Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Ascend Communications (ASND) -- Ignore unavailable to you. Want to Upgrade?

To: djane who wrote (47013)	5/16/1998 4:53:00 AM
From: djane	Respond to of 61433

This Year's 10 Hottest Technologies in Telecom (Part I) [ASND references in Part II] telecoms-mag.com What's on the radar screen this year for promising, new technologies that will drive the next wave of development? Read on and find out. Patrick Flanagan Technologies come and technologies go. Mostly, however, like the proverbial Top 40 hits, they keep on coming. At this juncture, the industry shows no signs of lessening the attempt to find better ways to conserve bandwidth, optimize transmission rates, and find new, more efficient, and less expensive ways to accomplish basic communications tasks. While it's true that there is some commodization taking place in local area and campus infrastructures, there are still plenty of challenges ahead as Internet telephony continues its paradigm-altering migration into the public network. As is the practice every year, we talked with a wide variety of experts, including the heads of leading telecom consulting firms. Their recommendations varied widely, but our qualitative approach clustered around a specific set of technologies where agreement was reasonably assured. In commenting on the technologies themselves, we have also tried to mention, where applicable, the leading-edge vendors who are putting these technologies out in the marketplace. Here in no particular order are Telecommunications magazine's 10 hottest technologies for 1998: 1. Multilayer Ultrafast Routing/Switching 2. Erbium Doped Fiber Amplifiers 3. Packetized Voice 4. LMDS for Broadband Interactive Services 5. Wireless Local Loop (WLL) 6. Internet Virtual Private Networks 7. Web-Enabled Call Centers 8. XML (eXtensible Markup Language) 9. TMN Technology 10. Java-Based Network Management Multilayer Ultrafast Routing/Switching: The Need for Speed With traffic on the Internet doubling every six months, backbone capacity continues to be an urgent carrier concern. The newest solution is a new generation of multilayer Gigabit and terabit switch routers that are emerging from the big four--Bay Networks, Cabletron, Cisco, and 3Com--as well as a number of start-up ventures. One impact is that the differences between routing and switching are disappearing on IP network backbones. The goal is to put as few switches and/or routers within the backbone infrastructure as possible, while increasing line speeds into the Gigabit and terabit ranges. Terminology to describe this emerging technology accurately is a problem. Currently, any number of descriptors are being used, including Gigabit switch routers (Cisco), multi-technology, Gigabit switching (Digital), terabit switch/router (Avici Systems), and Layer 3 switching (13 vendors ranging from Acacia Networks to Torrent Networking Technologies). The best umbrella term to describe the technology involved is multilayer switching/ routing because Layers 1 to 4 (SONET, frame relay, IP, and TCP) are all involved, as is what at least one vendor calls "Layer 0" for the deployment of wavelength division multiplexing (WDM) on fiber optic backbones. In addressing increased IP backbone capacity, the vendors are focused on three areas: performance, scaleability, and quality of service (QoS). The primary performance concerns of latency and network congestion are being addressed in terms of the impact on service levels. Recovery from congestion and efficient allocation of bandwidth are driving product development. Scaleability is being improved by technologies and platforms that can grow rapidly and elegantly as end users require more and more bandwidth from the core network. No longer is it possible to add another router or switch to resolve regional bandwidth problems. QoS is the most pressing concern. Without it, service providers cannot offer value-adds such as high-priority transmission of mission-critical applications and managed LAN/legacy services. The solution is for switches and/or routers to identify different service classifications of traffic and handle each efficiently without incurring large amounts of network overhead. One emerging technology for addressing these performance, scaleability, and QoS issues is packet-by-packet Layer 3 (PPL3) switches that act as full-blown routers at speeds up to 7 million packets per second. Another approach is cut-through switching, which doesn't peer into each packet as PPL3 does, but rather determines the destination from the first few packets in a series and then switches the flow to Layer 2 once a connection is made. Nick Lippis, president of Strategic Networks in Rockland, Mass., likes PPL3 because "it means really fast routing without having to change anything on your network." Other schemes abound, such as Avici Systems' Direct Connection Fabric (70-Gbps routing), the Pluris Massively Parallel Router (622 Mbps), Neo Networks' StreamProcessor (2.5 Gbps), and Torrent Networking Technologies' Shared Memory Switch Fabric (millions of packets per second). Argon Networks and Nexabit Networks are also developing products with similar capabilities. "These companies are coming at the architectural approaches for Gigabit and terabit rates from different angles, depending on their backgrounds, while they're all trying to crack the same nut," said Brendan Hanigan, an analyst with Forrester Research in Cambridge, Mass. Erbium Doped Fiber Amplifiers (EDFA): Enter the All-Optical Network Last year, wavelength division multiplexing (WDM) was a hot technology. Already, it is running up against limitations as technologies such as multilayer switching demand greater bandwidth. Today as many as 40 channels can be multiplexed on a single channel--a vast improvement from the original WDM capacity of four. However, each time the number of optical channels is doubled, there is a corresponding 3dB reduction in signal gain. Enter erbium doped fiber amplifiers (EDFA) with the ability to amplify several optical channels in the erbium passband, creating what is becoming widely known as dense wave division multiplexing (DWDM). There's new EDFA technology becoming available as well. Silica-based EDFA has been around for several years as a displacement for the need for costly electrical regeneration of an optical signal. Now the EDFA flat-gain amplifier with advanced filtering technology to provide a flatter gain curve is becoming available commercially. In the pipeline from Lucent Technologies' Bell Laboratories is a new amplifier based on silica EDFA and capable of supporting 100 wavelength channels having 100-GHz spacing. It uses two sections: One is optimized for long-wavelength channels beyond L band (1565 nm), while the other is optimized for conventional C-band channels (1525 nm to 1565 nm). The addition of a second gain-equalizing, fiber-grating filter provides uniform gain over the entire 80-nm spectrum. "EDFA is a key enabling technology in the continued evolution to the all-optical network," said Thomas Fuerst, a Lightwave product marketing manager for Alcatel Telecom of Richardson, Texas. One direction will be to expand DWDM network technology into metropolitan-area networks and eventually to LANs and local access. In addition to Lucent and Alcatel, several vendors are entering the EDFA marketplace. Ortel Corp. of Alhambra, Calif., has a number of fiber optic transmission products, particularly EDFA for 1550-nm broadband transmission and continuous wave pump lasers for EDFA. ADC Telecommunications of Minneapolis incorporates EDFA into its Homeworx HWX transmission platform. Optigain in Peace Dale, R.I., produces Model 2100 EDFA for SONET/SDH and DWDM systems as a pre-amplifier, booster, or in-line amplifier. Packetized Voice: Telephony By Other Means Packetized voice has reached the point that voice packets are delivered by cell or packet transport with the integrity and quality users expect from the public-switched voice network, and it is still rapidly evolving. Voice traffic is launched and received PC-to-phone, PC-to-PC, and phone-to-phone, with each possibility offering a slightly different application of the packetized voice technology. The marketplace is currently focused on pirating consumer voice traffic from traditional carriers by offering significantly lower prices while using the public Internet as the primary transport mechanism. The pioneers are start-ups such as Delta Three and VIP Calling, companies that worked the bugs out of the technology and educated the public. Now Internet service providers (ISPs) are entering the IP phone business. They see it as a way to increase revenues beyond the commoditized $20 per month they receive from all-you-can-eat Internet access services. The real potential