EPC over Cloud-Native Platform
As part of this blog, we wanted to address one key question
Should an Operator invest in implementing EPC over cloud platform?
Cloud platform for telecom is not up for debate anymore. Today, Tier-1 telecos across geographies are evaluating possibilities of moving to cloud platforms. In less than a year, we have gone from "How" to "When". However, most of these discussions are under the ambit of " Introduction of 5G infrastructure" with the presumption that 5GC will be parallel infrastructure and pose no hindrance to ongoing operations.
At Catenna, we see a major challenge with the above-stated logic. Availability and maturity of 5GC on cloud environment are at least a year away. Not to mention 3GPP Release-16 revisions which will only add to these delays. Then, each operator has challenges with upgrading existing infrastructure to support basic interoperability with EPC. Putting aside all the euphoria generated by 5G marketing campaigns, the reality is that most telecom vendors are still struggling to demonstrate end to end calls even in a controlled lab environment. So, how realistic is the expectation of 5GC over the cloud-native platform in Operator's environment in 2020?
Answer to availability of 5GC in some way also provides clues to our original question we wanted to address - " Should an Operator invest in implementing EPC over cloud platforms? "
Catenna thinks "EPC over cloud platform" is probably the best way forward and here are our reasons -
Build, Test and Deploy - In some sense, incremental growth is the cornerstone of every cloud-native application and in the current discussion, should be fundamental to new age telecom network. This incremental growth is almost impossible to achieve if we were trying to implement the whole network in one go (which some Operators are attempting today). In this scenario, EPC offers an opportunity to test waters with stepwise migration of EPC nodes with no dependency on 5G. Additionally, Operator can really see the incremental changes in platform in practice.
Reducing OPEX - with aging infrastructure and HW refresh around the corner, Operators have a hard choice to make. NFV was the short term answer but with the 5G bus arriving earlier than expected, NFV investments have vanished. If the choice is to extend current platforms or build cloud infrastructure, we think its a no-brainer.
Evolving EPC for next decade - While, the transition towards 5G will begun by 2020-21, EPC will play a major role in mobile connectivity for at least another decade. This is reason enough if the first two reasons were not convincing enough.
We also think that a node like PCRF could be one of the first products for such transformation projects considering the non-critical nature of the node.
We would love to hear what industry has to say !!