Cover Image

Not for Sale



View/Hide Left Panel
Click for next page ( 64


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 63
Generic Requirements for the Digital Data System (DDS) Network Ounce Channel Unit. BeRcore Technical Advisory TA-TSY-000083, April 1986. Available from BeUcore Documents Resists. Generic Requirements for the Subrate Multiplexer. BeDcore Technical Advisoty TA-TSY 000189, April 1986. Available from Be~core Documents Registrar. Digital Data System (DDSJ Multipoint Junction Unit (M]UJ Requirements. BeRcore Technical Advisory TA-TSY-000192, Apn! 1986. Available from Be~core Documents Register. A readily available reference on engineering of T! carrier repeatered lines is: Cravis, H., Crater, T. "Engineenng of T1 Camer System Repeatered Lines," Bell System Technical Journal Vol. XLE, March 1963, p. 431. A.~.2 Fiber Mediums and Terminals A.~.2.1 Fiber Cable Plants by Siecor Corporation-Raleigh, Norm Carolina Since the invention of practical optical fiber in the early 1970s, Be demand for optical fiber has grown tremendously. Today the uses of optical fiber are numerous. The applications of optical fiber are widespread from global networks to optical fiber extending aU the way to a subscnber's home or desktop computer. Applications for optical fiber include cable television, medicine, weapons guidance systems, aircraft control and instrumentation (fly-by-light), shipboard co~nrnunications, optical fiber gyroscopes, and remote lighting in automobiles, to name a few. Optical fiber provides many advantages for Be telecommunications system designer. It has extremely high information carrying capabilities and very low loss when compared to copper or coaxial telecommunications cables. This means Nat the distance between repeaters and regenerators can be very long (70 - 80 lan). This long distance between repeaters is best exploited when Be cables can be manufactured in lengths in excess of 10 km. Long lengths, light weight, and small diameters make optical fiber cable instaDabons much easier and less expensive Can copper. Optical fiber cables can be instaBed wad Be same ~:\NCHRPPhase:~pr NCHRP 3-51 ~ Phase 2 Final Report A1-55

OCR for page 63
equipment that is used to install copper and coaxial cables, with some modifications due to Weir small size and limited pull tensions. Optical cables can be instated in duct systems of 4000m or more depending on the condition, layout of the duct system, and installation technique. The remaining cable would be coiled and pulled further into the duct system. These long lengths of cable reduce the number of splice points making the total installation more efficient and reliable. The small size also saves on valuable conduit space. Aenal and direct buried installations also exploit the advantage of having long cable lengths with Me length limited only by the installation equipment. Another advantage is the dielectric nature of optical fiber cables. Since optical fiber is made of glass, it does not conduct electncity, nor is it affected by electromagnetic interference or radio frequency interference. This makes it the ultimate in electromagnetic transmission mediums and eliminates such issues as: unwanted, dangerous ground loops; requirement for separate electrical conduit; voltage spikes from the cycling of heavy electrical equipment; and isolation of control rooms. All dielectric optical fiber cable does not require grounding and bonding at building entrances. Since optical fiber cames a light signal instead of an electrical signal, Me cable does not emit electromagnetic signals that can affect nearby equipment and over cables. A.~.2.~.! Fiber Performance 1) Fiber Theory - Principles of Operation Every optical fiber used for telecommunications comprises three areas: core, cladding, and protective coating. The core consists of silica glass and a dopant to control the index of refraction of the glass. A typical dopant is germane. The cladding is concentric around Me core and composed of pure silica glass. The core and the cladding are manufactured as a single piece of glass and cannot be separated one from the other. The refractive index characteristics of Me core and the cladding are designed to contain the light in the core as it travels down Me fiber. Standard core diameters in use today are 8.3 ,um (single-mode), 50 ,um, and 62.5 Am (multimode). The cladding surrounding each of these cores is 125 I. Core sizes of 85 ~ "d 100 Am have been used in some applications but, they are not common today. L:~h~t NCHRP 3-51 Phase 2 Few Ream A1-56

