NGNs and the future of regulated voice services

Interested in this topic? Contact our expert Suella Hansen @

April 19, 2006

Determining a ‘cost based’ price for voice on an NGN will be difficult using the conventional economic (LRIC) approach and, if implemented, may lead to pricing anomalies between voice and broadband NGN services

As Next Generation Network (NGN) technologies and services begin to replace conventional telecommunications infrastructure, regulators around the world are recognising the need to review and adapt policies and processes accordingly. One of the key regulatory issues that arises from the introduction of NGN is the rapid replacement of PSTN voice services with Voice over Internet Protocol (VoIP).

Voice service regulation

Traditionally, PSTN voice has been the most heavily price controlled telecommunications service in the world, with regulators and governments setting:

  • price caps for PSTN residential line rentals
  • prices for the delivery of interconnecting traffic to and from PSTN customers
  • universal service commitments and funds for incumbent PSTN service providers.

The key to PSTN regulation has been the regulator’s ability to set prices based on its determination of the economic cost of access lines and minutes of traffic. Although often involving complex economic cost concepts – such as Long Run Incremental Costs (LRIC) – modelling of PSTN costs has been achievable because:

  • PSTN technologies are very mature and network component costs are well understood
  • PSTN network architectures and design methodologies are relatively standard
  • the PSTN voice service is highly standardised and the network resources required to transmit a voice call are precisely determined, allowing accurate identification of the network costs incurred in delivering a service.

As VoIP replaces PSTN voice, regulators must decide whether existing voice regulations are still appropriate, and if so, whether current regulated prices continue to reflect accurately the costs of providing the service.

Voice over IP

VoIP has grown rapidly over the past few years, principally in the areas of:

  • Replacement of enterprise PABX and private networks – using high quality commercial IP telephone equipment, often connecting to the PSTN using conventional ISDN lines.
  • Bypass of national and international PSTN toll and termination charges using private IP networks – often provided by Internet Service Providers (ISPs) using leased international capacity and IP call termination arrangements in a range of overseas countries.
  • Bypass of national and international PSTN toll charges using the Internet– often consumers or small businesses using free PC-to-PC services such as Skype.

Each of these services comes with an expectation of service quality, ranging from “excellent” for an IP PABX through to “best effort” for an Internet-based Skype call.

Cost and pricing of VoIP

As described above, regulated PSTN voice prices are often based on the estimated cost of delivering a voice service. However given the relative immaturity of VoIP services, there is little useful information on the costs of delivering the equivalent of a PSTN voice service over an IP network. Such a service would require guaranteed PSTN voice quality and the ability to dial any other PSTN or VoIP user using standard numbers.

Current pricing plans for VoIP services usually rely on subscription charges to cover the costs of on-net calling (within the service provider’s IP network) and charge termination rates for calls going off-net (such as to the PSTN). The VoIP service normally requires the subscriber to separately purchase:

  • broadband access (which pays for the access and any IP traffic costs)
  • Digital Subscriber Line (DSL) or cable modem/router
  • IP voice conversion device for a normal phone or a special IP phone.

Assuming the service is provided over DSL then, in most regulatory regimes, the subscriber must also purchase conventional PSTN voice access service from the copper loop provider. The value of VoIP to the end user is in its ability to bypass conventional national and international toll charges. For some users, these savings are sufficient to justify the additional cost of VoIP on top of existing broadband and PSTN line rental costs.

Often studies of VoIP costs look only at the additional marginal cost of providing VoIP, assuming that the remaining broadband infrastructure is paid for by the broadband and PSTN subscription. The bulk of the VoIP infrastructure (illustrated in Exhibit 1) is required to deliver the basic broadband access service, meaning that the only network elements remaining to be costed for VoIP service are:

  • IP voice call controllers
  • IP to PSTN gateways
  • any costs of national and international bandwidth.

Exhibit 1: NGN VoIP infrastructure [Source: Network Strategies]


When these costs are shared over a sufficiently large traffic base, the marginal cost of delivering VoIP services on a broadband network can be very small. If regulators were to adopt these costs as a basis for setting regulated voice service prices then, for example, a VoIP termination service is very likely to be significantly cheaper than PSTN termination.

VoIP and the NGN

If end users adopt VoIP as a replacement for their fixed line, then that service should reach the required quality standards and, in theory, pay for the true costs of the infrastructure it uses. The full cost of VoIP would therefore include broadband access and traffic charges and the costs of upgrades to IP networks to meet PSTN speech quality, call blocking and network availability standards. Contrary to the current VoIP pricing scenario, if an NGN network (including ubiquitous broadband access) were provisioned to carry only voice, the resulting VoIP costs are likely to be higher than current PSTN costs.

Since the NGN has additional costs and higher functionality than the PSTN, operators must utilise the new functionality to collect revenue from a wider range of services, thereby ensuring an economically viable investment. This is why many operators have a “triple play” strategy, involving delivery of Internet, television (Video on Demand, VOD or Live Stream) and phone (VoIP) on the NGN. Determining a “cost based” price for voice on an NGN will be difficult using the conventional economic (LRIC) approach and, if implemented, may lead to pricing anomalies between voice and broadband NGN services, which may not be priced on the basis of network resource usage.

For example, VOD may use massive bandwidth resources, but the service price is likely to be set by the comparable cost of hiring a DVD or subscribing to a conventional television service. In comparison, VoIP may use very little bandwidth, but may be of much higher value to the end user than VOD on a per kbit/s basis. It can be argued that if NGN voice service prices were regulated to a very low level, based on network resources used, the impact on the market could be a corresponding reduction in the ability of operators to invest and to re-coup their existing investment in widespread broadband infrastructure.

Regulatory re-think required?

In every regulatory regime, the overarching issue of whether voice service prices should continue to be controlled in the NGN world is a matter of national policy, driven by issues of investment, public good, cost benefits, encouragement of competition and universal service requirements. These issues will be addressed by the relevant policy makers and regulators.

In terms of setting regulatory prices for voice services, it is clear that the conventional economic LRIC approach can only be used reliably for the PSTN, where costs are known and can be reliably attributed to standardised services. For VoIP, network requirements and usage can vary significantly depending on the type and quality of voice service being provided and, by its nature, the IP network is dynamically shared between a wide range of services. This dynamic service and network environment makes it significantly more difficult to attribute network costs to services and could make the use of LRIC unreliable. Consistent use of LRIC would require widespread agreement on the network assets involved in delivering VoIP and a methodology for allocating costs to the wide range of possible VoIP service types.

If regulators simply consider the cost of delivering a VoIP call to an IP user today, they will find that it is principally borne by the called line in terms of its ongoing broadband access and traffic charges. In fact, in some scenarios, a VoIP service supplier need implement little or no physical network to deliver its service, indicating a significant paradigm shift from the PSTN environment.

Should regulators choose to continue the price control of voice services in an NGN environment, it is likely that they will need to significantly modify their approach to cost-based pricing. Future pricing of services by operators of public networks is likely to be driven more by content and value rather than network and bandwidth usage.