Service Provider Interoperability and the National Information
The vision of a national information infrastructure (NII) introduces many scenarios that imply the use and assumption of flexible, interoperable telecommunications services. The existing telecommunications infrastructure offers interoperability between local exchange carriers (LECs) and interexchange carriers (IXCs) for traditional circuit switched and dedicated services and among peers at agreed exchange points in the case of the Internet. The anticipated evolution of commercial telecommunications services does not guarantee that competing service providers will interoperate.
Current trends compromise the basis for interoperability among the providers of the telecommunications infrastructure. The merging of LECs and IXCs toward joint competition suggests a model similar to today's IXC environment, with its extremely limited interoperability. The commercialization of the Internet offers a second example where the Internet policy of free exchange of data tends to become compromised by the growing call for settlements between carriers. Early experience with advanced services such as ATM and frame relay, likely key enablers of the NII, also suggests a reluctance on the part of competing carriers to interoperate. For example, the long-term availability of X.25 public data networks has not resulted in interoperability among competing service providers.
The recommendations suggest several possible approaches to ensure the continuation and expansion of interoperability among telecommunications service providers. A first approach examines past models of interoperability, (e.g., for voice service or the Internet) and examines the suitability of these models to an NII. As technology-related issues will come into play, prototyping activities could be established to define interfaces, address performance, and consider management between service providers. The Network Access Points (NAPs) established by the National Science Foundation may also be considered as a government-sponsored vehicle for interoperability. Finally, it is advisable to continue the analysis of policy and regulatory reform currently under way to ensure continued interoperability among service providers in the NII.
Statement of the Problem
The vision of a national information infrastructure (NII) introduces many scenarios that imply the use of a flexible, interoperable telecommunications infrastructure. The existing telecommunications infrastructure offers interoperability between LECs and IXCs for traditional circuit switched and dedicated services and among peers at agreed-upon exchange points in the case of the Internet. The evolution of commercial telecommunications services does not guarantee that competing service providers will interoperate. In fact current trends compromise interoperability among future providers of telecommunications services, thereby creating the potential for many islands where exchange of information beyond island boundaries becomes difficult, less responsive, and uneconomical.
The following sections offer a brief background of the broad and diversified telecommunications market. The first section summarizes the current service providers (with a focus on local and interexchange carriers) and uses the Internet as an example of telecommunications services offered outside of the traditional carrier envelope. The second section describes the generic services currently available and focuses on the interoperability relationships in today's environment necessary for user-to-user communication. Interoperability today typically occurs at the lowest common denominators of service; the third section therefore provides an example of future services available to the NII. Interoperability among service providers in the realm of advanced services appears to be a significant challenge.
Telecommunications Service Providers
The telecommunications infrastructure is composed of an extensive set of physical communications facilities owned by a relatively small set of telecommunications companies. The LECs and the IXCs provide most of the bulk of the actual physical facilities that deliver telecommunications services. However, a broader definition of infrastructure is offered here that includes a large set of service providers that leverage the physical infrastructure to deliver service. This latter group includes the carriers themselves but also involves many organizations that build services based on specific markets or communities of interest. The Internet is the most prominent example of this latter group, where several thousand networks make up the Internet and only a small subset of the LECs and IXCs actually provide Internet services. In addition, new elements of infrastructure have emerged in the areas of cable infrastructure and wireless media. Although many of the cable and wireless providers have become tied to traditional carriers, their presence implies a strong potential for telecommunications evolution. With the exception of the Internet, the existing infrastructure offers limited interoperability, principally between a federally mandated tier of service providers and at the lowest common levels of information transport.
Local service implies that a service provider is principally restricted to offering connectivity within local access transport areas (LATAs). The local service area continues to be dominated by the regional Bell operating companies (RBOCs). However, many local service providers include traditional service providers and an emerging class of competitive access providers (CAPs). The traditional service providers offer RBOC-like services in independent areas of the country and include a broad spectrum, from large providers such as GTE and Sprint Local (both over 100 years old) to small providers such as Splitrock Telephone Co. of South Dakota. Traditional services offered by the RBOCs and other local providers have increasingly featured voice services, private lines, and a gradual influx of local data services. CAPs, such as Metropolitan Fiber Systems and Teleport Communication Group, have focused on the provision of dedicated access between high-density customer locations (e.g., government and business offices) and the IXCs. The CAP focus can be attributed to the strong business opportunity for local access and the large investment required to reach private residences. CAPs have also gradually increased their portfolio of services offered to include data services. The most prominent upcoming change in the local service area is the likely removal of restrictions on the RBOCs for the delivery of services beyond the local area.
