User Data Convergence (UDC) in Release 9 and its evolution

The below is mish-mash from the specs (see refrences at the end)

With the increase of service entities and the resulting user data types, User Data Convergence (UDC) is required to ensure the consistency of storage and data models.

UDC:

• simplifies the overall network topology and interfaces

• overcomes the data capacity bottleneck of a single entry point

• avoids data duplication and inconsistency

• reduces CAPEX and OPEX.

UDC simplifies creation of new services and facilitates service development and deployment though a common set of user data.

UDC promotes service and network convergence to support the increasing number of new services including Internet services and UE applications. A new facility User Data Repository (UDR) is considered for UDC.

In UDC, all the user data is stored in a single UDR allowing access from core and service network entities.

To achieve high performance, reliability, security and scalability, the UDR entity may consist of a network of different components distributed geographically, and exposes capabilities via open interfaces in multiple access entry points.

In the current 3GPP system, user data are scattered in several domains (e.g. CS, PS, IMS) and different network entities (e.g. HLR, HSS, Application Servers). With the increase of user data entities and the resulting data types, it is more difficult for integrated services to access necessary user information from plural entities.

The scenario mentioned herein is kind of called “User Data Silo”, which is the major paradigm of user data deployment for the time being, as illustrated by Fig.1. below

With the user data silos, user data are independently accessed, stored and managed independently. That brings many challenges to network deployment and evolution. Different user data access interfaces impose complexity on network topology as well as on application development, especially for booming Internet services and incoming IP-based UE applications; separated user data increases management workload. Moreover, new networks and services such as IMS are expected, so that the introduction of their user data only makes things worse, not to mention network and service convergence even if those user data have a lot in common and are correlated to each other. Separation also undermines the value of user data mining.

User data convergence is required to ensure the consistency of storage and data models. User data convergence will simplify overall network topology and interfaces, overcome the data capacity bottleneck of a single entry point, avoid data duplication and inconsistency and reduce CAPEX and OPEX. Also it will simplify the creation of new services and facilitate service development and deployment though a common set of user data. Finally it will promote service and network convergence to support the increasing number of new services including Internet services and UE applications. In this regard, a new facility User Data Repository (UDR) should be considered for user data convergence.

As illustrated by Fig. 2 above, User Data Convergence, as opposed to User Data Silo, is simply to move the user data from where it belonged, to a facility here called User Data Repository (UDR) where it can be accessed, stored and managed in a common way. Despite of the diversity of user data structures for different services, user data can be decomposed and reformed by a common data model framework (e.g. tree-like data model, rational data model) provided by UDR. In that case, user data categorized by services can be regrouped and identified by user ID, leaving no data redundancy. Also, convergence in data model will unify the user data access interface and its protocol, which will promote new service application development. Thereby, the capability of user data convergence can be open to creation of data-less applications.

There are plenty of data distributed in the 3GPP system which is used to perform the services, for instance, the configuration data of a network entity, the session data of a multimedia call, the IP address of a terminal, etc. With respect to user data, it refers to all kinds of the information related to users who make use of the services provided by the 3GPP system.

In 3GPP system, user data is spread widely through the different entities (e.g. HLR, HSS, VLR, Application servers) and also the type of user data is various. It is of paramount importance to categorize the user data before going through the convergence of user data.

The UDC shall support multiple application user data simultaneously, e.g. HSS and others.

Any application can retrieve data from the UDC and store data in it. The applications shall be responsible of updating the UDC with the dynamic changes of the user profile due to traffic reasons (e.g. user status, user location…) or as a consequence of subscriber procedures.

User Subscription Data: Before a user can enjoy a service, he may need to subscribe the service first. The subscription data relates to the necessary information the mobile system ought to know to perform the service. User identities (e.g. MSISDN, IMSI, IMPU, IMPI), service data (e.g. service profile in IMS) , and transparent data (data stored by Application Servers for service execution) are the examples of the subscription data. This kind of user data has a lifetime as long as the user is permitted to use the service and may be modified during the lifetime. User may be accessed and configured via various means, e.g. customer service, web interface, UE Presence service. The subscription data is composed of different types such as authentication data, configuration data, etc. Different type of data may require different levels of security.

