The challenge of intelligently ensuring correct network configuration

Ensuring correct network configuration at all times is far from being an easy task for mobile network operators (MNOs). It involves maintaining literally thousands of parameters, ranging from 2G to 5G radio access technologies, that will change over time. More specifically, the challenge is to ensure that network configuration is applied and maintained consistently across a variety of radio access network (RAN) technologies and hardware types from multiple equipment vendors. What’s more, frequent network software versions for each of the coexisting RAN technologies and vendors are issued and must be rolled out across the network as a part of the natural network modernisation process.

Can this onerous but essential task be carried out seamlessly and effortlessly in a more efficient and cost effective way?

1. The traditional approach

Traditionally, the process requires a centralised team of the MNO to consolidate the CM updates from several network equipment vendors, as well as to align to the operator’s baseline rules and policies (CM policy assurance). In practice, this activity involves a highly manual and error prone process where network configuration is compared against the baselines (CM policies). In this way, the potential network inconsistencies are identified and, if confirmed as such, work orders for their corrections must be issued by the MNO.

As network features become more and more sophisticated, setting many of those parameters is far from being a simple task. Many have a complex logic behind them as well as dependencies with other parameters, requiring for some conditional parametrisation to be undertaken.

In such a complex scenario, the network may operate with a configuration that could be far from optimal. This translates into a variety of costly side effects ranging from poor customer experience, inefficient resources utilisation, (i.e. misaligned energy saving policy) and unnecessary operation and maintenance costs such as network optimisation, field engineering, etc. This is where intelligent automation comes into play.

2. The urgency for a change

A more effective approach to applying and maintaining a correct network configuration at all times is urgently required by an increasing number of MNOs working towards the achievement of a zero-touch autonomous network. The journey towards this objective is addressed by the TM Forum model that defines five levels of automation, ranging from level 0, where operations and maintenance are manual, to level 5, where a fully autonomous self-healing network has been achieved.

In reality, many operators are still halfway through their journey, at level 2, with a partially autonomous network, or level 3, with a conditional autonomous network. This is the perfect stage along the autonomous network journey where adopting an intelligent and automated solution may help the operators achieve a consistent policy-compliant network configuration in a much more resource-efficient way.

3. The solution

Addressing such a complex challenge requires the deployment of a sophisticated technical solution that takes into account the operator’s business processes affecting multiple technical teams, which are the actual users/clients of such solution.

From an architecture perspective, the solution must address four major areas, making use of complex rule logic parsing, big data and even AI techniques:

• A cloud-based user interface, where approved baselines are uploaded (CM policies) and the logic governing the network parameters rules is reviewed and validated by converting such rules from a high-level- human-like language to a machine-readable format
• A digital twin ideal image of the network, where the defined parameter rules are executed automatically, that is, particularised for all managed objects. This represents the comparison base or the ideal network
• An audit and orchestration process, where the actual network configuration is compared against the particularised network image and discrepancies are detected and reported
• A network configuration alignment process, where the operator has the capability to either confirm and trigger correction (open loop mode), or automatically implement such corrections (closed loop mode), to resolve the identified discrepancies. In any case, a CM-provisioning process must be developed to complete this task

From a business perspective, the user interface serves multiple purposes such as ingesting network baselines, building and testing parameter rules, consulting network policies, reviewing detected network inconsistencies, etc. In general, the user interface must provide a flexible framework to be capable to deliver such services and to customise them to fit multiple operator user profiles and processes, all of them with different information requirements.

Having a network digital twin and orchestration processes maintains the entire automation process running and ultimately delivers the identified network inconsistencies to the CM alignment process, which should be able to run in a fully unsupervised manner. This last step of CM policy alignment must also fit MNO business processes and dependencies with multiple OSSs.