The IXCs, or long-distance carriers, provide telecommunications services between the LATAs. The divestiture of the RBOCs by AT&T in 1984 created the opportunity for new long-distance service providers and has resulted in the creation of several prominent competitors of AT&T, principally MCI and Sprint, and a strong cadre of growing service providers, such as Willtel and LCI. The resulting competition has brought rapid technology evolution, new services, and lower costs to long-distance services. The long-distance infrastructure has also become highly robust and diverse, currently including multiple high-capacity networks with independent control structures. For example, a typical span in a major carrier network may consist of several active fiber pairs, each carrying in excess of 1 gigabit per second (Gbps) of capacity. The IXCs make up multiple cross-country (east to west and north to south) routes that, taken together, represent hundreds of Gbps of capacity. The IXCs support many telecommunications services, dominated by voice (especially for the top three providers) and private line. In addition, the IXCs offer various public services for data or video-based services.
During this period of competition in the long-distance market, the local carriers have played a key role as the common exchange point among the services provided by the IXCs. However, in the local area, where the traditional telecommunications monopoly has remained in effect, technology, services, and prices have tended to lag behind the long-distance market, often significantly. Although the local carriers enabled the emergence of multiple long-distance carriers through a common and expensive local infrastructure, the capability gaps and higher prices in the local area have become considerable. The IXCs typically claim that approximately 40 percent of their total costs result from local carrier charges. The local carriers have moved to close these gaps as the promise of unrestricted competition has become feasible. The IXCs have also begun to posture for the entry of local providers through arrangements to build local infrastructures that will lessen the local monopoly of service.
The Internet is one example of telecommunications services that offer a value-added capability on top of the carrier infrastructure. It has become prominent over the last several years, principally due to an unabated growth rate. As a simple definition, the Internet is a collection of networks bound together using a common technology. The Internet originated with Department of Defense research on networking and sharing of computing resources. Primarily through government subsidies for research and education support, the Internet grew significantly through the 1980s and began to attract commercial attention due to its growth and strategic government and private investment. Currently the Internet has become a major source of commercial enterprise and continues to grow rapidly. An explosion of commercial Internet service providers has paralleled the growth of the Internet and offers wide area connectivity among the expanding population of Internet users and component networks. The Internet has also become a prime model for many visions of the NII, because of its ability to interconnect various technologies, people, and sources of information.
Overall, the telecommunications infrastructure, taken as an entity without regard for regulatory or business concerns, offers a multifaceted and ubiquitous platform for the delivery of services consistent with the NII. Current market and regulatory factors have placed restrictions on the delivery of services by local providers. As a result, local services tend to be less capable and more expensive than those provided by IXCs. This status quo has begun to evolve rapidly as local providers prepare for a relaxation of their restrictions. The IXCs continue to enhance their abilities to deliver many services at ever-lower prices. Numerous service providers, best represented by the Internet community, have capitalized on this infrastructure to provide effective information services.
It is also worthwhile to briefly review the services delivered by the service providers. The telecommunications infrastructure supports three types of connection that enable a wide variety of services:
Especially important to this paper are the relationships between the providers of these services that enable interoperability from user to user. Although transparent to a user, virtually all telecommunications services in fact require an exchange of information between multiple service providers. Today a subtle but complex web of relationships addresses the information transfer, operational support, and business mechanisms necessary to make telecommunications a reality. Interoperability among carriers has typically settled to the least common denominators of service such as switched voice or private line services. The demand of the NII, for more robust, featured services, poses a more difficult and a new challenge of service interoperability to the growing cadre of service providers.
