STANDARDS AND PROTOCOLS

Background

Early commercial satellites were designed for a single purpose: to carry international, analog telephone traffic. Because these systems interconnected domestic telephone carriers, they had to adhere to telephony standards developed by the International Telecommunication Union's (ITU) former Consultative Committee on Telephony and Telegraphy (CCITT). Since satellites, in general, provided only a single service, it was sufficient for satellite service providers to participate exclusively in the ITU standards process. Even when satellites began to carry analog television traffic, the number of standards bodies that service providers needed to be concerned with was still relatively small.

During the past decade, however, deregulation, the advent of digital communications, and the onslaught of demand for global data services caused an unprecedented explosion in both the number and technical diversity of domestic and international standards bodies. The past few years have seen the birth of entirely new organizations like the ATM Forum, the rapid growth of existing groups like the Internet Engineering Task Force (IETF) and the wholesale re-engineering of venerable standards bodies like the American National Standards Institute (ANSI), European Telecommunications Standards Institute (ETSI), and the ITU (see Figure 5.1, the snapshot of the standard bodies for satcom).



Fig. 5.1. Snapshot of the standards bodies universe for SATCOM.

For better or worse, these irreversible changes have completely transformed the old, cloistered standards community and created a vastly larger, faster-moving, and increasingly competitive international standards-making process. To remain at the leading edge of global communications, today's satellite developers, operators, and service providers are called to participate in an ever-growing number of standardization activities. In comparison to the "early days" of satellite communications, by some estimates the average amount of corporate resources that is allocated to standards participation-to the monitoring, developing, and assuring compliance with emerging standards-has grown at least one hundred-fold since the first commercial satellite transponder came online!

The changes that have swept the communications marketplace over the past decade are truly revolutionary. Service providers are moving away from the vertically integrated "stovepipe" systems of yesterday's network-oriented offerings toward a more consumer-oriented set of offerings characterized by a much richer mix of communications options wherein interoperability is the key word. The challenge now for the satellite industry is to develop the business strategies and technical capabilities that will enable satellite systems to effectively inter-operate with the terrestrial infrastructure as well as other satellite systems. New system architectures will be needed, and new or modified protocols will be required to provide the degree of interoperability that consumers will demand of future systems.

Trends

The tremendous demand for data services has brought to the marketplace a variety of solutions. Even in the broadcast arena where advanced digital systems are poised to replace the venerable analog TV systems, data services are being offered. The FCC, in December of 1996, adopted a recommendation by the Advanced Television Systems Committee (ATSC) for next generation digital television for terrestrial systems (ATSC Recommendation A/53). In doing so, the FCC established for the United States a standard for the development and deployment of high definition television (HDTV). This is a comprehensive standard that defines audio and video digital signal compression formats, packetized transport structure, and modulation and scanning formats. It also defines the digital data structures that will allow broadcast signals to carry ancillary data channels, a move which will help to accelerate the deployment of a whole host of datacasting and multimedia applications-all of this forterrestrial systems. In December 1997, the FCC was scheduled to begin examining responses to its Request for Information (RFI) in preparation for issuing a similar standard for broadcast satellite systems. The eventual adoption of such a standard is certain to create new opportunities for satellite systems, not only as a program distribution medium, but also as a key player in the direct-to-home (DTH) applications market.

Data communications is becoming an increasingly large component of the telecommunications budget for businesses. Consequently, businesses are especially eager to find ways to expand their data networks and at the same time reduce their overall communications costs. One solution that is gaining widespread acceptance is for businesses to replace or augment their traditional dial-up or leased-line services with Internet protocol (IP) based intranets and virtual private networks (VPNs). VPNs allow voice, and in some cases video, to be integrated along with the data services into a single local area network (LAN) and/or wide area network (WAN) service. The widespread availability and relatively low cost of IP based packet technology makes VPNs attractive for a wide variety of services including voice, voice mail, email, fax, document distribution, software updates, inventory management, and on and on for both internal and external communications.

Currently, most businesses with low to modest throughput demands (i.e., roughly 6 Mbps or less) find that Frame Relay and integrated service digital networks (ISDN) are the most cost-effective means of interconnecting their geographically dispersed offices, and of linking up with their business partners. Several other options, however, are being explored as well. For example, ADSL (asynchronous digital subscriber loop) technology and its variants-mainly SDSL (symmetric DSL), HDSL (high-bit-rate DSL), and VDSL (very-high-bit-rate DSL)-are being deployed by many local telephone service providers as an alternative to traditional T-carrier services like the 1.5 Mbps T-1. Typically, ADSL can provide up to 9 Mbps downstream and up to 640 kbps upstream; VDSL, on the other hand, can provide up to 51 Mbps downstream and up to 1.6 Mbps upstream. Its popularity is such that the ability of regional telephone companies to equip their central offices for xDSL services is a limiting factor in its deployment. The interest in xDSL services is matched only by that for cable data services. Businesses and consumers alike are installing cable modems where the service is available. The reason: cable data rates typically provide 10-30 Mbps downstream and up to 3 Mbps upstream, perfect for high speed file transfers and for streaming audio and video applications.

