The IMS model and the challenges facing new regulations

Different technologies for transmitting voice in data networks and, in particular, IP networks, have been available for a long time. In fact, the majority of telephone service operators have already made their first investments in Voice over IP services, mainly for International Long Distance applications. However, this technology’s potential is far more than a simple interconnection or providing a new means of transportation. VoIP systems permit high optimization of existing network structures and, at the same time, the introduction of new, advanced services, mainly for broadband users. Unfortunately, the lack of a global agreement on how to implement this technology and the growing number of components required in order to offer new services produced serious concerns among different operators, who perceived risks in the short term, particularly with regard to the integration of solutions from different suppliers.

The mobile community, organized in 3GPP, was the first to address this problem and it defined a new network architecture focused on the implementation of multimedia services. So it was that the IMS (IP Multimedia Subsystem) model came about. This model proposes a network architecture based on the following premises:

• Perfectly defined and demarcated network functions

• Clear interfaces between the different network functions

• Use of standard protocols (SIP, Diameter, RTP) in said interfaces.

This model’s biggest achievement is that it provides a flexible architecture that permits easy, fast implementation of new services and also defines open interfaces that ensure interoperability of solutions from different providers.

This model was subsequently complemented so that it would also cover the needs of fixed operators. This task was undertaken by ETSI, and so it was that the TISPAN model emerged. It is worth noting here that the TISPAN model includes the IMS model, without altering it but complementing it with new functions, such as PSTN Switch Emulation, Video Streaming systems, etc.

The IMS/TISPAN model proposes a layered architecture (see Fig.1). The most important aspect of this architecture is that it proposes a single control layer and a single applications layer, regardless of the access network used by the client, i.e. fixed (ADSL, WiFi) or mobile (3G, WiMax). That way, the user’s experience, when it comes to using his services, will not be affected by the access network used, making implementation of a true convergence between fixed and mobile worlds possible.

The fact of having a defined, standard architecture for VoIP services favors rapid adoption of these technologies by operators, creating the need to update existing regulations and, therefore, raising the question of what should be regulated and how.

In connection with this issue, there are different tendencies on how VoIP services should be regulated, which cover an assortment of alternatives. The option that has gained most acceptance internationally proposes regulating the service rather than the underlying technology for providing the service. Consequently, this option establishes that three perfectly distinguishable services exist, but not the Voice over IP service as such:

PSTN Emulation: This is the case of operators who replace part of their traditional switching systems with NGN or IMS components in order to reduce maintenance costs or expand their installed base. In these cases, the user does not perceive that he is a VoIP user, simply that he receives a traditional analogical pair in his home, and the VoIP technology resides in the Access Node and in the switching system. In this case, the tendency is to apply the same regulations as the existing ones for a traditional switched network, as, in fact, the service received by the user does not vary to any great extent.

Voice over Broadband Services: This is the case when the end user has access to broadband over which he receives his telephony services and when at least two conditions are met: the user is assigned a public number under the national numbering plan and there is full interconnection with the PSTN network. These services will be more complex than the traditional telephony services and they will, more than likely, include video, presence, instant messaging, etc. It is precisely this service that raises the largest number of questions when it comes to regulation, although some agreements have already been reached:

• It is generally understood that this service is in no way related to geographic position, which means that long distance rates do not apply (in fact, many countries have a special series of numbers, 0-700 or 0-500 for example, for these users).

• Some countries (particularly in Europe) see this service as a possible replacement for the traditional telephone service and, consequently, they are working on regulations that are similar to existing ones, taking into account the technology’s limitations (e.g. Lifeline).

• In other countries, the tendency is for there to be no in-depth regulation of these services (Australia, for example). Countries that are following this policy also establish that the “traditional PSTN” service that fully complies with current regulations should be available as such, if the user so requires, with the Voice over Broadband services being an alternative option for the end user. This tendency involves a double investment for operators, who must maintain both services.

• Basic requirements such as legal interception, emergency services, etc. are maintained.

• In countries that are considering this service as a replacement for traditional telephony, the quality of the service needs to be guaranteed. The IMS model permits ensuring adequate voice quality throughout the NGN/IP network.

Voice over Internet Services: covers the user-user communication services offered over the Internet, provided they do not offer connectivity with the traditional PSTN/PLMN networks (if they were to do this they would be considered Voice over Broadband services). In general, the tendency is not to regulate these services. Blocking these services by the operator is forbidden in a number of countries. Generally, the voice quality of these services is lower, as it is not possible to prioritize voice package traffic with regard to other types of traffic on the network, precisely because of the lack of an agreement with the operator administering the access network.

This classification merely establishes the bases of a future regulation. It would seem that there is still a long way to go, particularly with regard to the regulation of the Voice of Broadband service. For example, will it be left up to the operators to agree among themselves on whether a user subscribing to a broadband service with one operator will be able to obtain VoIP services provided by another? Will there be regulations that oblige an operator to open its broadband network to other operators for VoIP services in exchange for a given fee? The IMS model addresses and solves interconnection problems from the technical viewpoint and even goes as far as to allow the operator providing VoIP services the possibility of indicating to the operator providing broadband access that it assign a user more resources, temporarily, for the duration of his call, so assuring optimum voice quality. Open interfaces for this communication between networks to function correctly are already perfectly defined. But which of these interfaces will be compulsory for an operator regulatorywise and which will be optional? This is the job that still has to be done.

Maximiliano Auer
NGN Solutions Archtect Manager
FSD - RSC Latin America
ALCATEL/México