lies in the corporate use of packetized voice technology over enterprise networks, particularly virtual private IP networks. Two examples of this are currently in operation. A pioneer in this area, NetWorks Telephony Corp. of El Segundo, Calif., offers a PC-to-any-device calling service that delivers calls globally on and off the Internet. The traffic is carried over Infonet's managed worldwide backbone network on which voice packets are prioritized using a proprietary multimedia cell technology. NetWorks Telephony is primarily marketing its services to ISPs, including the building of gateways at ISPs' points of presence (POPs)--a cost savings for ISPs that want to get into the IP telephony business. Infonet, also of El Segundo and which was spun off NetWorks Telephony as a separate entity, uses the same backbone for its Integrated Media Services that target multinational corporations; the company positions it as the "world's largest integrated voice, fax, and data network offering service and local support in more than 35 countries." There continues to be a lot of debate about the viability of packetized voice services. From the IP telephony viewpoint, Dataquest analysts predict a $3 billion market by 2001. The corporate user world is interested, according to analysts at Forrester Research, who found that 42 percent of 52 Fortune 1000 telecom decision-makers surveyed planned to "experiment with" IP voice and fax by 1999. Whether or not these predictions come true depends a great deal on how well the switched voice carriers defend their services. "These are the cheapest, highest quality networks in the world," said Thomas Pincince, founder and executive vice president of New Oak Communications, which is now part of Bay Networks. LMDS for Broadband Interactive Services: Cable Modems--Watch Out! Local multipoint distribution service (LMDS) is a wireless, two-way multichannel data, video, and telephony technology that could provide the long-promised, ubiquitous broadband telecom nirvana envisioned for the now rarely mentioned information superhighway. LMDS occupies one of the broadest expanses of radio wave spectrum devoted to any one service, a bandwidth of 1.3 GHz surrounding the 28-GHz Ka-band. This is a much higher frequency than most existing wireless applications. Baud rates are in excess of 1 Gbps downstream and 200 Mbps upstream, and since LMDS uses a very small cell configuration (2- to 7-mile radius), it is able to polarize and reuse spectrum in a highly effective manner over a small area. In addition, this enormous bandwidth means that LMDS is not hampered by the interference found with other wireless systems. (Line-of-sight blockages occasionally obstructed signals and storms caused "rain fade" in early deployments. Newer versions of the technology appear to have overcome these drawbacks.) Since there are no wires or cables to deploy and maintain, the economics of LMDS are one of this technology's strongest points. But unlike mobile wireless, LMDS works only between fixed points--offices, towers, homes, and other structures. Total sales for subscriber and network equipment are projected by Insight Research Corp. of Parsippany, N.J., to be $3.1 billion in 2001. The initial uses for LMDS are to distributed cable TV in areas where in-ground infrastructure is too expensive to install. Related to this is the use of LMDS as a less expensive way of bringing competition into entrenched cable TV and local telephony markets. The longer range prospect is for LMDS to blossom into the technology for linking corporate campuses to the enterprise network at speeds rivaling those now offered by fiber optic networks. The critical step of FCC licensing is now completed with an auction that netted $578.6 million. Venture capital-fueled start-ups dominated the auction, as was expected, because they got a 45-percent discount. Of the top 10 bidders, only US West Communications in 10th place was an established carrier. In part, this is because local exchange carriers, particularly the RBOCs, are only allowed to purchase 10 percent of the LMDS spectrum in their service areas. The top five bidders were WNP Communications, NEXTBAND Communications, WinStar LMDS, Baker Creek Communications, and Cortelyou Communications. WNP and NEXTBAND bid in excess of $325 million for more than 2 million POPs, compared to $156 million for the other eight top bidders. Predictions about when LMDS will become widely available vary greatly, particularly in geographic areas with a high density of existing broadband capacity. "Realistically, 1998 is a year of definition and positioning, and there will be many false starts. But next year, we will get a much better picture of what LMDS is and what it's useful for," said Richard Sfeir, director of marketing for CommQuest Technologies in Encinitas, Calif. Laurence Swasey, an analyst with Allied Business Intelligence in Oyster Bay, N.Y., predicted that business users will be pragmatic, "looking to utilize LMDS if there are no other high-speed carriage options available." The marketplace will get more competitive as well, with wider availability of cable modems, DSL, satellite systems, and wireless local loop. LMDS has an edge because it is "viable at low-take rates," which is industry lingo for the ability to make money with a low percentage of potential subscribers. The start-up costs can be as low as $16 per household within a typical service area with a population of 13,000, including seven business cells and 22 residential cells. Profitability could be obtained in the third year if 60 percent of major employers and 10 percent of residences sign up. Those are big ifs, but they present a much rosier scenario than can be constructed for any of the current in-ground broadband alternatives. Wireless Local Loop: Getting Unwired The idea behind wireless local loop (WLL) is "think small," which is the exact opposite of the usual wireless logic. At this time, there are three primary uses for WLL: to provide advanced services in urban areas and within business complexes, to replace wireline services in residential areas, and to bring telephony inexpensively to remote and underserved areas, particularly outside the United States. The technology comes in four flavors: analog, digital cellular, proprietary (broadband CDMA), and low-power, or cordless, radio systems. Today and in the future, the focus is on broadband CDMA and low-power radio except for emerging, or "first phone," countries. The universal strength of all forms of WLL is significantly lower costs than wireline services. On a per-line basis, the cost for WLL ranges from $500 to $1000 and is decreasing, while wireline telephony provisioning can run as high as $2500 per subscriber. The big WLL push for the future is in wireless office solutions deploying low-power radio technology. Hughes Network Systems, Ericsson, AT&T Wireless, AG Communication Systems, RogersCantel, and others are actively launching in-building systems. The workforce can be reached at anytime over a portable handset anywhere within the coverage area, within a single building or over a campus. When the user leaves the coverage area, the handset continues to function by being handed off to a public wireless carrier on either 850-MHz cellular or 1900-MHz PCS. In residential areas, WLL is being marketed as an alternative to wireline POTS, but one with greater available bandwidth for data, particularly Internet access. The hope is that the convenience of a single wireless phone installation for use in and out of the home, combined with lower rates for in-home use, will cause customers to abandon their twisted-wire POTS connections. Since 1995, AT&T has offered its AirLoop system as a way of meeting competition and curbing the maintenance and repair costs associated with traditional wireless infrastructure. Now a large number of manufacturers are involved, including Lucent Technologies, Qualcomm, Motorola, and Nippon. Estimates vary on how large the market will be for WLL. MTA-EMCI, a research firm in Washington, D.C., projects 60 million installations by 2000. In developing nations, the estimate is $125 billion in revenues by 2004, plus another $10 billion from the industrially developed group of seven (G7) nations. Melanie Posey, a senior analyst with Northern Business Information in New York, predicted that "the market will begin to take off after 1998, as an increasing number of wireline operators incorporate WLL in their network expansion plans and as new operators in competitive markets try for rapid market entry and value-added mobility." She expected WLL to account for 17 percent of all local loop construction by 2000, but noted that this number could be as high as 30 percent "if all cost, technical, and regulatory obstacles suddenly clear up."

