While collecting and aggregating OSS data to gain network performance insights is critical, what about monitoring the control plane? Understanding system health is one thing, but OSS data doesn’t intrinsically provide oversight of service delivery. Control plane insight is needed to understand and spot discrepancies in real-time that can affect what your end customers experience. And as 5G arrives, monitoring the control plane is set to become a critical function.
Your OSS and the data that can be captured from the elements within it can do much to identify and help you to address issues of network performance. But that doesn’t necessarily help you deliver services over the network seamlessly, meeting end-user expectations and providing the desired quality of experience.
If you want to do the latter, it’s the control plane you need to monitor because, for example, if messages sent to establish a service are faulty or incorrectly formatted, your OSS isn’t going to alert you to the problem. As 5G and with it service oriented architectures arrive, ignorance of issues like that definitely won’t result in bliss.
Thus, it’s safe to say that without control plane monitoring, MNOs will have a growing problem, albeit one that can be by passively accessing control plane data to enable them to understand the discrepancies and deviations that affect service delivery. With this in place, the operator benefits; so, if indirectly, does the end customer. Which makes control plane monitoring a win-win proposition. Despite this recognition, many MNOs rely on OSS data, rather than dedicated network monitoring systems that can capture, process and filter control plane signalling.
Before we get to monitoring in any more detail, let’s backtrack and define what we mean by the “control plane” In general terms, an efficient 3GPP specified mobile network provides terminals with access to services (voice, text, data, etc.). Part of the process of so doing requires a control plane, aspects of which are manifested in both the access (RAN) network and the core network (for instance, 3G).
The ‘control plane’ is generally where all signalling needed to support the functions in the network system that establish and maintain the user plane are handled, principally signalling which in this context means the exchange of information required to enable (but not to provide) the end-to-end communication service itself.
As such, the control plane is essentially a forwarding path to exchange information necessary for the operation of a service. Bearing this in mind, the control plane must be efficient, scalable, reliable and suited to the needs of mobile network operators. And, of course, it has to work. To ensure this is the case, the need for monitoring should be immediately obvious. And, while we’ve used mobile networks as an example, the same is true for fixed operators or for any that deliver network-based services to users.
It’s worth adding that control plane monitoring differs fundamentally from traditional network monitoring. Rather than focussing on the health of the network nodes and other components in isolation (are they working correctly? Are they not?), it focuses on monitoring the interactions between those assets, broadly following the path a service takes as it’s in the process of being executed via various elements, depending on the generation of technology concerned.
Monitoring the Control Plane allows operators to more rapidly troubleshoot scheduling and service orchestration issues that arise. The metrics related to the Control Plane (as opposed to the OSS) give a detailed view into its performance, including real-time data on the plane’s overall workload and latencies. In so doing, monitoring provides an insight into what the customer is likely to be experiencing and thus enables important service-related issues to be quickly addressed.
This matters not least because the delivery of any service is inherently complex, involving multiple touch points in a chain, for example the IMS core, the IMS AS, a policy controller and other network assets. To deliver any service, this chain will run from the initial attachment to the network to request for a service, to authorisation and other links all of which are initiated by and orchestrated within the control plane. If there’s no monitoring of the plane’s performance, what happens when a problem arises? For example, what if a device is functioning correctly but policy requests are not being responded to in the right time window? Monitoring and analysis through capture of the control plane information can reveal where the problem lies.
Today, the need for control plane monitoring is made more urgent by the introduction of service-based architectures arriving with and driven by 5G. In previous network generations, procedures defined all interactions between network functions as a series of message exchanges, carried out by protocol interactions in the control plane. With 5G, Service Based Architectures permit network functions to offer and consume services of other network functions. This permits direct interactions between functions.
The advantage here is the promise of allowing services offered by network functions to be reused for other purposes rather than simply processing one-off control procedures defined for implementing a specific function. But with these changes, the need to understand how the control plane is performing becomes more important. And there are numerous other issues, for example the question of state management. In service-based architectures serving nodes are no longer fixed requiring information to be transferred between old and new entities.
And, of course, there’s a great deal more. The key point here is that as the shift to 5G progresses, monitoring the control plane is likely to quickly become a required ingredient to ensure optimal service delivery and, ultimately, to achieve commercial success.
It’s also worth adding that the sum of both traditional network monitoring and control plane monitoring delivers benefits greater than either parts could alone. Objective data drawn from the control plane can work in tandem with traditional network monitoring to create a comprehensive picture of both basic network health and service delivery performance that can be used in combination for analysis, KPI collection and benchmarking – as well as troubleshooting and problem solving.
For today’s operator Quality of Experience is the key metric against which commercial success is benchmarked, so control plane monitoring makes sense. Most contemporary operators already have complex networks and and many are planning for further evolution. Since the control plane is composed of a long list of protocols – legacy, current and next generation – you need to ensure that you can keep pace with these shifts and deploy monitoring across the complete network, for all domains end-to-end – not just, it should be noted, new 5G investments.
Service delivery won’t become simpler, and the loads placed on both individual network components and the nodes involved in delivery chains will become increasingly demanding.
Making sure that monitoring and optimisation challenges are met comprehensively is a foundation stone for a commercially successful future. At Utel, we’d welcome the opportunity to further discuss with you how control plane monitoring can complement the network insights you’re likely already accruing.