Telephone-based voice services are by far the dominant mode of telecommunications. This dominance can be expressed in many ways, including percentage of infrastructure consumed, ubiquity of service, and generation of revenue. Voice services have evolved over the past 100 years to where significant complexities in the delivery of service have become transparent to the typical user. The voice service infrastructure also supports a large proportion of all data-oriented services through its support of modem-based communications. In general, voice services involve two handoffs between carriers: A local carrier passes a call to the IXC, and the IXC passes the call to the destination local carrier.
Telecommunications carriers also provide private line services, which allow users to connect separate sites with basic transport services. A private line represents an extension of the carrier infrastructure to an individual user; the carriers devote a dedicated time slot of their network to that customer. A user may attach a large variety of customer premises equipment (CPE) to the private line to support virtually any type of service. Large users of telecommunications use private line services as the cost of dedicated services tends to be high relative to that for switched services such as voice (note that while customers only pay for voice services when they are in use, private lines incur flat rate charges independent of usage). Paradoxically to the higher relative costs, private lines represent the least possible value added by the service provider. However, private line services provide the underpinning of virtually all services. For example, the Internet has long been based on the connection of devices that implement the Internet protocols (i.e., IP routers) with private lines.
In fact, service providers distinguish their services from customer use of private lines by the movement of value-added equipment from the customer premises to the service provider premise. For example, in the case of Internet services, many large enterprises have deployed Internet technology within the enterprise as IP routers interconnected by dedicated services. In effect, these enterprise networks are analogous to a private internet (i.e., many enterprise networks interconnected). Conversely, an Internet service provider deploys Internet technology within its infrastructure and offers service through dedicated connections from individual user locations to the backbone network. Similarly, enterprise networks based on local intelligent equipment such as smart multiplexers, X.25 equipment, or frame-relay technology continue to proliferate.
Public data services originated with the same government development activities as the Internet but moved directly to the support of commercial services. The emergence of public data networks based on the X.25 protocol suite through the 1980s created a global infrastructure for data transport. However, unlike the Internet, X.25 networks tended to focus on services for individual users over a virtual private network, rather than between independent networks. These virtual private networks (a concept later extended to voice services) allowed many users to share the backbone of a single service provider. These X.25 networks have recently found new life as the access vehicle of choice for consumer information services, such as America Online, owing to their deployed dial-up facilities. X.25 networks also offer a working model for carrier network interconnection in the form of the X.75 gateway, the technical designation for interconnection between X.25 networks provided by different public-service providers. However, while X.75 offers the technical mechanism for interoperability, competing service providers have been reluctant to deploy X.75 connections. Public data services have also reemerged in the form of frame-relay and ATM services.
Similar to the emergence of public services for data communications, carriers have initiated the delivery of public videoconferencing services over the past several years. A video-based service takes the form of providing telecommunications service between public or private video rooms. The video rooms, essentially conference rooms modified to support video teleconferences, are interconnected via carrier facilities. The carriers have offered reservation and gateway services to allow users to schedule facilities ahead of time and ensure that the video systems at both the source and destination ends will interoperate. The ability of a service to offer interoperability between disparate types of video equipment has become an important feature of the overall service. In a limited fashion, the video teleconference capabilities of the carriers support interoperability, although primarily in the context of specific user implementations rather than generic service features.
The NII can be expected to increasingly take advantage of a wide range of service capabilities. Today's common model for the NII, the Internet, tends to focus on data exchange modalities. Conventional telecommunications services offer limited models for interoperability among competitive carriers or for advanced services. We can expect that the NII will increasingly embrace multimedia services that will require a robust
service infrastructure and place demands on service provider interoperability that exceed the scope of existing interoperability models.
Emerging Infrastructure and Services
The strong growth and competition in the telecommunications industry have accelerated the introduction of advanced technologies to the existing infrastructure. The result is a dynamic set of services that offer greatly enhanced services and new capabilities to the NII. These capabilities will begin to change the manner of delivery of information services, likely with significant improvement in price and functionality. However, the early definition of these services has tended to follow existing models for carrier interoperability (i.e., via private line or eventually local carrier interconnection to the IXC). Rather than attempt a comprehensive review of emerging technologies, the following section discusses the potential of asynchronous transfer mode (ATM), a good model for the introduction of new telecommunications services enabled through advanced technology.