To: djane who wrote (47013)	5/16/1998 5:02:00 AM
From: djane	Respond to of 61433

This Year's 10 Hottest Technologies in Telecom (Part II) [ASND references in Part II] telecoms-mag.com Internet VPNs: The QoS Conundrum Today's data communications traffic is beginning to resemble switched voice traffic--for a lot of reasons: Users expect to be connected at will to the enterprise network from their homes and while traveling; there is an escalating need for communications among business partners, customers, and suppliers; and there are increasing demands for access to the public Internet. Existing corporate private data networks cannot meet this demand for ubiquitous connectivity. The Internet VPN (virtual private network) is emerging as the new WAN architecture by carving out a private passageway through the Internet while using the public Internet backbone as an appropriate channel for private data communication. The objective is to create a single, seamless IP network for mission-critical applications as well as remote access, closed user groups, public Internet access, Internet-based commerce, and a Web presence. "This requirement for pervasive data communications connectivity cannot be met by existing corporate private data communications networks," said Michael Kennedy of Strategic Networks in Boston. Therefore, ISPs will become strategic suppliers, a move that is "inevitable," he said. The ISPs are pulling together the technology required to implement Internet VPNs in terms of backbone architecture, security, and configuration/management software. Among the providers are six major U.S. ISPs--AT&T, WorldNet, PSINet, GTE Internetworking (BBN), networkMCI, Sprint, and UUNet. Global ISPs are emerging as well, particularly Infonet, EQUANT Network Services International, Global One, and IBM Global Network. The key technology for successful delivery of the Internet VPN is interoperable end-to-end QoS guarantees for mission-critical traffic. "Too many QoS standards have been proposed," Kennedy said. He cited RSVP, Ipv6, ATM, QoS, and MPOA and predicted that a compromise of RSVP for signaling and ATM for wide area QoS delivery will prevail. "This confusion over standards will delay widespread adoption of Internet VPNs until the early 2000s," he said. Not all of the ISPs agree. Infonet, by partnering with Nortel and deploying its Magellan Passport switches across its private backbone, now offers a hybrid Internet VPN in 38 countries with three classes of service options. EQUANT has deployed the same technology across its global backbone. Cisco, Bay Networks, Ascend, Cabletron, 3Com, and many others are developing end-to-end QoS technology. The flaw in the Internet VPN is the NAP (network access point), where traffic crosses from one provider's backbone to that of another. "Once you hit a NAP, pray," said Timothy Kraskey, marketing vice president for Ascend's core systems division. There's a strong motivation to resolve the few stumbling blocks to widespread availability of Internet VPNs. While flat-rate Internet service is now a commodity, Internet VPNs are customized to meet specific needs. "This provides a now-missing opening for an ISP to achieve value pricing through targeted offerings," said Kennedy. Web-Enabled Call Centers: A Match Made in Heaven The marriage of call centers and World Wide Web sites is a match made in telecom heaven. The Web page visitor, the potential purchaser, or dissatisfied customer wants human contact when making a purchase or resolving a complaint. The dream solution--a mouse click on a Web page "talk" icon from a multimedia PC--creates a packetized voice channel, or hot link, to the call center over the same circuit in use for the Internet connection. "This is the modern equivalent of the door-to-door salesman's foot in the door," said Kennedy of Strategic Networks. The technology behind Web-enabled call centers is essentially the same that permits IP telephony over the Internet. A key component is the sound card now routinely supplied with a multimedia PC and the device that converts the analog voice signal into IP packets. Near toll-quality is achieved at 5 kbps (compared to the public-switched network's 64-kbps standard) through digitizing algorithms and IP packet switching that intrinsically exploits the pauses and silences of speech. Products for Web-based hot links are rapidly rolling out. Among the recent announcements are MCI's Call Center Connection for 800-number business customers; Infonet's GIS voicelet service; the Call Center solution developed by Executone Information Systems in collaboration with VocalTec Communications and Dialogic Corp.; and Siemens GEC Communications Systems' Call Server. AT&T and Sprint also have announced Web-enabled call center products. XML: Raising the Web's IQ One of the better- kept secrets of the Web is that hypertext markup language, universally known as HTML, is pretty dumb. There's a better option ready to debut: eXtensible Markup Language, or XML. The fatal flaw in HTML is that it has no attributes for handling the business world's complex and endless demands for searching the Web. HTML makes Web pages easy for humans to read but hard for search engines to decipher in a meaningful way, particularly the pinpoint culling of information to the specifications of demanding Web surfers who want 10 highly specific references, not 150,000 random keyword references. XML will clear the way for countless new activities and applications that are now too cumbersome. This is because applications require the Web to mediate between heterogeneous databases, to distribute a big portion of the processing load from the Web server to the Web client, or they require the Web client to present different views of the same data to different users. One example is the purchase of discount airline tickets. An intelligent agent will be able to search among XML price "tags" for the most favorable fares and trade in tickets when lower fares become available. XML developers will use these tags to describe Web page contents according to agreed-upon categories, making searches much more productive than the keyword criteria used today. CommerceNet, a nonprofit Palo Alto, Calif.-based organization promoting commerce on the Internet, is the driving force behind XML. According to Chairman Marty Tenenbaum, XML is a "silver bullet, with the real possibility of fundamentally restructuring the way a given industry works." XML was created last year by programmers working for the 229-member World Wide Web Consortium (W3C) as a way to simplify standard generalized markup language (SGMO), or ISO 8879. It omits the more complex and rarely used parts of SGMO, making it easier to use on a widespread basis. While the W3C is unsure exactly when XML will be finalized, it will be this year. The big boost will come when Microsoft's Internet Explorer 5.0 and Netscape's Navigator 5.0 appear. Both will support full-blown XML, while Explorer 4.0 now supports some XML applications. Microsoft is using XML for its new Channel Definition Format for Web broadcasting and Netscape is integrating it into the Gemini Web-design software suite, which is to be the centerpiece in the Apple and Sun network computer platform they hope will challenge the Window's de facto standard. TMN: Toward A Smarter Network With telecom networks becoming more intelligent, distributed, and larger every day, a new management network architecture is emerging to fill a compelling need for making carriers more competitive. The International Telecommunications Union/European Telecommunications Standards Institute (formerly CCITT) has created a network management framework and standards known officially as the Telecommunications Management Network (TMN) for operations, administration, maintenance, and provisioning. A primary objective is to support a changing environment that includes voice services giving way to data and multimedia, low-speed