Assessment of VoIP location capabilities to support emergency services
This document is the final report of a study conducted for Ofcom by Analysys Mason into the ND1638 architecture designed by the Network Interoperability Consultative Committee (NICC) to provide location information for emergency calls made by VoIP users. The objective of this study is to help Ofcom to develop its understanding of the NICC architecture, including:
- describing the capabilities and any limitations of the architecture
- identifying the challenges of implementing the architecture
- quantifying the implementation and on-going running costs.
The ND1638 architecture has been developed to meet a specific requirement to provide real-time emergency service location capability for VoIP users contacting UK emergency services from DSL access points via the existing architecture used by the stage one Public Safety Answering Points (PSAPs). The architecture achieves this specific requirement and remains, where possible, compatible with international interface standards, namely those developed by the National Emergency Number Association (NENA) in the United States and the Internet Engineering Task Force (IETF). In a European context, ND1638 represents the most detailed work that has been done to address a specific national requirement. ND1638 is, as a result, well placed to influence the Europe-wide initiative by the European Emergency Number Association (EENA) that is due to progress towards specifying VoIP emergency location standards during 2011 (although a common pan-European approach is unlikely due to variation in emergency service architectures between countries).
However, ND1638, which only addresses DSL, is just the first stage in the development of the emergency services location architecture to meet the developing needs of users. The use of different means of VoIP access (e.g. cable, Wi-Fi hotspots, private networks) as well as other next-generation services (e.g. text, images and video) are not currently covered. Work is continuing on addressing these issues in the NICC working group, and it needs to be ensured that the initial implementation of ND1638 is compatible with its on-going development. Our investigations suggest that this is likely to be the case, as for example the ND1638 architecture does not exclude an evolution from a network-centric to an end-device-centric, physical location request model which may be more appropriate to all-IP networks in the future.
To date, there has been very little (if any) progress towards implementation of ND1638 in the UK, and there remain considerable challenges in achieving this. During the study we spoke with a number of UK-based ISPs, VoIP service providers (VSPs) and emergency handling authorities (EHAs), which raised several concerns about the current architecture, relating to:
- Range of access types supported only DSL is supported, and not other access types such as Wi-Fi hotspots, cable networks or private networks.
- Alignment with other standards such as those defined by NENA and IETF, rather than being a UK-specific standard.
- Need for additional implementation guidance to ensure consistent and robust implementation.
- Costs of implementation significant investment is likely to be required.
- Challenge of managing the implementation across so many VSPs and ISPs the UK has a large number of VSPs and ISPs.
- Ensuring the participation of the ISP and access network provider (ANP) community there may be reluctance to participate, particularly amongst ISPs and ANPs not providing VoIP services directly themselves.
The engagement of ISP and ANP organisations is particularly important to the success of the project, as their participation is required to implement the Location Information Server (LIS), which is required to determine the physical geographical address of the VoIP 999 caller. As ISPs and ANPs are not currently involved in emergency calling, and as VoIP callers will in many cases not be their taking a voice service directly from them, this may prove difficult (as economic incentives are misaligned). It is of note that in the United States, where an LIS has been included in the NENA i2 architecture for some time, actual LIS implementation is very sparse: the usual method for determining the location of callers is still to rely on the non-real-time registration of this information by end users, as in the current UK situation (pre-ND1638). The large number of VSPs, ISPs and ANPs in the UK will also provide a challenge in terms of implementation programme management.
The lack of progress towards implementation has also made it difficult to establish definitive costs for implementation across different entities, including EHAs, VSPs, ISPs and application service providers (ASPs). Cost estimates provided by the industry vary widely: estimates of capital costs by study participants ranged from 200 000 to 1 million plus, while for operating costs, estimates varied from 2000 to 200 000 per annum. The wide discrepancies in the cost estimates provided reflect both a certain lack of focus on this area to date, as well as the different starting positions of the various organisations. However, it is apparent that significant investment will be required by a large number of parties.
While it appears that ND1638 provides a viable way forward for VoIP emergency location, the progress of EENA standardisation and the on-going consultation by the US Federal Communications Commission (FCC) on the NENA i3 "next generation 911" architecture should be closely monitored. Both are due to report during 2011. EENA should provide a clear indication of the position of the ND1638 architecture in the context of European compatibility, while the level of NENA architecture implementation that the FCC mandates in the United States may provide some guidance on what can reasonably be expected to be implemented in the UK.
In this section
Report for Ofcom