The three-layered approach to network data monetization

Automated operations within the trendy state-of-the-art telco networks, reminiscent of 5G Core, O-RAN and SDN, all observe the identical sample. The principle enabler is the entity that streamlines assortment and consumption of the community state. That is Community Information and Analytics Operate (NWDAF) in 5G Core, Radio Clever Controller (RIC) in O-RAN and SDN Controller in Software program Outlined Networks. Every of those entities applies a three-layered strategy to community information monetization. Whereas the performance of those community state customers may be very comparable, their implementations are totally different. This ends in vendor particular options, silos per community area, and extra importantly, much less alternative to monetize the potential of community information.

Figuring out the three layers of complexity

Trendy telco networks produce loads of information. Every of the above-mentioned community entities first gather the community information and exposes it to the community state customers. The community state customers are varied analytical modules which will fluctuate from a easy threshold-based algorithms that analyze the present community state to very complicated AI/ML-based algorithms that predict the long run community state. Lastly, primarily based on the insights from these algorithms totally different actions are pushed in the direction of the community, relying on the carried out automation use instances, and thus the loop is closed.

Allow us to now take the instance of a European CSP which has loads of concepts on enhance their community operations. They’ve carried out many various Proof-of-Ideas, all specializing in the Analytics a part of the mannequin. The information- and automation features have been uncared for in these PoC’s, leading to very sub-optimal outcomes. Many use instances yielded disappointing outcomes, as a result of information limitations, and when the outcomes have been good, the required automation capabilities weren’t (but) accessible, which considerably lowered the enterprise advantages of the use case. In a single different case, an vitality saving use case was not deployed, as a result of the required coverage administration operate was not accessible.

IBM advocates to disaggregate information and analytics features and mix them with automation capabilities, as actionable data-driven analytical insights with out automation don’t carry the total enterprise worth.

Right here we establish three layers of complexity, the place every of the three proposed layers brings particular worth: Information Democratization within the Information Layer, an open “plug & play” AI/ML module within the Analytics Layer, and eventually the flexibility to behave instantly on the actionable insights within the Automation Layer.

Information Layer: accumulating the info and making it accessible and comprehensible to all customers.

Analytics Layer: analyzing the info for the varied Use Circumstances to offer actionable insights. 

Automation Layer: appearing upon the actionable insights in an automatic method.

Plenty of the business hype right now is round AI/ML and Automation. Nonetheless, in terms of Telco Networks the complexity of the Information layer is commonly missed.

Information is the brand new pure useful resource; nonetheless, this doesn’t indicate that it’s straightforward to make use of and to have it monetized. To get the utmost profit from it, the proper information have to be effectively mined from its hiding locations and solely afterwards precious insights will be produced out of it. Along with this, community information is generated on a regular basis and everyone has it – certainly, every CSP has an considerable limitless information supply that by no means stops. Subsequently, information mining is the enterprise of each CSP these days. The query is – how to make sure that your information mining is environment friendly and that it permits you to unlock the ability of your information?

Use instances on get began

Automated operations opened a world of prospects for the telco operators. The idea got here similtaneously the necessity within the period of technological evolution in the direction of disaggregation which enormously elevated the complexity of recent telco community operations, but in addition because the principal enabler for the continued evolution of telco companies in the direction of SLAs by no means seen earlier than, reminiscent of ultra-low latency or enhanced cell broadband. From right now’s perspective an increasing number of operations in telco shall be automated as guide operations can not deal with the challenges. Nonetheless, it’s but not clear what are the successful use instances that may carry one of the best financial values to the operators. We name this the Enterprise Worth Speculation. Immediately we are able to distinguish the 2 classes:

The apparent or foundational use instances the place the enterprise worth speculation is intuitively clear. Examples of such use instances are vitality saving, anomaly detection, automated root trigger evaluation, failure prediction. The worth of those use instances has already been confirmed by others, the duty at hand is to develop and deploy these use instances within the personal surroundings, which can not at all times be straightforward to do.

The much less apparent or aspirational use instances the place the enterprise worth speculation is but to be confirmed. Information will be mixed from community, buyer domains and name information information in addition to transactions that go over the community to reach at novel insights which are of worth, e.g., unsuspected correlations. We refer right here to the concepts, inside intestine emotions, and many others. of mixing community information with different sources of information which will yield outcomes however their actual worth and financial of achieved high quality enhancements has but to be proved.

The above elaboration on the nonetheless current uncertainty of the financial values of the data-driven analytics use instances in telco underscores the necessity for the CSPs to host their very personal playgrounds for analytics. Actually, CSPs want an open platform, by which they’ll simply develop and deploy their very own analytical modules but in addition deploy third celebration better of breed analytical modules, as CSPs are unlikely to create all analytical fashions by themselves.

It requires the straightforward entry to community information in an comprehensible method, to keep away from large time loss in looking for the proper community information units. Having available information fashions can scale back effort and time to implement use instances by 30 to 50%.

Step one in the direction of the monetization of your Community Information is to create a Worth Tree. On this course of, it’s obligatory to acknowledge what are the varied classes of worth that may be recognized and to specify these classes in an increasing number of element. The Community Information Worth Tree isn’t static, it continues to evolve as new concepts emerge, and efficiently carried out use instances spark new concepts.

Within the evaluation of which use case to develop, you will need to assess the supply of the proper information, and what you wish to do with outcomes of the info analytics: what’s the required automation to behave upon primarily based on these newly obtained insights. Insights that can’t be acted upon have far much less worth than insights that may be acted upon in an automatic method.

Wish to be taught extra? Contact us at chris.van.maastricht@nl.ibm.com and Maja.Curic@ibm.com.