Meanwhile at the high end, other businesses, especially those with higher total throughput requirements, are adopting ATM (asynchronous transfer mode) as the preferred LAN and WAN technology. ATM is a highly adaptable technology, one that is particularly well suited for integrating voice, data, and video services. It can be used effectively at bit rates that range from the low to moderate rates employed for direct-to-desktop applications (i.e., 1 Mbps and below) all the way up to multi-gigabit per second rates employed for trunking applications. Nevertheless, the business community is not rushing to embrace ATM. The relatively high cost of implementing an ATM network is a major deterrent. If and when the price of ATM routers, bridges, and interface hardware come more in line with the price of Frame Relay equipment, ATM is likely to take over as the technology of choice. In the interim, Frame Relay, ISDN, and xDSL for low to moderate bit rate applications along with Gigabit Ethernet for high data rate backbone applications will most likely persist.

This proliferation of network technologies brings with it the need to develop interoperable solutions. That includes interoperability among the various network technologies and interoperability among the various delivery systems, e.g., wired and wireless systems, cable and twisted pair systems, satellite and terrestrial systems, etc. One approach is to develop hybrid networking concepts. The ongoing activity to combine IP-based and traditional telephony services is a prime example. The objective here is to cross-enable circuit-switched and packet-switched services. The result could include: (1) voice-over IP networks, (2) supplementary telephony services over IP networks such as call waiting, call forwarding, caller ID, and 3rd-party calling, (3) lifeline services (e.g., 911) over IP, (4) multimedia over IP, (5) fax over IP, and (6) voice access to IP (esp. Internet) content. Alternatively, it would result in circuit-switched access to IP-based (esp. Internet) information content. Still other concepts are evolving that would further blur the distinctions between circuit-switched and packet-switched networks, and between connection-oriented and connectionless networks.

Standards bodies are now engaged in the process of turning this vision of universal access into a reality. Efforts are targeted at evolving the necessary networking by the year 2000. In Europe, ETSI is laying the foundation for a Universal Mobile Telecommunications System (UMTS) that would unify cellular, GSM, cable, wireless local loop, and satellites with the fixed network. A similar effort is underway in ITU activities designated as the Future Public Land Mobile Telecommunications System (FPLMTS) project and the International Mobile Telecommunications in the Year 2000 (IMT-2000) project. As a part of this work, new network management and control standards must be developed for these expanded services.

Concurrently, the concepts for universal access by anyone, anywhere, anytime as envisioned by Vice President Gore in his 1995 landmark address on the National Information Infrastructure/Global Information Infrastructure (NII/GII) are beginning to acquire substance. From the onset there has been considerable debate as to what the essential elements of the NII/GII are, what the architectural structure of the NII/GII is, and what the basic service capabilities should be. In September 1997, a Canadian standards body, the Telecommunications Standards Advisory Council of Canada (TSACC) was the first to produce a document that defines these top-level concepts. Its purpose is to guide the development and deployment of the Canadian Information Highway. Similar undertakings by the International Organization for Standards/International Electrotechnical Commission (ISO/IEC), the Joint Technical Committee (JTC1), Special Working Group on Global Information Infrastructure (SWG-GII) and by T1P1.1 are nearing completion. Likewise, in early 1998, ITU-T SG13 is expected to issue a set of documents that will establish the basic principles, framework, and architectures for the GII.

The Issues

With the top-level concepts for the NII/GII now beginning to solidify, standards bodies are beginning to focus on the next tier of issues associated with the make up of the NII/GII. Expert groups within the various standards bodies will thus begin to hone in on the technical details that will guide the implementation of the various networking technologies. In the past, these expert groups have been dominated by terrestrial network service providers and equipment manufacturers. Without a more active participation on the part of the satellite industry, the ensuing standards may not be particularly satellite-friendly. That is the case for many of today's communications protocols, which are largely fiber-centric. Frequently, requirements affecting bandwidth utilization, bit error rate performance, and latency (i.e., propagation delay) needlessly find their way into the standards and unduly hamper the design or applicability of satellite systems. Clearly, if satellite systems are to be effectively integrated into the communications fabric of the NII/GII, the satellite industry must play an active role in the development or modification of these standards.