User content Data: Some applications may have to store content defined by the user and that here may be quite large (e.g. Photos, videos) User content data can reach very high volume (e.g. Hundreds of Mbytes and more), and the size required to store them may largely vary over time. They generally do not require the real time constraints as user profile data may require. Storage of user data content is not typically subject of UDR. Storage of user data content is not typically subject of UDR. UDC on user content data can be achieved by converging them with links or references, such as URLs, to other entity.

User Behaviour Data: Such data concerns the usage of services by a user as services are consumed. Generally there are event data records that can be generated on various events in the usage of services by a user and that can be used not only for charging or billing purposes but e.g. for user profiling regarding user behaviour and habits, and that can be valuable for marketing purposes. The amount of such data is also quite different from other categories, they present a cumulative effect as such data can be continuously generated by the network implying a need for corresponding storage. Usage data may require real time aspects about their collection (e.g. for on line charging), they are also often characterized by a high amount of back office processing (e.g. Billing, user profiling). Processing of user behaviour data such as for CRM, billing, data mining is not typically subject of UDR. Those might be processed with lower priority or by external systems whereby UDR supports mass data transfer.

User Status Data: This kind of user data contains call-related or session-related dynamic data (e.g. MSRN, P-TMSI), which are typically stored in VLR or SGSN. These dynamic data are only used by their owner transitorily and proprietarily, and hardly shared by other services in the short term.

Figure 4.1-1 below presents the reference UDC architecture. UDC is the logical representation of the layered architecture that separates the user data from the application logic, so that user data is stored in a logically unique repository allowing access from entities handling an application logic, hereby named Application Front Ends.

In the architecture, the User Data Repository (UDR) is a functional entity that acts as a single logical repository of user data and is unique from Application Front End’s perspective. Entities which do not store user data and that need to access user data stored in the UDR are collectively known as application front ends.

NOTE: Depending on the different network deployment, there may be more than one UDC in an operator’s network.

Application Front Ends (FE) connect to the UDR through the reference point named Ud to access user data.

Figure 4.1-2 shows how the UDC reference architecture is related to the overall network architecture by comparing a Non-UDC Network with an UDC Network. In the non-UDC Network, the figure shows NEs with their own database storing persistent user data and a NE accessing an external database; in both cases, when UDC architecture is applied, the persistent user data are moved to the UDR and concerned NEs are becoming Application-FEs (NE-FEs) according to the UDC architecture. This figure also shows that network interfaces between NEs are not impacted .A Network Element (NE), which in its original form represents application logic with persistent data storage, when the UDC architecture is applied, may become a NE Front End, since the related persistent data storage is moved to the UDR.

3GPP TS 23.335 gives more details and information flows for User Data Convergence

Further evolution of UDC is being studied part of 3GPP TR 23.845. The TR tries to address the assumption of multiple UDRs in a PLMN, to identify consequences and the possible impacts on existing UDC specifications. From a practical point of view, even if the aim is to have one single logical repository, a certain number of considerations may drive to have more than one UDR in a PLMN.

For very large networks with a very large amount of users, although an UDR may be implemented in a distributed architecture and multiple database servers with geographical distribution and geographical redundancy, an operator may consider to deploy several UDRs between which it will distribute the users.

More details in the technical report.

References:

3GPP TR 22.985: Service requirement for the User Data Convergence (UDC) (Release 9)

3GPP TS 23.335: User Data Convergence (UDC); Technical realization and information flows; Stage 2 (Release 10)

3GPP TS 29.335: User Data Convergence (UDC); User Data Repository Access Protocol over the Ud interface; Stage 3 (Release 10)

3GPP TS 32.182: User Data Convergence (UDC); Common Baseline Information Model (CBIM) (Release 10)

3GPP TR 32.901: Study on User Data Convergence (UDC) information model handling and provisioning: Example Use Cases (Release 11)

3GPP TR 29.935: Study on UDC Data Model.

3GPP TR 23.845: Study on User Data Convergence (UDC) evolution (Release 10)