A new study by Juniper Research forecasts that the global number of active retail roaming subscribers will reach 1.25 billion by the end of 2024 and rise to 1.6 billion by 2028. With the number of roaming users growing, how can you automate processes and reduce workloads for a mobile carrier’s team?
Roaming services are typically paid for based on the consumption of services, including voice, data transfer, and SMS. This applies to both end-customer-to-operator relations and operator-to-operator relations. Additionally, commitments made in agreements between operators are also tied to service consumption.
Steering of Roaming (SoR), a method necessary to control the fulfillment of these commitments, is done by accepting or rejecting devices’ network registration attempts long before they are capable of generating any billable traffic.
Such an approach poses a significant challenge as it requires a specialized team of people to constantly monitor target fulfillment by comparing data from wholesale agreements with consumption from data clearing house (DCH) reports. Furthermore, a team needs to adjust the IMSI-based steering policy in the SoR system. By utilizing Usage-based Steering, these challenges can be overcome.
What is Usage-based Steering?
Until now, it has been the industry standard to close this gap through human data analysis and manual input, but that can change. Usage-based steering is a concept of making decisions by inputting commercial targets only (e.g., data transfer consumption).
How would that work? From a technical perspective, the system will still have to steer by distributing SIM cards between networks during the registration process – current technological constraints on networks mean no other solution is available. However, one solution is to give the SoR system the necessary knowledge (targets, consumption) and implement logic that would enable it to make the right decisions.
Technical realization – data source
One of the significant decisions that every company needs to make in the process of implementing usage-based steering is the choice of the data source. Each possible option has its advantages and disadvantages, so let’s examine them:
TAP stream
The core of inter-operator billing is TAP files (Transferred Account Procedures). TAP files are sent from a Visiting Public Land Mobile Network (VPLMN) to a Home Public Land Mobile Network (HPLMN) and are essentially a “bill” for services that the subscriber in question has consumed.
TAP files contain many values that can be significant from a steering perspective, including:
- VPLMN information (networkInfo)
- Total charge (auditControlInfo) and currency (accountingInfo)
- Details of service usage in the form of CDRs (callEventDetails)
After supplementing this with knowledge about targets and discounts (which can also be automated through integration with an external tool), the SoR platform would have all the means to track target fulfillment. If paired with appropriate logic and the ability to analyze all data and convert it to necessary adjustments to the steering policy, such a platform could work fully autonomously without any additional input from a human.
That said, using transferred account procedure (TAP) files has drawbacks, with the most significant technological limitation being the delay. A VPLMN can aggregate Call Detail Record (CDRs) for up to 30 days before sending it to HPLMN in a TAP file. With more-or-less steady traffic, this might not be a problem, but it is often not the case in roaming. If, for example, the platform is on the right track to achieve all its targets for the year and then during the December holiday season, the traffic skyrockets and changes its typical characteristic, the platform might go overboard with its standard distribution rules and only become aware of it long after the damage has already been done. It is possible to try to compensate for such fluctuations by analyzing past traffic (roaming is usually highly seasonal), but this adds another layer of complexity to this already challenging endeavor.
DCH integration
Integrating with a Data Clearing House is essentially based on the TAP files as well (with all their advantages and limitations) but with the addition of a proxy system. This method of accessing TAP files has the potential for huge benefits, as DCHs are typically powerful tools for the analysis of TAP data.
Some DCH tools can even send ready recommendations to SoR on which network is missing traffic to reach the target. Utilizing DCH’s potential, it is possible to significantly limit the cost of development work on the SoR side and lower the cost of computing power needed for the project since a significant part of processing would happen on the DCH side.
The main drawback of this approach (apart from the TAP files delay covered in the previous section) is dependence on the DCH. First, the operator’s DCH must agree on the integration, and then there is the matter of what they could send towards SoR, and how.
Internal operator’s nodes
Another option is to use internal operator nodes, which provide near-real-time access and the greatest flexibility during integration. However, it is not always possible due to local breakout roaming implementations (connections done internally within VPLMN, without routing them to HPLMN).
Technical realization – scope of integration
A full-blown usage-based steering that is aware of all transactions generated by a customer is not always justifiable from a commercial perspective. Given the rapidly growing importance of packet data connection, it might be advantageous to limit the scope to steering based on packet data usage only.
Integrating with an operator’s internal nodes responsible for tracking packet data connections (e.g., GGSN and/or PGW) can provide all the benefits of usage-based steering for an operator whose outbound roaming is data-centric – while keeping the cost at a reasonable level.
Technical realization – level of automation
Feeding SoR with relevant data is just the first step. As described in the introduction, at the end of the day, the SoR platform needs to make decisions on International Mobile Subscriber Identity (IMSI) distribution across networks. With that in mind, the next challenge is implementing the logic to enable the platform to “translate” commercial data into the actual steering policy.
At this point, the article arrives at a difficult question that decision-makers in the age of AI have to ask themselves – how much do you actually want to delegate to software?
You have several approaches, depending on the answer to the question above (the bolded sections illustrate the differences between them).
Full automation
A system has access to all the knowledge about business targets and their fulfillment rates. Based on that, it generates a roamers distribution policy and enforces it by controlling all steering-related parameters autonomously.
Partial automation
A system has access to all the knowledge about business targets and their fulfillment rates. Based on that, it generates a roamers distribution policy. It enforces the policy within a framework of steering rules configured by a human.
Recommendations
A system has access to all the knowledge about business targets and their fulfillment rates. Based on that, it generates recommendations on how to adjust roamers distribution policy and steering rules. All configuration is done by a human.
Data preview only
A system is integrated with DCH and/or discount tools and presents relevant data inside the SoR GUI. It does not automatically analyze usage data. This is not usage-based steering per se, but it is a step toward bringing the commercials and steering closer together.
The future of steering of roaming
Usage-based steering of roaming presents challenges and opportunities for mobile network operators. While it offers the potential for more efficient fulfillment of inter-operator commitments and reduced manual monitoring, implementation requires careful consideration of data sources, integration scope, and desired level of automation. Ultimately, usage-based steering represents a significant step in aligning technical operations with commercial objectives in the mobile roaming landscape.
If you are interested in implementing up-to-date roaming solutions and boosting your mobile network, use this form to contact our experts.
About the authorJakub Aleksejczuk
Product Manager
A product manager with over 5 years of experience in the roaming services business, Jakub oversaw the development of the Amplitiv's SMS Welcome platform from scratch. Currently, he manages Amplitiv’s roaming portfolio of products and works closely with roaming managers during implementations. His technical background and cooperation with operators on a global scale enable Jakub to understand the dynamic nature of the roaming services sector.