Among the most pressing protocol issues to affect satellite communications are several associated with ATM. Among the biggest issues in this regard is one that has to do with the way voice services are carried. Because of ATM's overhead structure, a standard 64 kbps digital voice channel, if not compressed, would expand to approximately 80 kbps. For satellite systems, where bandwidth is a precious commodity, the adoption of a standard that incorporates voice compression and silence suppression would eliminate the channel inefficiencies that would otherwise result. An equally important ATM-related issue is that of quality-of-service (QoS) requirements. Here the propensity within the standards community to arbitrarily impose fiber-optic consistent requirements on the standards has unnecessarily burdened various ATM services classes. Studies have shown, for example, that the requirements regarding cell loss and cell error ratios for MPEG video applications need not be as stringent as fiber bit error rates would support.

Another pressing protocol issue affecting the ability of satellites to carry data file transfers is that of the TCP/IP traffic management protocol. Current implementations of the TCP/IP protocol, which were developed for low-latency terrestrial applications, include a rudimentary slow-start congestion avoidance control mechanism and relatively small data buffers. For high latency satellite links, the effect of the slow-start mechanism combined with the small data buffers severely limits the data transfer rates. Solutions must be found that would enable satellites to handle file transfers as efficiently as low-latency terrestrial systems (see Chapter 4).

Efforts to provide traditional telephony services over IP based networks bring with them a number of unanswered questions. In particular, there are issues to be resolved relating to call control procedures (i.e., information flows and protocols), IP-to-E.164 address translation, charging/billing, end-to-end quality-of-service requirements, and the traffic loading effects that IP protocols could have on PSTN and satellite networks. Similarly, the are many unanswered questions regarding circuit-switched access to IP based (e.g., Internet) content.

Communications Satellite-Terrestrial Interoperability Issues

The interoperability of satellite and terrestrial communications facilities is increasingly becoming a major issue. In the past, satellites tended to be used in a stand-alone mode or were integrated into the voice network with few problems. As we look into the future, it is becoming increasingly evident that communications across the terrestrial and satellite facilities needs to occur seamlessly to create a large global network that is capable of handling low as well as high bit rate traffic. The first issue is the mismatch between the bandwidth of satellites and fiber. Present commercial satellites tend to have a maximum bit rate capability of about OC-3, or 155 Mbps, while a single fiber can handle 2.5 Gbps, or many times that when DWDM is used. Another important issue is the use of protocols that were designed for the low latency of the terrestrial fiber network and not the round trip delay of GEO satellites. The use of different standards between the United States and the rest of the world is an issue for both the terrestrial and satellite facilities. As mentioned in Chapter 4, considerable work is being directed at solving the protocol issue. The system, or interoperability of terrestrial and satellite facilities, to form a single global network, is the subject of numerous experiments around the world. These experiments are driven by either application and service programs or standards and technology programs (see Figure 5.2). The experiments are listed in Table 5.1.



Fig. 5.2. Application and service programs/standards and technology programs.

The Year 2000 Problem

The year 2000, or as it is typically called, the Y2K problem, has not gone unnoticed in the satellite community. Several transponder service providers have studied this and stated that they will not be affected. Satellite manufacturers also believe that there are no known problems since dates are not included in any of the onboard software or firmware. The Global Positioning System satellites are Y2K compliant and the support systems will be compliant before 2000. If there is a satellite Y2K problem, it will probably originate in the earth stations. Of concern is the software that controls the satellites, but this is being studied in great detail. Another area of concern is the end user networking software. All public announcements by the service providers indicate that they are Y2K compliant.

Table 5.1
Experiments in Interoperability-U.S. and Europe
 

Technology-Driven

Service-Driven

United States

  • GIBN
  • COMSAT ATM
  • INELSAT ATM, SDH
  • NASA/TIA CIS
  • NGI
  • KA-band
  • US/Japan HDVN
  • NREN
  • NYNEX API
  • W-ATM
  • Hughes DirecPC
  • Sydaya Frame Relay

Europe

  • GIBN
  • ESA/ESTEC CODE
  • EUTELSAT ATM, SDH
  • BT SDH
  • RACE Sat/ATM
  • COPERNICUS
  • NICE
  • ACTS
  • CSELT SDH
  • SECOMS
  • TOMAS
  • SINUS
  • ABATE
  • Multiserve (GMD)
  • DIGISAT
  • ISIS
  • VANTAGE

Asia-Pacific

  • GIBN
  • CRL NII/GII Testbed
  • CRL ultra-high speed and MM testbed
  • APECAPII testbed
  • KDD ATM, SDH
  • Cable & Wireless ATM, TCP/IP
  • Sony ATM MPEG2 DTH
  • Sony MM CATV
  • AsiaSat MPEGs

Canada

  • BADLAB
  • Teleconferencing



Published: December 1998; WTEC Hyper-Librarian