By Fu Xiaoyang
The vast majority of mobile networks involve numerous technologies and forms of access, making integration of user data a must to ensure its consistency, improve user experience, simplify network structure, reduce OPEX, and boost operational efficiency, especially in the LTE/VoLTE era.
Convergence is best
On a legacy network, user data is scattered across various network elements (NEs). The home location register (HLR) will store 2G/3G data, the system architecture evolution home subscriber server (SAE-HSS) will store LTE data, and the IP multimedia subsystem home subscriber server (IMS-HSS) will store voice over LTE (VoLTE) data. This division of labor causes numerous problems, including authentication synchronization failures for LTE users, inconsistencies in PS and evolved packet core (EPC) data, and difficulties with the VoLTE handover process. What's more, maintaining separate user databases increases OPEX and prolongs service deployment times. Erasure of garbage data is also difficult.
Re-synchronization of user authentication data
Authentication data for 2G/3G users is stored in the HLR while that for LTE is stored in the SAE-HSS; when users switch between 2G/3G and LTE, terminals must obtain data independently generated from each. To avoid discrepancies, constant synchronization of the HLR and the SAE-HSS is required. Unfortunately, this is a bandwidth- and signaling-intensive process that impacts user experience. A converged HLR/SAE-HSS database avoids this.
Consistent PS & EPC user data
According to 3GPP protocols, about 70% of user data in the HLR and SAE-HSS is the same, leading to redundancy, cost and potential inconsistencies. Discrepancies in access point name (APN) data can lead to service interruption when users switch between 2G/3G and LTE, but problems caused by data inconsistency are very hard to locate. Telcos must expend a lot of resources to ensure HLR-HSS integrity. A convergent database that encompasses HLR and SAE-HSS user data would simplify service provisioning and facilitate network maintenance, while ensuring data consistency.
Voice service consistency for CS and VoLTE
For circuit-switched (CS) networks, supplementary user service data is stored in the HLR. For VoLTE, it's stored transparently in the IMS-HSS. This leads to an impaired service experience because these databases are independent of each other and cannot be synchronized. CS forward numbers are stored in the HLR; when they roam to a VoLTE network, these numbers become invalid. Users can only change supplementary services (forwarding/blocking etc.) on one network domain on their mobile terminal, so comparable data on 2G/3G and VoLTE networks may be inconsistent. With an integrated HLR/HSS database, telcos can ensure service data consistency. If users change supplementary services through the service delivery system or through their terminal, data changes are universal, guaranteeing voice consistency.
Costly network complexity
Coexistence of independently-built 2G/3G/LTE/VoLTE networks is a complex situation. VoLTE requires eight NEs to manage user data and 36 interfaces for integration and interwork, with at least four interfaces in the upper-layer service provisioning system adapted, making network structure simplification a real challenge. Integration of user data management NEs for unified data storage and processing not only reduces their interworking complexities, it also simplifies the service provisioning system interfaces.
Network simplification will ultimately boost operating efficiency, making a convergent database the best choice for VoLTE. 3GPP R6 introduced user data convergence (UDC) architecture, which logically separates user data and application services, providing a solution for unified user data storage and management. It's no surprise that major vendors are now adopting it.
Converged user data network build scenarios
Most telcos have heavily invested in their legacy HLRs. They can apply one of these two methods for deployment of a converged user data network, based on the status of the legacy HLRs.
Method 1: Retaining live HLR while deploying a convergent HLR/HSS database
Telcos can retain the current HLR and build a new convergent HLR/HSS database to manage all VoLTE user data from the IMS, CS, PS, and EPC domains, with HLR data migrated gradually to the convergent database, and the HLR itself eventually retired at the end of its lifecycle. However, this requires allocation of new international mobile subscriber identity (IMSI) and mobile station international ISDN (MSISDN) numbers to LTE/VoLTE users.
The new independent convergent database will be responsible for unified management of 2G/3G/4G/VoLTE user data. The challenging part will be the coexistence of the HLR and convergent HLR/HSS database. Some users will migrate from the HLR to the convergent HLR/HSS, so flexible number routing (FNR) is needed for correct routing to the convergent database for processing. What's more, the provisioning system has to interoperate with the original HLR and the new convergent database. If the user is an LTE/VoLTE subscriber, the BOSS delivers 2G/3G/LTE services to the convergent HLR/HSS. Otherwise, 2G/3G service is delivered to the HLR.
Method 2: Live network HLR replacement with a convergent HLR/HSS database
If an aging network HLR is proving too costly to maintain, a new HLR/HSS database can replace the HLR completely and be responsible for storage of 2G/3G/LTE/VoLTE user data and support services for the IMS, CS, PS, and EPC domains. However, a cutover can be challenging; network security and a quick fallback must be ensured if problems emerge.
Major industry vendors have cutover experience, with some having developed special tools to secure and facilitate user data cutover/migration. Huawei's Smart Cutover tool can migrate data of various types while ensuring the consistency and completeness of authentication data before cutover, greatly boosting the efficiency of the overall process. Smart Cutover has successfully cutover/migrated over one billion lines involving non-Huawei devices, with no cooperation required with original device vendors.
Future-oriented user data monetization
When they provide customer services, telcos gather a large amount of user data, including subscription data, location information, behavior patterns, and content information. If properly refined, this data is valuable to customer service and marketing efforts. It can also be shared with third parties, if user privacy can be guaranteed, with potential benefit to urban planning and shop location selection. A convergent user data solution aggregates user data for systematic analysis, and allows telcos to share the data internally or externally through secure interfaces, maximizing its value.