
Technological issues, like how to implement, what are the most suitable tools are also a typical software development problem.
MEC is a relatively new concept, where do you see it moving over the next few years?
MEC is a new concept and is closely tied with NFV, which has today a lot more traction than MEC. In our view, MEC will benefit by the adoption of the NFV technology in the networks access, namely C-RAN, vOLT, etc. For this purpose, telcos will need to move already IT resources to the edge, which can be used and shared by MEC, reducing the investment and making MEC cheaper and more attractive.
“I think the feasibility of MEC implementation on a real mobile network (Lab but using real production equipment),” he said. “It is easy to integrate and can perform well in a production environment. However, many things are still open like how to keep working, accounting, charging, policing, lawful interception, etc. when some traffic is not going to the core.”
Download the MEC Congress brochure to see more information about Carlos presentation.