transport losing out to high-speed networks, closed architecture being replaced by open standards, closed access bowing to open access, and integrated systems supplanting equipment-specific configurations. TMN is an enormous and complex body of technical standards that defines the functions and capabilities of the different levels of communications management--business, service, network, and element--and how they all inter-relate. The challenge for the telecom management vendors is to assemble this complex body of standards into viable products that can be readily deployed. The bottom line on TMN is it clearly defines three essential areas of communications management: an architecture that breaks "management" down into layers and groups of functions (a triangle architecture), a methodology for defining the management behavior of managed devices (GDMO, or guidelines for definition of managed objects), and a set of protocols for management information that defines a standardized interface at all seven open system interconnection (OSI) model layers, with options for wide area and local networking. "For equipment suppliers, TMN standards are a mixed blessing," said Keith J. Willetts, the president of NFM of Morristown, N.J., a global consortium of service providers and suppliers focused on network management. He believes that while standards help increase the market for products, meeting the sophisticated TMN standards increases product costs that customers are often unwilling to absorb. "This has led to a slow acceptance of TMN standardization," he said. He believes this is changing, however, with the need to integrate diverse management systems into a cohesive whole, making it a mandatory requirement during new equipment procurement. The strongest force behind TMN may be the vendor community. NMF identified four component sets and the companies actively developing them: TMN Basic Management Platform (Bull, Digital, Sun, IBM, and HP); Multi-Network Bandwidth Management (Newbridge, NCR, Siemens, MPR); Generic Alarm Monitoring (Alcatel, Alantech, Digital); and CMIP/SNMP Interworking (Bull, Sunsoft, Netmansys). Java-Based Network Management: Using the Old Bean "Java represents a real opportunity to consolidate existing disparate network management protocols, while also broadening the view into all layers from the applications down to the network device," said Mary Slocum, a group marketing manager for Sun Microsystems, the Java developer. Surprisingly, it's almost impossible to find anyone who disagrees. The Java Management Application Programming Interface (JMAPI) was specifically designed to simplify obtaining SNMP information, particularly application management where the SNMP footprint is too large for new hand-held communications devices. The multilingual adaptability of Java matches up well with the management needs of distributed networks. Of particular importance is the ability to download executable code reliably to any device, while SNMP programming languages such as Tool Command Language have not been widely successful. The implementation of JMAPI enables a network with devices running small Java virtual machines (JVM), which in turn permits a central server to push new software simultaneously to every device running a JVM device. An unfulfilled, but quite feasible, promise is Java intelligent agents that would travel the network monitoring performance. The current emphasis is on assuring that JMAPI and other Java network management tools can coexist with SNMP as well as leverage it to provide a radically improved network management tool. One example is the management of SNMP devices from the same protocol-independent manager as Java devices. The SNMP management information base (MIB) describes the network elements being managed. By translating these MIB definitions and combining them into Java, programmers create reusable software known as Java beans. These beans perform a wide range of functions, such as spontaneous report generation. The ultimate objective is to have Java beans become intelligent agents that lead to self-managed networks. There are a number of hurdles facing JMAPI and its Java-bred tools. One is fitting into current environments and working with the existing SNMP infrastructure. A second challenge is legacy systems and equipment that could slow adoption. The time frame for overcoming this problem will be determined by how quickly network managers upgrade incompatible and less intelligent hardware. The good news is that once networks are JMAPI-enabled, there is an infinite ability to upgrade execution software on equipment. Confusion in the marketplace also holds the potential to slow adoption. Already, Java has become a language with several flavors, with Microsoft and Hewlett-Packard in particular adding proprietary features. Microsoft, along with Computer Associates International, Compaq, Cisco, Intel, HP and 3Com, is pushing Web-based Enterprise Management (WBEM) as a "non-Java" protocol that works with Web browsers. Java is a protocol-agnostic language, raising questions as to how well WBEM and JMAPI will coexist. Supporting JMAPI are Sun, Bay, Cisco, IBM, Novell, Platinum Technologies, and 3Com. Sun, as would be expected, is working with a wide variety of other vendors to overcome these potential hurdles, including sworn enemy Microsoft. Microsoft is a licensed Java vendor and has developed the JIT Compiler to recompile Java code as it is downloaded to the network device, which speeds the process of translation into native machine format. Patrick Flanagan is a contributing editor with Telecommunications. He can be reached via e-mail at pflanagan@mcimail.com. 1997's 10 Hottest Technologies 1. Web Broadcasting 2. Remote Access Servers 3. Extranets 4. Internet Telephony 5. Enterprise Network Directory Services 6. Web Site Management Tools 7. IP Switching 8. Wavelength Division Multiplexing 9. Digital Subscriber Lines 10. Higher Speed POTS Modems

To: djane who wrote (47013)	5/16/1998 5:05:00 AM
From: djane	Respond to of 61433

This Year's 10 Hottest Technologies in Telecom (Part III) [ASND references in Part II] telecoms-mag.com In the Pipeline: More Technologies Due to Have an Impact In addition to the 10 Hottest technologies, Telecommunications asked analysts and industry experts to identify other developments that bear watching. Jim Mollenauer, president, Technical Strategy Associates, Newton, Mass.: Universal ADSL. The ADSL people have finally gotten their act together and seem to have critical mass. ADSL varies from DSL in that it's a different flavor with some different twists, such as not requiring professional installation because it's splitterless; users just plug it into the existing telephone jack. A disadvantage is that some phones will cause noise in the line. To cure the problem, a user will have to purchase a filter, which is not an expensive device and easy enough to work around. When ADSL was invented, the expected application was video and this didn't work out. For Web access, ADSL is awfully fast. Packet over SONET. It simply means TCP without the ATM cells. One reason the carriers put in SONET is because it makes maintenance of the system relatively uniform. There are physical layer aspects that check on the health of the network quite irrespective of what's running over it. The carriers are going for packet over SONET, although access to dark fiber can simplify by running ATM straight over the fiber without SONET. Also, dark fiber is getting pretty scarce. Network loads due to the Internet have gone up and sucked up the fiber that's already buried, unless it was buried in the wrong place. Rick Malone, principal, Vertical Systems Group, Dedham, Mass.: Dedicated end-to-end transport over the Internet. A lot of the new technologies, such as packetized voice, packet over SONET, and ultrafast switching/routing, are being dictated by ISPs, not carriers. Carriers are more concerned about