Evolving 3GPP Standards and Testing Challenges
3GPP specifications are evolving, revised versions of many of these specifications are produced up to four times a year following the quarterly TSG plenary meetings. It has always been challenging to ensure that the network cores are compliant with evolving 3GPP standards. Private LTE and 5G use cases add even more test scenarios on a repetitive basis. Compliance verification, therefore, becomes a humongous task for network engineers, developers and testers.
Let’s dig a little deeper into actually how big the problem is.
There are 70+ 3GPP call flows (attach, detach, DDN, Handover, Dedicated bearer, event triggers, etc.) in the core network. Each call flow carries around 20+ messages (CSR, DSR, CBR, DBR, UBR, Session Initiation, DNS Query, PFPC, Diameter, etc.) in and out & between core network components. Each of these call flow has around 400+ (TEID, EBI, FQDN, etc.) Informational Elements.
So if we consider a basic but complete call flow, UE attach has around 27+ messages exchanged between different core network components. Together, they have about 100+ IEs to be verified.
In these 3GPP call flows, the core components also generate system-specific, time-bound Informational elements. The core components need to pass some Informational Elements (QCI, IMSI, MSISDN, etc.) through different components. Adding to the depth, the call flow is not valid if the messages’ sequence does not happen in the specific order (UBR before CBR, DSR before CSR, etc.) as specified in 3GPP specifications. But there is a breadth too; some of the messages can be conditional, optional, and/or can occur in asynchronous order also (example: Modify Bearer request). Moreover, different core components play different roles and hence work on different protocols such as GTPv2, Diameter, DNS, and PFCP, etc.
Each network function has multiple installations and mobile users get handed over between these different installations. (Selection of SGW, PGW, etc.) As if this was not enough, there is a possibility of a local break out and home-routed scenarios for roaming users, where different components from different network cores need to have a dialogue with each other. E.g., cases for MVNO, MNO, Private LTE, etc.
Wi-Fi Calling adds a different dimension to the conversations happening between network core components too.
Here is a quick glance at various interactions within EPC and EPC peripherals for a simple basic call flow.
The efforts needed to assert all of these are mammoth work and demand a good deal of time and money. When the network core component gets updated for the evolving 3GPP standards, it introduces regression verification. This results in a cascading effect where a single unwanted change can result in a non-standard call flow and/or the network core may cease to function correctly.
So if we consider 30+ 3GPP complaint call flows and various scenarios, network engineers will have to test the flow and compliance for over 100,000 informational elements at each node. So, core network verification is a mountain-like activity which is time consuming, tedious, error-prone, requires weeks of effort again and again.
Automation in 3GPP Testing and Compliance.
If you consider the above scenarios, testing is becoming more and more complicated. As new use cases start rolling out with 5G implementation, the complexity will increase even more. Testing and optimization will have to be automated. GS Lab has developed an end-to-end Whitebox testing solution that can overcome this challenge. Most of the testing tools in EPC space are either traffic generators or perform black-box testing. The Whitebox testing tool at GS Lab called “Octopus” automates the complete EPC testing at each node.
Octopus is an end-to-end, completely automated 3GPP compliance tracker, which automates the unit, functional, regression, and compliance testing and reduces the EPC and vEPC testing time from days to minutes. Octopus features help you test all the required call flows across various components and scenarios. It even detects and pin-points all the IE failures at each connection point for compliance. It can work with different test traffic generators. It does not require any change in EPC code, infrastructure, or deployment. Octopus seamlessly works across various protocols and call flow scenarios. It also comes with a user-friendly dashboard to check all traced messages’ success or failure and their sequence. It is compatible with physical servers, VMs, or containers and supports multiple simulators. Octopus quickly compares the Old PCAP for 3GPP call flow, sequences, and timing validation along with many packets in DP.
Octopus helps companies bring down regression testing time and reduces dependency on niche resources. We have already automated 550 test cases for 15+ 3GPP procedures for one of our customers, where the total testing time is reduced from 18 business days to 5 hours.
As the standards keep evolving, the industry will have to keep a closer watch on how the networks are managed and adopt automation as much as possible.
Sagar Neve | Associate Manager – Engineering
Sagar Neve is an Associate Engineering Manager at GS Lab and has an experience of more than 17 years in IT. Apart from heading the GS Lab’s open source contribution for ePDG; he is leading efforts on GS Lab’s Octopus – a 3GPP Compliance Tracker. After working in areas like Telecom, Video On Demand, Storage arrays, Video Surveliances, QA, and Automations, Sagar is now working in the telecom domain since the past 2 years and is keen on understanding new trends in the telecom industry in 5G and Private LTE areas.