Since its definition in the mid-1980s, ATM has represented something of a holy grail to the telecommunications community. ATM, as defined, supports all communications services and thereby represents a merging of the telecommunications and computing portions of the information technology industry. The formation of the ATM Forum in 1991, now including over 700 members from all segments of industry and government, sparked the development of ATM specifications to support early ATM implementation. The emergence of early ATM capabilities in late 1992, along with ATM's steady growth over the past 30 months, has created considerable excitement as the harbinger of a new era of networking consistent with the NII. Over this time ATM has taken a preeminent position on the leading edge of the telecommunications industry. ATM simultaneously offers service providers more effective operation and users greater flexibility. Service providers, from large public carriers to the telecommunications groups of private organizations, have embraced the ATM concept as a means to streamline internal operations and improve the delivery of user services. For example, ATM will enable carriers to significantly reduce overall provisioned bandwidth and allow instantaneous provisioning of service through the statistical multiplexing capability of an ATM backbone. Although fiber-optic networks have greatly reduced the overall cost of network capacity, carrier investments in network facilities still represent billions of dollars in investment and operating expense. The ability of an ATM approach to reduce costs and improve operations constitutes a significant market advantage. In addition to enhanced internal factors, ATM will enable service providers to deliver advanced services at a fraction of current prices and create new markets for services.
The vision of an ATM-based network also offers significant advantages to the user community. Existing telecommunications services mirror the carrier networks, where multiple independent networks share a common time division substrate to deliver separate services. For example, the voice, Internet, and video services mentioned here exist as separate networks within the carrier facilities, requiring separate capacity, operational support, and management. Users tend to have a relatively small set of options for service, which are rigidly defined. ATM offers a dramatically different view. A customer may send any type of informationvoice, video, data, image, or multiple types at onceat transfer rates from 1 to 1 billion bits per second. ATM also promises independent quality of service guarantees for each application, thereby allowing users to select services within their own context of budget, urgency, and quality, all on a dynamic basis. Despite the great promise of ATM, true interoperability of ATM services offered by competing carriersthat is, addressing technological, operational, and business concernsappears several years away.
The evolution of information technology has significantly enhanced the capabilities and range of information services. Numerous reports and articles have already described emerging information services, but three trends become important to future service provider interoperability:
The continued evolution of information services and their combination with telecommunications will strongly influence the NII. The influence will be seen in the changing nature of competition between telecommunications carriers, information providers, and new hybrids of the two. We can expect that commercialization of information services will affect interoperability as a market driven by information content may evolve.
Analysis and Forecast
Interoperability among telecommunications carriers is a key attribute of the infrastructure. For example, voice or switched services and private line and Internet services all support interconnection between users connected to different carrier services. Meanwhile, private networks and emerging services tend to struggle for interoperability beyond specific provider boundaries. The current models for interoperability among service providers offer a reasonable baseline today. However, key trends such as regulatory changes, market forces, and technology development are emerging that affect the continued assumption of interoperability.
Current Models for Interconnection
There are three significant models for service provider interoperability that may come into play within the future NII. Since divestiture, local carriers have set up agreements with IXCs for the exchange of services. These agreements have focused on two types of service: switched service supporting voice; and private lines that transport many types of information, with little value added from the carriers aside from provision of the connection. Although these arrangements are technically sophisticated, the most important feature involves the business mechanisms that track the exchange of information and convert the volume of information to financial settlements. Meanwhile, the data community has set up interoperability models consistent with data exchange. The X.75 protocol was developed to address the exchange of information between national public data networks that employed the X.25 protocol. The Internet is thriving and is synonymous with the concept of interconnected networks. It presents a powerful model for the NII, often used interchangeably with future NII concepts. However, today's Internet does not support explicit business mechanisms for financial settlement but has begun to address settlements between service providers and is likely to evolve to some appropriate model. The existing models for interconnection offer limited guidelines for future service provider interoperability.
The most important instance of service provider interconnect, in terms of the volume of traffic handled, supports the connection of switched traffic, such as voice, and private lines between LECs and the IXCs. This relationship emerged with the divestiture of AT&T in 1984, the establishment of many local carriers, and the gradual introduction of competition in the IXC area. The interconnects are mandated by the agreements surrounding the divestiture providing for equal access to local facilities to any IXCs prepared to provide service.