standards and an orderly rollout of a technology, whereas the ISPs are more focused on interfacing with end users and providing tremendous performance without a lot of concern about how this happens. Many IP networks today are a hodgepodge of technologies with gateways that patch them together. ISPs are pushing bandwidth providers to supply the facilities that guarantee dedicated transport, particularly within the context of quality of service standards. If Internet traffic performance is unpredictable to the end user, then the ISPs will not be successful commercially over the long run. Delivering raw bandwidth. Because of access to cable modems and DSL, the general public is starting to gain faster access to the Internet than most businesses. Businesses pay 10 times as much as the public for a 56-kbps dedicated access line, which for $40 a month gets a cable modem that can deliver 1 Mbps to 8 Mbps. Once fat access pipes such as cable modems and DSL become more common, there will suddenly be 40 million public users, as well as the now 8 million businesses, all wanting raw bandwidth with at least 1-Mbps throughput. I don't see the possibility of ever building a core network that is too big, and key components will be huge switches and lots of fiber with advanced WDM to split up the fiber. Herschel Shosteck, president, Shosteck & Associates, Wheaton, Md.: Wideband CDMA has been proposed as a new 3G (third-generation) technology by the Japanese and there's been a negative reaction by the European Telecommunications Standards Institute to protect GMS. At the same time, the regulators in Europe have allocated wideband CDMA spectrum, puttingthe carriers in a quandary. They may not believe in 3G, but they are compelled to say they favor this new wideband CDMA technology so as not to alienate the regulators or worse--lose the valuable new spectrum. The bottom line is a coming political barrage supporting wideband CDMA as a marketing solution, while in actuality many phone systems worldwide are barely into second-generation wireless systems. This situation will be played out by GSM in its next two iterations and CDMA in its next two generations both delivering on third-generation promises, while wideband CDMA struggles to find a market. Multimode, multiband wireless services. Over the next three years, there will be a shift toward multimode, multiband wireless phones and base stations. These will neutralize the religious wars over the wireless interface. Carriers in rural areas, driven by demand for roaming services, will install base stations that will serve subscribers at 800 MHz and 1900 MHz, using TDMA, CDMA, GSM, and AMPS. Over the next five to 10 years, as the price of chips drop, we'll see combined terminals and base stations. When the wireless industry started in 1983, a base station measured 8 feet by 8 feet by 20 feet and a wireless handset weighed about 35 pounds. Now, the base station is a 35-pound compact package and the handset weighs a few ounces. Michael Kennedy, director of consulting services, Strategic Network Consulting, Boston, Mass.: Enterprise-wide network management. We did a project where one enterprise network required 10 network management systems to provide a complete solution. This in itself is unmanageable. The business side is not well coordinated either. There are a lot of billing and reporting tools available that are not being used. There's great potential for full integration of all the management systems needed today for an enterprise-wide network. Web-based network management tools are part of the answer when there's a common interface to all of these tools for the network manager. Report generation needs to be made a lot more efficient as well. With outsourced Internet VPNs gaining in popularity, you've got to be able to monitor your network more fully than in the past because you've turned it over to someone else. ISP operational support systems. The challenge of ISP networks is to get the cost of administering them down. They need the equivalent of what the telephone industry calls OSS, or operational support systems. The challenge is to scale the Internet without adding a lot of people and resources. This will come through bringing to the ISP and Internet worlds the kind of technology that the telephone industry has taken for granted for a long time, particularly for facilities management and maintenance. As it stands now, the ISPs have sky-high operating costs and are losing money. Fred J. McLimans, chairman and CEO, Current Analysis, Sterling, Va.: Microwireless. Voice connectivity is moving down to the scale of a microcell in the form of home-based or small-office-based wireless. This is not quite the same as wireless local loop. It's a step smaller, usually communicating within a 50-foot radius of the microwireless cell. There's a big challenge to come up with a low-cost device that would actually allow users to go from the home-based cell to the area-based cell to the global cell with a single device and billing mechanism. Once this is done, there will be a tremendous demand for number portability, and eventually everyone will get his or her own unique telephone number at birth. This will eliminate dependence on land lines, with the exception of providing interactive bandwidth for data applications. There will be a demand to expand the cellular phone into a PDA (personal digital assistant) soon after microwireless takes hold. Speech recognition with computer telephony integration. This isn't real sexy stuff, particularly when it's used in PBXs or voice mail. It gets interesting when you take it to the level of computer telephony. Users are experimenting with speech-to-text capabilities to increase productivity and free those who write reports, memos, and presentations from their keyboards. Further ahead is the coupling of speech recognition with PDAs and Internet access capabilities, both as a navigational aid and for directory and information look-ups in combination with the enterprise data network and the PBX. We're beginning to see true integration of server-based PBXs into the enterprise network. This really starts to enable real voice and data integration and the ability to have voice recognition-based gateways for the transport for voice across the Internet or IP data service networks. How We Decide What's Hot and What's Not For all who participated in the selection process, we set down these guidelines for qualifying as a hot technology: Hot means capable of entering the mainstream of telecom systems/operations within one to two years, while now having sufficient development dollars and industry support to become economically viable. Telecom is defined broadly to encompass cable, interactive media, and other emerging forms of communication, as well as the standard inclusion of voice, data, and video transmission. A technology is not a product, but a product can execute a technology. To qualify, a new technology also has to have the potential to make a large impact on existing systems, operations, or procedures. Reader Action Poll Do you agree with our choice of the 10 hottest technologies? Are there others that should be on the list? E-mail your comments to tcs@telecoms-mag.com. We'll publish the results in a future issue. [TOP]

To: djane who wrote (47013)	5/16/1998 5:15:00 AM
From: djane	Respond to of 61433

