Challenge
In 5G Non-Standalone (NSA) networks, subscribers consume more data than in 4G networks, even though the network infrastructure is the same (5G NSA over LTE), as well as the connectivity between the visited and the home network. Better quality and broader bandwidth increase the data consumption of the roaming subscribers. Therefore, assigning a subscriber with high data demand to a roaming network with higher QoS will result in higher data consumption. The subscriber will gain better quality while using roaming data and potentially purchase an extension for his/her data package when the existing package is exhausted.
There are various challenges in QoS-based steering:
- Limited control and visibility over data QoS.
- System integrations from offline QoS monitoring tools to real-time steering execution.
- Subscriber complaints leading to steering interruptions (i.e., no steering enforced) and additional costs.
- Increased demand for analytical solutions.
However, these challenges are also opportunities for mobile operators:
- Examine and control the performance of your own network internationally.
- Implement accurate steering policies that can benefit commercial agreements.
- Drive better roaming experience to meet consumer and enterprise data demands.
Solution
The Adaptive Steering algorithms adjust and automate steering decisions based on partners’ network scoring, data QoS performance, and dynamic subscriber segmentation. It prioritizes better data QoS networks to boost usage and ensure user experience. By combining data connectivity requirements/SLAs from all enterprise verticals, Adaptive Steering provides the framework for a holistic understanding of the roaming ecosystem and enables the execution of QoS business models.
The analysis is based on GPRS Tunneling Protocol (GTP) traffic and SS7/Diameter LTE/HTTPS 5G signaling traffic feeds. The system can detect problematic scenarios such as registration failures, no data session creation, low vs. high data throughput, high handover ratio between roaming networks, and other QoS parameters.
The system dynamically clusters roaming subscribers into level-of-usage segments and scores roaming networks according to their roaming QoS. Each cluster is associated with different steering policies based on the network scoring. For example, high-data roamers are assigned to high-QoS preferred networks. Silent or low-data roamers can be assigned to low-QoS networks. Subscribers experiencing coverage issues can be assigned to networks with fewer coverage shadows. All the clustering methods are dynamically calculated and adapted in real-time according to field conditions.
Result
Adaptive Steering continuously monitors multiple roaming events, analyzes signaling and data usage QoS metrics in real-time, and adjusts steering decisions based on partner network performance and subscribers' data consumption behavior. It transitions roaming management from fixed to automated by aligning subscriber usage profiles with appropriate network QoS levels. Adaptive Steering leverages the right KPIs to classify subscribers into QoS-sensitive communities, each with tailored steering policies. This approach aligns QoS with business objectives, targeting specific subscriber segments to prevent revenue loss caused by QoS issues.