The LEC-to-IXC interconnect employs sophisticated technology oriented to the establishment of connections on a dynamic basis between users that often span three service providers. Both LECs and IXCs establish multiple connections between their networks and other carriersfor example, Bell Atlantic supports interconnects to Sprint, AT&T, and MCIand the interconnect discipline has thus been replicated many times. Each of the carriers has invested heavily in improving the signaling that controls these interconnects. Improved signaling both increases performance (e.g., in the sense of call set-up time) and enhances feature availability. In addition to the technology inherent in these interconnects, the establishment of business mechanisms that measure traffic exchange and enable financial settlements represents a major contribution to the interconnect model. The volume of both the traffic (tens of billions of call minutes per month) and the financial exchange (billions of dollars per year) makes the current LEC-IXC a powerful model for interconnect.
The LEC-IXC model does not necessarily represent the best possible approach. The interconnect is mandated through regulatory means rather than through the dictates of an open market. The absence of any substantial interconnect scheme between competing service providers suggests that the current scheme may not extend well to competitive-service providers. For example, interconnection between Sprint, MCI, and AT&T is mostly nonexistent; it is unclear whether typical LEC to IXC agreements would translate to two IXCs.
The technologies associated with this interconnect have not been shown to adapt well to other types of services. The IXC-LEC emphasis on the lowest common denominators of telecommunications, switched voice services and private lines, does not extend well to an NII vision that takes advantage of various advanced telecommunications services. Recent efforts have begun to address the interconnection of advanced services such as ATM and frame relay between LEC and IXC. However, operational interconnects have only reached planning stages, with issues like service guarantees and appropriate financial settlements still poorly understood.
Public Data Network Interconnection: X.75
A second form of carrier interconnect in place since the early 1980s and widely deployed is the gateway between public data (X.25-based) networks, referred to as X.75. This technology applies to gateways that may exist between public X.25 data networks, often those operated by national telecommunications carriers (i.e., France Telecomm; in the United States, the primary X.25 carriers are MCI and Sprint). The X.75 interconnect offers an historical model for large data service provider connection in a commercial environment. Though limited in functionality and becoming obsolete, the X.75/X.25 model is important because it supports the data environment, involves financial settlements between carriers (including international entities), and bears some similarity to the emerging ATM environment (i.e., it supports connection-oriented service).
X.75 operates by setting up virtual connections from user to user as a spliced connection of multiple virtual connections. In this sense X.75 represents the opposite of the Internet IP protocol, which uses a connectionless model for service delivery. X.75 also thereby tracks current efforts toward the development of ATM services, which is also connection oriented and a potential model for future services. The X.75 gateways also include mechanisms for the tracking of data exchange and settlement procedures between networks. As an interconnect between public networks, the X.75 model has the added advantage of defining interconnects between international entities. Many countries still operate a single, monopolized telecommunications carrier and therefore a single public data network. Notably, the evolution of public data networks has tended to downplay the interoperability offered through X.75 connections. The expansion of global services by primary carriers has created direct competition between service providers and resulted in reduced desire for interoperability.
The Internet offers an interesting model for interoperability. From its inception, virtually anyone could connect to the Internet, given compliance with the set of procedures and formats dictated by an evolving set of Internet standards. By the late 1980s an architecture emerged that included a backbone transit network, the NSFNET (which supported connectivity between regional service providers) and an acceptable use policy (which restricted traffic to research and education). The NSFNET exchanged traffic with other major government backbone networks, through Federal Internet Exchanges (FIXs) on the East (FIX-east) and West (FIX-west) coasts. The FIXs provide a technical and policy-based interchange point, but they do not include mechanisms typically found in business environments for collection of traffic data and billing.