The Brave New World [Overview of 1998 worldwide telecom deregulation] Bhawani Shankar, May 1998 telecoms-mag.com Even for the telecom industry, which is used to seeing events hyped, the past few months have been a ceaseless blur of conferences claiming to unravel the mystique of deregulation. What exactly is this phenomenon and what, indeed, was the dawn of 1998 supposed to bring? Do any of liberalization's lofty implications have any immediate bearings on the services customers get and pay for? Despite the many hours of debate at the podium, the answer has to be that there are as yet no telltale signs of this long-awaited, much-debated deregulation. Change, however, is not totally absent. It emerges in different ways, depending on the operator. Large, dominant incumbents will stress that it is business as usual, but they are nevertheless preparing new game plans. Newer and the so-called alternate operators may like to believe that the free-for-all that was promised in Europe as of January 1 this year is theirs by right, and they will be more demanding when they encounter resistance from their larger neighbors in arranging bandwidth and interconnects. Tangible changes that are more relevant to end users will take their time coming. The European Commission, Europe's legislative body, passed a whole range of directives aimed at achieving 1998 objectives. But the Commission has not had a good record in enforcing directives within a reasonable amount of time. According to the Yankee group, "Despite the best efforts of the Commission, liberalization of telecom services in Europe will continue at a very uneven pace and the gap between most and least liberal nations will continue to be very wide for the foreseeable future." In fact, for an overwhelming number of customers in Europe, January's deadline made little difference: Many markets, such as the United Kingdom, have already made significant advances toward the European ideal; at the other end of the scale are those, such as Greece, that have deferred deregulation for up to five years. Estimates of what customers can reasonably expect this year would need to be based on several key issues. The first is an assessment of what the incumbents perceive to be their realistic positions and threats, and how they may react. In addition, the plans and objectives of competitive operators will indicate what markets can anticipate. Both sets of results would also need to be considered objectively in order to make a calculated guess about the future. Incumbent, Bent Double For operators such as British Telecom (BT), still nursing its disappointment over the MCI misadventure, it would seem that the options are limited, but Chief Executive Peter Bonfield disagreed. "More opportunities do exist, but we are not going to rush out and spend all our money," he told a conference in London in late 1997. Speculation is rife about what BT will do with the $7.4 billion that it earned by bowing out of the MCI deal. New partnerships are being forecast with likely candidates ranging from Bell Atlantic (currently acknowledged as the most powerful Bell operating company), AT&T (still the largest U.S. long-distance operator and brand name with which to reckon) and Japan's NTT (full of technology and potential, but far from a surefooted, streetwise international telco). To make matters worse, the banking and analyst communities believe almost without exception that the incumbent community is in for several rude shocks over the short to medium term. "Incumbents will be at a fundamental disadvantage. They have old technology, narrow product ranges, higher unit costs, as well as regulatory restrictions," said Andrew Harrington, the London-based managing director of Salomon Brothers. "The majority will tend to be too complacent and there are sure to be few big winners and some big losers," he cautioned. In many cases, deregulation has meant price caps and these could remain in force in some regions through the rest of the century. This was geared to give competing operators sufficient financial muscle to invest in technology and establish themselves before price caps are removed entirely. "Being an incumbent is not easy," said Bonfield. "On the one hand, shareholder value and price-earnings ratios are becoming key performance indicators and on the other, the only way for us to move forward is to move into international markets while retaining the United Kingdom as a core business." The incumbents need to be considered in the historical context of deregulation. By the end of 1998, 80 percent of the world's telecom markets are scheduled to liberalize. "This is a fundamental restructuring of the world's fifth biggest industry," said Harrington, "and it has the full backing of governments around the world." The combined directives of the EU and the World Trade Organization (WTO), when fully ratified, will cover more than 90 percent of global markets. While some governments may drag their heels and attempt to extend protection of incumbents through regulation, licensing, and various other measures, they will have to ultimately open up. The political costs of resisting reform are too high and as such, reformation is an irreversible process. For instance, telecom deregulation is becoming a reality even in Japan, which has resisted it for as long as possible. Given the free market status of its dominating electronics industry, the protectionist telecom sector had been a sad contrast to its more liberal competitors. Having posted fervent opposition (on the grounds that it would dilute its ability to retain its leadership in leading-edge technologies), the incumbent, NTT, is now reconciled to the growing competition. In fact, now that the tables have turned, NTT's stated opinion is that restrictions should have been lifted sooner. "We have been slow in opening up the communications industry," said Jun-ichiro Miyazu, NTT's president, explaining the key difference between Japan's electronics and telecom industries. "Where there has been restriction, progress has been slow. Protectionism does not help commercialization, and we understand that we can no longer survive in that environment." A New Paradigm? Telecom analysts, awed by the ground-breaking changes, are at a loss for words to describe what the outcome might be. The words "paradigm change" have been offered and, whatever the possible interpretations, there is general agreement on what tangible effects may be seen at an industry level. For one, about 3000 new operators are expected to be created over the next couple of years; all of these companies will try to whittle away at incumbent market shares and create new markets. The majority of these new operators will be investing in technology that is more advanced, cost-efficient, and flexible than what incumbents own, including higher level SDH systems, IP over ATM, and so on. In comparison, incumbents will have no option but to retain PDH, copper, and 800-MHz technologies in some parts of their networks. The comparative cost-ineffectiveness of these technologies will have to be reflected in bottom-line financial performance. On the other hand, most alternative networks would be well capitalized and have powerful backers. This means that new operators could offer broader product ranges, have much lower unit costs, and be in a better position to offer bundled services. As price caps melt away, the new industry structure will, no doubt, give rise to a new economics. "In five years' time, the telecom industry will look like any another free-market industry," said Harrington. "Markets will become highly segmented, marketing costs will increase, approaching those of popular branded products in the retail sector, and the industry in general will have low predictability." Market segmentation trends are already apparent. Large incumbents are grouping together to form consortia whose sole aim seems to be to serve business customers. As experience has shown, this is where the most revenue is at stake and this is where being big and global counts. Low-revenue residential markets, on the other hand, are better served by companies similar to the United Kingdom's cable franchisees. Increasingly, residential markets will be served by cable operators or similar companies whose accents are on a mix of telephony and entertainment/interactive services. In the business services area, the emphasis on network reliability and end-to-end manageability will lead to the increased importance of owning infrastructure. Leasing and