A Model for Emergency Service of VoIP Through Certification and Labeling
Douglas C. Sicker
Tom Lookabaugh
Executive Summary
Voice over Internet Protocol (VoIP) will transform many aspects of traditional telephony service including technology, the business models and the regulatory constructs that govern such service. This transformation is generating a host of technical, business, social and policy problems. The Federal Communications Commission could attempt to mandate obligations or specific solutions to the policy issues around VoIP, but is instead looking first to industry initiatives focused on key functionality that users have come to expect of telecommunications services. High among these desired functionalities is access to emergency services that allow a user to summon fire, medical or law enforcement agencies. Such services were traditionally required (and subsequently implemented) through state and federal regulations.
Reproducing emergency services in the VoIP space has proven to be a considerable task, if for no other reason then the wide and diverse variety of VoIP implementations and implementers. Regardless of this difficulty, emergency service capability is a critical social concern, making it is particularly important for the industry to propose viable solutions for promoting VoIP emergency services before regulators are compelled to mandate a solution, an outcome that often suffers compromises both through demands on expertise that may be better represented in industry and through the mechanisms of political influence and regulatory capture. While technical and business communities have, in fact, made considerable progress in this area, significant uncertainty and deployment problems still exist.
The question we ask is: can an industry based certification and labeling process credibly address social and policy expectations regarding emergency services and VoIP, thus avoiding the need for government regulation at this critical time? We hypothesize that it can.
To establish this, we will develop just such a model for VoIP emergency service compliance through industry certification and device labeling. The intent of this model is to support a wide range of emergency service implementations while providing the user some validation that the service will operate as anticipated. To do this we first examine the various possible technical implementations for E911 VoIP. Then, based on these implementations, we will define a set of possible implementation models for the certification and labeling process. We will then consider how each of these models might operate within the certification process. This will include describing the details of how these models would function and respective requirements. In determining the appropriate certifications and labeling model, we will examine similar mechanisms that presently exist within the federal government, e.g., the Part 68 process of attached devices.[1]
We evaluate the success of our model in an immediate sense by verifying its internal consistency and consistency with historically successful precedents. We also propose the specific methods that would be used in an actual implementation of our model over a longer time period.
Our experience in this area makes us well suited to the task at hand. We have substantial industry and academic experience with implementing VoIP technology and substantial lab resources for testing the various configurations. We also have extensive experience with the federal and state government regulatory process, including working within the FCC as well as serving on advisory boards that focused on state regulatory issues, as well as experience in technical and executive roles in industry. Lastly as academics, we have participated in a variety of conferences and program committees that examine the interaction of technology and policy.
In summary, we believe that such a model will allow industry to develop VoIP services with minimal government oversight while ensuring socially desirable functions are provided, that we are well qualified to develop such a model, and that development and dissemination of a model in the next few months is very timely relative to the rapid increase in regulatory discussion around VoIP.
Introduction
The transformation to VoIP will ultimately impact all sectors of the telecommunications services industry – traditional incumbent local exchange carriers, cable providers, wireless service providers, emergency service providers, etc. – and is only in its very early stages. Similar to the service providers, both hardware and software developers are struggling to understand and take advantage of the opportunities in this area, including prospects for new technologies (such as integrated messaging and mobile collaboration). Moreover, business, governmental and residential consumer users of telecommunications equipment and services are looking for guidance on when and how to upgrade to a new technological frontier. Finally, governmental agencies are struggling to ensure that social policy concerns are addressed in a very different technological environment.
The disruption of VoIP promises to upend a century old model of voice telephony by creating a more dynamic marketplace and changing the point of control in from the central office switch to the end user’s device. For all of the relevant stakeholders affected by VoIP (i.e., service providers, hardware and software vendors, customers, and governmental agencies), there is a great need for clear insights and analysis of the issues raised by the transition from the traditional Public Switched Telephone Network (PSTN) to a packet-switched, Internet Protocol-based architecture.
The social policy implications of VoIP present regulators and incumbent businesses with a true dilemma – i.e., choices between mutually exclusive and equally unfavorable options. Under the current PSTN-based voice telephone network, a number of critical policy goals, such as reliable E-911 service, are provided effectively and reliably. The VoIP world, at least as it exists today, does not deliver the same quality E-911 service, leading some state regulators to impose legacy regulation on this service to require it to do just that (along with other requirements), regardless of the dramatic technical and business model differences in nascent VoIP services.[2] The FCC has only begun to develop its regulatory strategy for how to address VoIP, although Chairman Powell has made clear that it will need to differ from the legacy model used in the PSTN context.[3] Given this situation, it is unclear how such services might evolve in a meaningful and timely fashion. To help address one aspect of this dilemma we propose the following work.
Objectives
The objective of this project is to develop a model for VoIP emergency service compliance through a process of industry certification and device labeling. The intent is to support a wide range of emergency service implementations while providing the user some validation that the service will operate as anticipated. At a high level our goals are twofold, (1) to examine possible solutions for VoIP E911 and (2) to provide a strategies for implementing VoIP E911 in a way that satisfies the various stakeholders.
The first part of our analysis involves the examination of potential technological solutions designed to address the barriers to public adoption of the technology and to ensure that longstanding social goals of E-911 are addressed. In particular, there are numerous technical shortcomings and uncertainties that must be resolved before VoIP will mature into a replacement for traditional telephony. We will thus define a set of technology recommendations that explicitly considers the larger economic, policy and technology concerns that are integral to VoIP’s success. The goal of these recommendations will be to provide direction for the models of certification. To do so effectively, we will conduct some testing of different solutions so as to identify the relevant advantages and disadvantages of particular approaches.
