In the era of 5G, the operators can provide more diversified and low latency service experience to the end user by providing different kinds of network slices and constantly securing the SLA/QoS for each scenario. The major source of income of operators will be changed from direct selling of voice erlang and data throughput to common users, to corporate sales with the vertical industries, providing services with different combination of SLA/QoS standards, such as latency, throughput and number of IoT connections.
For the operation teams, their main job will become to ensure these are achieved in the live network.
However, as stated in above, the future 5G network may contain dozens of dynamic network slices with different SLA/QoS requirements and more physical sites. These factors will result the work load of the operation team being increased, and hence it is difficult for operation teams to control the OPEX. In order to solve this problem, the industry expects that AI will widely infiltrate all aspects of operations; these AI applications must be supported by a more intelligent OSS to improve the operational efficiency and the level of automation.
ZTE has published its uSmartNet AIOps solution, which is able to provide a comprehensive intelligent operational solution, from field service to full network value driven under Pre5G and 5G network. This solution is based on AI technology and uses a big-data platform to perform deep mining of the network, service and user data in multiple dimensions, achieve cross domain resource integration, dynamic resource scheduling, dynamic threshold adjustment, intelligent root cause analysis and more. This solution strongly supports the transformation of manual operation to full automation.
Together with the dramatic increase in the amount of network equipment and services, the traditional way of NOC operation is lacking the efficiency and unable to control the service quality under the dynamic environment. The intelligent NOC solution can use the related AI engine and perform the analysis of network data, predict potential risks, improve the accuracy of the prediction and hence achieve the transformation from traditional NOC to intelligent NOC based on AI. This intelligent AIOps solution can drive the 5G system, service platform, and related modules (such as Global Assurance, Global Inventory, and Global Provisioning) to maximize the visualization of 5G operations, resources, deployment and configuration of services. The 5G daily operation can achieve end-to-end automation and improve the production efficiency.
Function abnormality detection can automatically detect and analyze abnormality in QoS or other KPIs , and deduce a possible root cause. The function is also linked with the automatic RCA function, which can correlate the abnormality with the known alarms and improve the accuracy of the root cause analysis. The result of which will be inserted into AI data base for further reference.
The root cause analysis adopts multiple machine learning algorithms, achieved cross layer and cross domain root alarms and root cause analysis, the result of which can correlate with the abnormality of the performance and be able to reveal the effect on network slices and drive the 5G service system platform to invoke the right policy and perform self-healing.
Intelligent assurance of access network
In the 5G era, as the bandwidth and coverage requirement is higher, the number of physical sites will be increased a great deal and hence the first thing that operation teams should improve is the efficiency and accuracy of the scheduling of work orders.
In order to achieve this, the AIOps provides 3 major modules. They are intelligent scheduling, fault prediction and intelligent SPM. The intelligent scheduling adopts AI algorithms such as a genetic algorithm and ACS, which can dynamically achieve the optimal scheduling of the tasks and resources considering many aspects such as SLA, skill level, travel routes and etc. The Intelligent SPM function and the fault prediction function can predict the proper amount of stock for local warehouses and possible causes of future problems based on historical alarms and root causes and guide the field team to schedule preventative checks and reduce the number of failures.
With the fast development of the digital services in recent years, many of the operators’ focus points have changed from network quality to service quality and customer experience. Some operators have developed SOC’s (Service Operation Center) to conduct service quality monitoring and control. In the era of 5G, it is expected that many of the vertical industries will take the chance to evolve their services, create new services or service models. For operators, it is essential to be able to identify the right opportunity, quickly develop the appropriate service slices, and ensure the QoS as required by the industries. In these aspects, AI will be widely adopted to perform analysis on huge amounts of data and help the operation team and marketing team in decision making. ZTE SOC has the following solution in AI:
- Linking user portfolio and user behavior, to identify the potential value
- Linking user experience with service KPIs, to locate hidden problems
- Quickly locating the users’ problem, issuing an automatic ticket and enabling the pre-handling of the problem
- Closed loop service quality issue handling and efficiency improvement
ZTE uSmartNet AIOps solution has been trailed by many operators, helped them to cut down the time, workload, and resource expenditure in function and service links, improving the network quality, shrinking down the time required for fault handling, greatly improve the operation efficiency and hence enabling closed loop automation in many aspects of operation. Overall, it can greatly accelerate the level of automation and intelligence of your operation.
This Knowledge Network article was written by ZTE, it does not necessarily reflect the views of Total Telecom. To include your own content here contact firstname.lastname@example.org