interconnect arrangements will be the choice of the typical reseller or small operator serving domestic or regional markets. This is borne out in the case of WorldCom, which, in the early days of its alliance with the long-distance operator LDDS, was more of a reseller and rose to a $1 billion revenue level without owning significant infrastructure. WorldCom's strategy took a U-turn with its acquisition of MFS, a company dedicated to high-revenue, city-centric business markets. As a result, the international operator is estimated to spend about $50 billion acquiring other telcos and building new fiber-based infrastructure. "We are in the process of building the first, global broadband network," said John Sidgmore, WorldCom's chief operating officer, adding that there is more to come. In fact, monumental changes are forecast in global backbones. The commissioning of submarine cables and terrestrial fiber backbones over the next two years will effectively quadruple the bandwidth on some international routes and increase by an even greater factor on others. Together with the fact that domestic services will benefit from the use of SDH and other cost-efficient technologies, this is expected to lead to the commoditization of voice. "We expect basic voice service tariffs to fall by about 80 percent over the next five years," said Harrington. The billions of minutes of voice traffic that will be carried on these networks by then will mean that per-minute costs of plain old telephone service (POTS) will effectively be zero. This will emphasize the value-add that service providers will have to offer in addition to basic packages. Strategic Choices These prophecies only help strengthen the view that incumbents will bear the brunt of the market rationalization that is to come. Conservatively, if the market grew by 5 percent per year and incumbents lost 4 percent of their market share and costs rose by 3 percent during the same period, the scenario for 1997 to 2002 looks frightening for the dominant operator. In this simplistic experiment, a telco's earnings could decrease by about 4 percent by the year 2002. "Bear in mind that BT is losing about 10 percent of its residential customers to cable operators in their franchise areas and this estimate starts to look very cautious indeed," said Harrington. In the business sector, too, incumbents will fare perhaps only marginally better. Operators such as WorldCom and Colt are already a formidable challenge for them. The degree of their vulnerability becomes apparent when it is considered that one of Europe's largest operators, Deutsche Telekom, relies on business customers--who represent just 4.5 percent of its customer base--for 41.5 percent of its revenues. In fact, 18 percent of its customers generate about 70 percent of total revenues. More bad news for the incumbents is in the shape of recent research done by the France-based OECD, which has found that tariff falls do lead to higher usage but not enough to compensate lost revenue. During the period 1985 to '95, when telecom tariffs fell by about 40 percent or more, telecom revenue as a percentage of GDP remained more or less constant in OECD countries. For the dominant carrier used to high margins and protected markets, revenue loss is inevitable--perhaps even just--and incumbents can, at best, only attempt to minimize losses. "An incumbent's survival strategy should include writing-off old technology and assets, increasing investment in new technology, and trying to manage market share loss by growing into new, international markets," said Harrington. Today the problem with most strategies on the drawing boards is the failure to recognize that such revenue loss and other associated changes are inevitable. The Global and the Local In the short term, Europe is expecting a flood of new operators and investment, primarily from the United States and the Asia-Pacific region, originating from a combination of private and public operators. Medium-size operators who have achieved significant success, such as Ameritech in the United States, are eyeing Europe as an obvious market. Bell operating companies have expressed their frustration with the U.S. Telecom Act's seemingly empty promise of access to long-distance markets. "Many state-owned and formerly monopolistic operators in Europe are privatizing and they will need our marketing skills," said Ameritech's chairman and CEO Richard Notebaert. "Our investment in Europe already has a book value of more than $5 billion." Over time, the distinction between public and private operators will blur and perhaps in the period after 2002, the industry is expected to rise to a new exalted status of equilibrium with four to five global supercarriers and between 2000 and 4000 domestic and regional operators. The so-called supercarriers are expected to be the world's backbone networks and the primary carriers while the smaller, second-level operators provide domestic and feeder services. Does that mean the medium-size operators that are not tied-in with one of the larger consortia are an endangered species? Ameritech's Notebaert feels otherwise: "Even though Ameritech may not be part of a global consortia, there's always room for us to participate. No matter how big an operator is, everywhere in the world is just too big." Bhawani Shankar is managing editor of Telecommunications International. [TOP]

To: djane who wrote (47013)	5/16/1998 5:21:00 AM
From: djane	Respond to of 61433

Frame Relay: Trends to Watch. High-speed frame relay provides a viable alternative to end users who are not ready to commit to ATM services Mark Kaplan [NN], May 1998 telecoms-mag.com One of the truisms of data communications is that the need for speed is ever-increasing. People want things to happen faster than they did before and that demand expands to the limit--and then beyond--the capabilities of the available mechanism. From its initial public service offering in 1991, frame relay has become the data technology of choice for organizations around the world that need to implement networks at speeds of T1/E1 and below. According to the 1997 Frame Relay Market Study by Distributed Networking Associates in Greensboro, N.C., there are now approximately 25,000 frame relay users worldwide with a total of about 270,000 ports in production. Close to 70 percent of these ports are at the rate of 64 kbps or below (DS0). Bandwidth Drivers Demand for increased bandwidth is emerging from both the residential and business communities. New-generation applications such as audio/video clips and interactive multimedia shopping and gaming sites appearing on the Internet are gaining rapid acceptance with consumers. We are in the midst of growth in both the density and volume of applications, as well as geometric growth in the sheer numbers of people accessing these sites. This will continue to grow as the new generation of high-performance, low-cost PCs are snapped up--each with a 33k or a 56k modem--and these new users get on-line. Inside the Internet, service providers are deploying high-speed backbones to reduce congestion. Increasingly, users are demanding 56-kbps, T1, and even T3 connectivity for their Web servers. To keep pace with these pressures, Internet service providers are turning to frame relay to provide high-performance, cost-effective solutions to their customers. At the enterprise, the growing demand on corporate information systems is creating the need for yet more bandwidth, especially in high-traffic areas of the network. The richness of the content within the enterprise network continues to consume all available bandwidth. As corporations become geographically dispersed, users are increasingly less content with WAN-speed bottlenecks and are looking for near-LAN speeds to all information servers within the enterprise network. Additionally, the sheer volume of data exchanged daily between major computer centers within the enterprise is driving many corporations to look at trunk speeds in excess of 2 Mbps. In recognition of this market-driven need for greater capacities and speeds, the Frame Relay Forum recently amended the User-to-Network Interface (UNI) FRF1.1 and Network-to-Network Interface (NNI) FRF 2.1 implementation agreements (IAs) to meet users' demands for access speeds up to 45 Mbps (DS3). The Frame Relay Forum is currently developing an implementation agreement for multilink frame relay (MFR) to address needs for capacities between T1/E1 and T3/E3. MFR is a software-defined means of inverse multiplexing several low-speed links to act as a single higher speed link. For example, a site requiring a 6-Mbps access link could tie together four T1 links (or three E1s) to create a logical 6-Mbps line at a cost below that of installing a T3 access line, depending upon local tariff structures. The major frame relay switch vendors currently support frame relay access and trunk speeds up to and including DS3 (45 Mbps). 