Traditionally, a telecoms Operations Center has been a place where engineers have listened to the heartbeat of the network and its services, with the objective of observing irregularities and correcting them. As 5G technology takes over mobile networks and telcos begin to deliver enterprise services, the Operations Center needs to become the brain of the network. It is in the interest of the telecom operators to turn this nerve center of their network into the most intelligent and self-operating part of their network. This will ensure that 5G will deliver its promise to launch enterprise services, in-time and error-free.
Speed and accuracy of 5G networks is critical to deliver enterprise services, such as connected car, remote healthcare, and robotic factories, which take the top 3 commercial slots when it comes to 5G use cases. However, the 5G network slices dedicated to these 3 services need constant and dynamic configuration, which is only possible when the Operations Center offers a certain level of automation. While a fully automated Dark Operations Center is still away, there are several steps that a telco can take to reach this goal.
Why the sudden focus on automation of operations?
5G is primarily about delivering network slices, which connect the traditional mobile network to its enterprise customers who are already launching bandwidth hungry AR/VR- infused videos and other critical services, such as connected health care and connected transport.
For this, 5G needs to be enterprise-ready, extremely efficient, and error-free. This translates into improved latency, responding to problems in real time, to reconfigure slices in real time and to ensure that the 5G core (which is microservices based) can handle the new dynamic traffic.
If the enterprise’s business investments into Industry 4.0 and Connected systems must take off successfully, a high level of automation of 5G operations is critical and immediate.
How Orchestration and Assurance come together for automation
5G network slicing is starting to be a commercial reality, but the 2 key components for telcos to provide the automation needed for network and service operations are assurance and orchestration.
While automated assurance points out the network and services performance automatically, automation of an inventory system is also important to achieve this. Dynamic orchestration of the network and its services automates the network and service configurations and is required to close the loop for auto-correction and auto-remediation.
It is established that assurance and orchestration must work in tandem to drive automation. That said, there are challenges to be overcome in integrating multiple orchestrators and multiple assurance systems for seamless operations to deliver the required levels of automation.
Many automation gaps can be filled by AIOps
Using AIOps, telcos can recognize patterns in multiple large datasets to predict and solve network problems accurately and at scale. With AIOps, service providers can perform anomaly detection and generate predictive capacity and coverage analytics. AIOps also offers predictive root cause analysis to speed up troubleshooting, prioritize resolution and automatically identify the next best actions for remediation. These recommendations can be activated through closed-loop automation. AIOps, based on cloud infrastructure, is now recognized as an efficient path for network and service automation. The cloud environment is also known to better handle the spikes in 5G traffic. With AIOps, not only are vast amounts of data processed efficiently, but operational errors are also reduced.
Some quick steps towards Operations Center automation
4 key steps that can be undertaken right away for Operations Center automation are:
- Automate the generation of network and service performance KPIs
- Automate the trouble ticketing process
- Automate the monitoring of service level agreements for enterprise services
- Automate root cause analysis leading to auto-remediation
These steps form the edifice of the fully automated dark Operations Center of the near future. Here we should mention the role of some TMF APIs that can help bring some of these steps into the real world.
- TMF 628 between Performance management and Service Quality management
- TMF 621 for incident/problem tickets
- TMF 655 for change management tickets.
- TMF 633 / 634 to onboard service and resource catalog
- TMF 634/639 to publish resource catalog.
- TMF 656 in the datastore of Service Problem management
- TMF 642 in the Fault management store for internal communication between layers.
Conclusion
The success of 5G services for the enterprises is dependent on its speed, scale, and dynamicity. Only a Service Assurance system that responds to these critical aspects of 5G, reacts and remediates in time, will help the 5G mobile operator to stay true to its promise to the enterprise. Some of this is achieved through the inherent design of 5G, i.e., microservices-based 5G core and the dynamicity of 5G network slices, however, 5G performance is highly dependent on predictive analytics and error-free reliable operations. The automation of the telco Operations Center is now a necessity for all mobile, broadband and enterprise networks. Not only for faster time to detect problems, but also for reducing the time and effort to resolve them. The rebuilding and automation of an intelligent Operations Center, with AIOps at its center, will ultimately deliver the critical services that the enterprise requires.
This blog was first posted on TMForum Inform on 9th Sep 2022: https://inform.tmforum.org/features-and-opinion/how-aiops-enabled-automation-will-deliver-the-dark-operations-center/