Second, we will examine strategies for implementing a certification of these alternatives. In so doing, we will take special care to address the lessons that emerged from the challenges of addressing the above issues during the development of wireless telephone service. In that case, for example, the effort to ensure the development and adoption of E-911 solutions took a long and convoluted path.[4] In the VoIP arena, some of the positive and negative lessons from the wireless telephone service context may well apply whereas others will not. In all likelihood, the E-911 challenges for VoIP will be much more difficult to address than it was for wireless services because, among other reasons, there are likely to be a greater number of carriers in the VoIP environment and those carriers do not necessarily control the network. Moreover, the many different “flavors” of VoIP – such as VoIP over Wi-Fi (including through “dual mode” telephones, which can switch between cellular and Wi-Fi)[5] – suggest that a “one size fits all” solution may fail in this context. Thus, the appropriate approach to this issue may well resemble the self-regulatory model used by Internet standard setting bodies more than the traditional model of common carriage regulation. In tailoring such a regime, it will be crucial to account for the fact that equipment defined more by its software than hardware presents novel regulatory challenges.
Expected Output
The expected output of this research is multifold. First, we expect to provide a complete summary of the existing solutions for E911 in the VoIP space (something that would place the regulatory discourse on much firmer ground and not currently found in the literature). We will examine the proposed solutions for VoIP E911 and create a set of service criteria based on such aspects as architectural design and level of service. We will also assess emerging VoIP E911 applications, in particular their feasibility, ease of adoption, and whether they raise new social policy questions (e.g., intruding on privacy). Second, we intend to provide a number of categories by which to define these solutions. After identifying the most promising technological solutions for implementing VoIP E911, we will develop a more comprehensive way of identifying the various solutions. Third, we will assess the existing models of certification from related industries to provide a template for this problem. Fourth, we will examine the possible solutions in the context of our proposed model and provide a set of criteria on which to judge the specific E911 VoIP solution. Lastly, we will describe a labeling process for informing the consumer.
Clearly, this is not economic or legal analysis; rather it is a technology assessment and a subsequent proposal for a non-regulatory alternative to a daunting policy problem. The intent of the work is to propose a series of certification and labeling models that could be applied to this problem in hopes of easing the transition to VoIP while ensuring social needs are met.
Techniques
The techniques used in this research project include:
1. Historical Analysis and Literature Review. Most successful policy proposals, including techno-policy, are grounded in a thorough understanding of precedents and experience. This is particularly true given the role of law and politics in establishing and maintaining policy. This kind of analysis is typically much farther reaching than the type of literature review used in pure scientific and engineering research and requires both familiarity with literature threads from multiple disciplines (including law, sociology, economics, business, political science, engineering, and computer science) as well as diverse argument analysis skills.
2. Systems Engineering. The analysis of complex technical systems (such as practical VoIP implementations) requires the ability to sharply define external requirements and interactions, subsystems, and the internal interactions between subsystems, and to generate overall performance and behavior expectations for the system as a whole. We will apply system engineering techniques to identify, decompose, and evaluate emergency service impacts in diverse VoIP implementations. Importantly, we will highlight the impact of business model aspects of overall systems on the practicality of certification as a solution to emergency services.
3. Technical Experimentation. Based on our experience in similar problems, we expect a small number of key points will be best illuminated by limited implementation experiments. Predominantly, these are meant to test feasibility in the sense of generic computer science research (by creating running code). We may also occasionally be interested in a particular performance measurement (e.g., the effectiveness of internal location techniques). We will use our laboratory facilities, well equipped with traditional telephony equipment, VoIP equipment, and generic computing resources, to perform any necessary experiments.
Steps
1. Technical assessment (performance, reliability, accuracy, etc.) based on an iterative process of classification of fundamental trends in technology, implementation, and business models around VoIP and a system engineering analysis to illuminate key trade-offs and limits. We expect to use our laboratory facilities for small experiments at key points in our analysis. This step draws on all three of our key techniques above.
2. Secondary research into existing certification processes. Secondary research draws on historical analysis and literature review techniques, and includes collecting and analyzing original descriptions as well as outcomes research on certification processes, how they have been embedded in overall policy systems including government regulation, and the synthesis of an overall model for successful certification tailored to the particular technical and industrial circumstances of VoIP.
3. Build a model of a new certification and labeling process. Model building involves iterating between proposed solutions and analysis and critique using the findings of our preparatory work and draws on system engineering techniques. We note that the model does not yet exist – it is necessarily created as a result of the work in (1) and (2) above rather than a priori. But there are obvious precedents that we will be building from, including, for example, the Underwriter Laboratories model for industry self-certification with respect to safety, or the WiFi Alliance model for industry certification of compatibility. There is a risk that such precedents cannot be extended to something perceived as such a critical public good as emergency service or that they cannot be extended to the diversity of technical implementations and business models associated with VoIP; this is the heart of why there is an open research question involved in what we are doing.
4. Assess the Model. During the immediate scope of the project, we assess the model by critical and documented analysis of internal integrity and credibility with respect to successful precedents, an analysis that includes dissemination to policy oriented audiences for critique and debate, a process we are familiar with through professional experience in policy formulation. In particular we will validate our model against each of the major categories of VoIP implementation we identify and evaluate in (1). For the longer term, the analysis in (1) and (2) will not only motivate our particular model but also specific strategies for assessing it once used; we will document these strategies. Elements would likely include (i) monitoring for technical and business model evolution that is out of scope of that envisaged in our model, (ii) tracking specific performance metrics of certified processes and comparing against criteria derived from consumer expectations and social policy goals, and (iii) voluntary adoption rates of certification compared against a template derived from past experience in industry sponsored certification processes.