1997 saw some vendors announce support for OC-3 frame relay, with higher speed support expected to come to market in the near term. Multilink support will rapidly follow the adoption of the standard. ATM and Frame Relay Most carriers are offering ATM (asynchronous transfer mode) services in parallel with their frame relay services with connection points between the two. In fact, some frame services are offered with a higher speed ATM service as the core transport mechanism between the frame relay switch points. It is important to understand the two types of interworking available. The first is network interworking (NIW), which can be thought of as encapsulation or tunneling frame relay frames through the ATM network to interconnect two (or more) frame relay attached devices. With NIW, the variable length frames are segmented and packaged into the payload of the ATM cells without disturbing the frame header information. The increase in overhead is offset by the higher switching speeds and the larger trunks interconnecting the ATM switches. If NIW can be thought of as tunneling, service interworking (SIW) can be described as a translation service between frame relay and ATM. Here, by mapping the frame header information into the ATM header, a frame relay device can establish communications with an ATM device. Together, NIW and SIW foster the coexistence of frame relay and ATM and allow users to choose the technology that best meets the traffic requirements and budget allowance of each site. ATM vs. Frame Relay Frame relay was initially designed to provide transport for the delay of insensitive data sent by higher level applications capable of recovering lost or dropped frames. Frame relay has no inherent frame correction mechanism; errored frames are simply discarded and higher layer protocols determine what frames need to be re-sent. Since the frames vary in length and traverse the buffers and matrices of the switches, delay across the network is of a non-determinative nature. While prioritization mechanisms are sometimes employed to provide differential services, to date there are no true quality of service (QoS) levels available from frame relay networks based on standard metrics. The Frame Relay Forum and ITU-T (International Telecommunications Union-Telecommunications Standardization Sector) translation are attempting to address this issue during the coming year. In contrast, ATM was built from the ground up to provide differential QoS levels (e.g., constant bit rate, variable bit rate, available bit rate, unspecified bit rate) with consistent characteristics (at least for the constant bit rate and real-time variable bit rate service levels). Utilizing a fixed-length payload of 48 bytes and a 5-byte header, ATM switches are able to provide highly determinative service classes each optimized for specific types of data and applications. The question then is which technology to use for a given network? Frame relay is historically very good at transporting data which is not highly dependent upon precise delivery intervals. Examples of this are typical client-server database queries, e-mail and file transfers, and broadcast video applications. ATM has typically been tagged as the transport of choice for delay-sensitive information such as interactive video and voice. However, recent enhancements to the basic frame relay service have tended to blur some of the distinctions between the two. FRF.11 has defined standards for carrying voice over frame relay. FRF.12 has defined fragmentation issues to create a more determinative delay pattern to the frames by chopping up the large data frames into smaller pieces to better match the size of the voice frames; thus it is able to minimize frame delay variation through the network. So, the choice of which technology to use where depends on the character of the predominant traffic. If your network will be transporting a high degree of voice traffic and/or near-broadcast quality video, ATM is most likely the better choice due to its ability to support constant bit rate (CBR) traffic as well as inherent broader bandwidth (currently up to OC-12, or 622 Mbps, with plans to extend to OC-48). If the dominant traffic type is non-delay-sensitive data (which could be voice, video, fax, imaging, multimedia, file transfer, or e-mail), frame relay is a better choice even if bandwidth in the range of 45 Mbps is required. Frame has less overhead than ATM, is more readily understood by most networking professionals, and is easier to install, and frame equipment and services are generally less expensive than ATM devices and services. Topology Drives Speed Many frame relay adopters have deployed their networks in a star topology similar to previous leased line networks. This has allowed users to realize an immediate benefit of frame relay--primarily lower cost. In this configuration, one location (typically the corporate computer site) receives traffic from numerous branches or remote locations. Since a large number of relatively lower speed locations are concentrated for delivery to this site, the bandwidth required at the central site is set at some percentage of the total bandwidth of all remote sites. Initially, customers ordered 64-kbps access (DS0) for remote office locations and 1.55 Mbps (DS1) for the headquarters site. When traffic or the number of locations increased and signs of congestion became apparent at the headquarters site, customers were relegated to ordering another DS1 facility (either as a separate DS0 or as part of an NX64 service) from their service providers, which increased access cost and required incremental hardware (ports) and customer premises equipment (CPE). Now these same customers can obtain service at speeds up to 45 Mbps without requiring incremental hardware. The individual CPE may need replacing or upgrading but additional devices are typically not required. The need to integrate or replace with ATM may not be required in this type of network based on the mix of applications and services offered. No gain would be realized by moving traditional or legacy applications to an ATM network despite the promise of greater bandwidth than available via high-speed frame relay. High-speed frame relay provides a viable alternative to end users who are not ready to commit to ATM services. Standards-based solutions exist today for access at DS0, DS1, and DS3 rates. The Frame Relay Forum is close to agreement for an IA for MFR for rates between T1 and DS3 in increments of either 1.544 Mbps (T1) or 2 Mbps (E1). With support on the horizon for SONET rates beyond 155 Mbps, frame relay insures that your investment in equipment and services can be maximized into the future. Mark Kaplan is the senior marketing manager for frame relay products at Newbridge Networks Inc. He is also the chairman of the Market Development and Education Committee of the Frame Relay Forum. He can be reached at (703) 736-5792. [TOP]