The growth of the Internet has precipitated commercialization. At first, small service providers and then gradually larger service providers (including Sprint and then MCI) developed services that employ Internet technologies and offer connectivity to the Internet (a combination usually referred to as Internet service). The commercial service providers created a Commercial Internet Exchange, the CIX, to allow the exchange of commercial, policy-free traffic. The shift to commercialization, with the government as a major catalyst, is prompting the gradual withdrawal of government influence. Through the NSF's solicitation 93-52, the NSF has decommissioned the NSFNET as the transit backbone for the domestic Internet and promulgated a commercial infrastructure that includes three primary Network Access Points (similar to FIXs) and commercial network service providers (NSPs). Both the NAPs and the CIX are provided by third-party vendors (the facilities for each are actually provided by carriersSprint, Ameritech, PacBell, and MFS) and support the exchange of traffic.
At this time there are limited barriers or policy for connection at a NAP or a CIX. The NSF has set a policy that any entity funded by the NSF must employ an NSP that is connected to all three primary NAPs. In this way the NSF has ensured that the research and education of the Internet will be fully connected, at least for the 2 to 3 years planned for the NSF funding. Within each access point members must establish separate bilateral agreements to facilitate interconnection. In some cases, service providers have included placeholders for settlement agreements in these bilaterals; however, no agreements have been set at this time.
Attributes of Service Provider Interconnection
The preceding profiles of major interconnection methods suggest several key attributes for service provider interconnection.
Key Trends Affecting Carrier Interoperability
Today's telecommunications environment exists through a web of interoperability agreements among numerous service providers. In general, these agreements exist to facilitate the fundamental exchange of information. In virtually all cases the agreements exist between noncompetitive entities where cooperation benefits both parties. However, many commercial and regulatory trends have begun that threaten this basis for service provider interconnection. The expected relaxation of constraints on LECs for the provision of long-distance services will create a new competitive relationship among LECs and IXCs. Both groups have rushed to form strategic relationships and extend capabilities to prepare for this new era. In addition, as telecommunications services have become more of a commodity, service providers have looked for ways to differentiate their services through technological improvements as well as in combination with information services. The Internet is undergoing a parallel change where the trend toward commercialization will create competitive relationships among previously complementary entities.
Current trends in the telecommunications regulatory structure suggest strongly that local carriers will begin to extend their role beyond local transport services to include wide area connectivity. The LECs have already won the right to provide long-distance services for cellular-based customers. Local carriers will also continue their expansion into other nontraditional service avenues, especially in the area of information service. Although the pace of this evolution of the LEC role is unclear, it is likely that the local carriers will tend to dominate the local infrastructure for many years to come. Despite major IXC investments, the extensive, in-place LEC connectivity to every residence will be extremely difficult to duplicate. Early in this evolution, it therefore appears likely that the LECs will dominate locally and have multiple options for long-distance exchange.
The IXCs have also begun extensive efforts to move into alternate forms of service, including Internet services, local service, and cellular/wireless-based services. The IXCs have recognized the large portion of costs apportioned to local service, along with the growing threat of LEC competition in the long-distance arena, and have begun to focus on the extension of facilities to local areas. For example, a major IXC announced a partnership with three major cable companies and a CAP in 1994. These activities recognize the likelihood that the IXCs will soon directly compete with the LECs for customer-to-customer service. The evolving roles of LECs and IXCs suggest the possibility of a paradigm shift in agreements for traffic exchange between carriers. For example, today the flow of financial settlements travels in one direction, from the IXCs to the LECs; future environments where today's LECs and IXCs become competitors suggest a two-way flow.
Both LECs and IXCs have also recognized that simple transport services will become more like commodity services. As a result, many carriers have begun to expand their services to include advanced services and information capabilities. Advanced services, such as ATM, enable the service providers to create value with their offerings and differentiate themselves from other providers based on the features and capabilities of the services offered. The addition of information services on top of the transport presents a further opportunity for carriers to differentiate. A likely evolution of information services and telecommunications will bind the two together. The purchase of information services will therefore imply the selection of a specific carrier.
The commercialization of the Internet presents a parallel scenario for interoperability. As in the traditional carrier networks, interoperability is a defining feature of the Internet. However, the open exchange of information that created the Internet has occurred without the influence of business considerations. As the Internet has grown, its potential value to commercial ventures has become apparent. These trends have resulted in an Internet dominated by commercial service providers (note that Sprint's SprintLink service recently became the largest global supporter of the Internet, surpassing the NSFNET). As commercial ventures, Internet services still represent a small fraction of carrier revenues. Therefore, today Internet service providers exchange information more or less freely through agreed-upon exchange points without traffic measurement or settlements. However, as revenues grow, one can reasonably expect that competitive influences will drive the Internet toward more concise, business-oriented operation.
Interoperability tends to contradict service provider efforts to move away from commoditization. As the carrier networks tend to interoperate, they share features and functionality and offer access to the same sources of information, thereby reducing the differentiation associated with information-based services. Interoperability between emerging services, such as ATM, has also evolved slowly. Although this is partially due to the
immaturity of the technologies, the desire for stable customers also contributes to inertia over service interoperability. This is desirable from an NII perspective, but the creation of interoperable networks tends to counter marketplace tendencies among competitors.
Expansion of the roles and capabilities of LECs and IXCs will eliminate the current stratification between local and long-distance services. It will also erase the natural requirement for interoperability between LECs and IXCs. Financial settlements for the exchange of services will follow market-based formulas dictated by individual service providers. The larger carriers may be able to leverage their greater reach of services to force unbalanced settlements. Although the evolution of these trends is uncertain, the coexistence of parallel, competing service providers suggests an unstable environment that may affect end-to-end connectivity and the support of emerging NII applications.
The commercialized Internet will begin to mirror the evolving carrier environment over time. The major IXCs have already captured prominent positions in the commercialization of the Internet. As Internet services reach traffic/revenue levels comparable to those of mainstream carrier services, service providers will move toward settlement scenarios that will once again favor the larger providers. The traditional carriers own the facilities and a growing share of the expertise that make a growing, commercial, unsubsidized Internet feasible. However, numerous nontraditional service providers now make up a substantial portion of the Internet. Though these providers depend on the traditional carriers for physical connectivity, they have established strong foundations based on Internet-level services and a growing ability to delivery information services. Segmentation of the Internet appears highly unlikely, but the effect on connectivity is at least uncertain.
The possibility of limited interoperability among service providers, including those offering Internet services, should have a profound effect on the emerging NII. In principle, many of the applications contemplated in NII literature would still be feasible; but numerous obstacles would almost certainly emerge. For example, the concept of telemedicinethe delivery of medical services over the NIIimplicitly assumes a seamless telecommunications fabric. Otherwise, medical resources that reside on one service provider's network could not reach a patient on another network. Although this scenario is consistent with the desire for service providers to differentiate, it clearly threatens the realization of telemedicine services.
The recommendations presented here suggest several possible approaches to ensure the continuation and expansion of interoperability among telecommunications service providers. Overall, further analysis needs to be conducted on the nature of interoperability for the NII and the suitability of the existing models for this requirement. Today it seems that interoperability among competing service providers has been assumed as a natural attribute of a telecommunications infrastructure. On the contrarythere are numerous examples where today's competing service providers do not interoperate. Furthermore, the evolution of the telecommunications infrastructure in terms of technology, economics, and service relationships will tend to eliminate existing areas of interoperability and provide disincentives for further development of interoperable services.
The regulatory changes under way, one of which will relax restrictions on LECs, represent a major influence on service provider interoperability. Explicit consideration of the effect of interoperability of these events should be undertaken. In addition, serious consideration should be applied to additional regulatory actions necessary to ensure continued interoperability among service providers in the NII.
The Internet represents a national and global treasure of information. As the Internet continues toward commercialization, the effects on interoperability among service providers become less clear. The growth of the Internet has brought traditional service providers into the role of Internet service delivery. Although the traditional carriers play a key role in the scaling of the Internet because of their ownership of the physical infrastructure, they also bring the traditional service models and notions of interoperability. The future of the
Internet is widely debated, but specific attention should be given to interoperability as the Internet migrates to a fully commercial enterprise.
Numerous technological factors will affect the evolution of the NII and the interoperability among service providers. Prototype models for carrier interconnection for various services, technologies, and special user communities should be established to determine feasibility and preferred approaches for deployment in the future NII. The NAPs offer an excellent platform for these prototypes, as they represent by definition a means for interoperability among service providers.