OCR for page 63
The Bird section of an optical fiber is Me outer protective coating. The typical diameter of a coated fiber is 245 ,um. The protective coating is typically an ultra violet (UV) cured acrylate coating applied during We manufacturing process. The coating provides mechanical and environmental protection for the fiber. a. Apex of Refraction In order to understand how light travels down a fiber, one must have a basic understanding of Be principles of how light behaves when traveling Trough a medium. In a vacuum, light travels at approximately 3.0 X 108 m/s. In any over medium (air, water, glass, etc.) light travels more slowly. The ratio of the speed of lift In a vacuum to He speed of light in a medium is caned the Index of refraction. It is typically represented by Be letter "n" and its value is always greater than one (~) unless the medium is a vacuum, Den Be index of refraction would be equal to one. The slower light travels in a medium, the larger Be index of refraction. b. Total Internal Reflection When light passes from one medium to another, and the second medium has a different index of refraction, Be light is either reflected back into Be first medium at the interface or refracted (bent) as it passes Into the second medium. This pnnciple is caned SneR's Law.) The effect is similar to placing a stick in water, i.e., it appears bent. Whether the light reflects or refiacts depends on the angle of incidence and the index of refraction of the two materials. There is an angle at which the lift is refracted along Be interface of the two materials. This angle is called the cntical angle. When light approaches the intersection of two materials win different indices of refraction, at an angle less than the critical angle, the light is refracted. For angles of incidence greater than the critical angle, Be light is reflected back into the medium. The angle of reflection is equal to Be angle of incidence. This property is caned total internal reflection. This can be visualized by the following example: When standing on the shore of a smooth lake with a mountain range in the backdrop, at a certain distance Be mountain range is reflected off Be water's surface; however, Were is a point at which the mountain is no longer reflected and Be bottom of the lake is visible. This is similar to light traveling down an optical fiber. L:;wCHRP`Phasc2.rpt NCHRP 3-51 Pie 2 Fly Rein A1-57

OCR for page 63
c. Numerical Aperture To ensure that the light reflects and stays in the core, He light must enter the core through an imaginary acceptance cone. The size of the acceptance cone is a function of Be refractive index difference of the core and the cladding. The acceptance cone is not specified for a typical optical fiber, it is expressed as the numencal aperture. The numencal aperture of an optical fiber is a measure of Be maximum angle of light entering the end of Be fiber Cat watt propagate in the core of Be fiber. The numencal aperture in mathematical terms is the sine of the cntical angle. d. Modes After light is coupled into an optical fiber, it travels down the fiber in propagation states calved modes. There can be from one to hundreds of modes, depending on the type of fiber. (See single-mode and multimode fiber). Each mode Cannes a percentage of the launched light. A separate signal cannot be transmitted down He different modes. The approximate number of modes is based on the core diameter; He lower the wavelength of operation, He larger He number of modes In an optical fiber. This gives a simplified reason for the different data rate capacities of different multimode fibers. The greater He number of modes, He greater He modal dispersion among those modes and subsequently He lower the data rate capacity; however, He larger fiber cores (in multimode fibers) have a larger numerical aperture which makes it easier to couple light into those fibers. 2) System Parameters a. Wavelength Multimode fibers generally operate at two primary wavelengths: 850 nm and 1300 nm. The wavelength of operation can affect the data rate capacity so higher data rate systems generally operate at He longer wavelength. Single-mode fibers generally operate at two primary wavelengths: 1310 nm and 1550 nm. For single-mode fibers aD of He optical power is contained in only one mode. (Data rate capacities in single-mode fiber will be discussed later.) As a note, ah four of these wavelengths (850 nm, 1300 nm, 1310 nm, and 1550 nm) are invisible to He naked eye and in He near infrared region of He electromagnetic spectrum. L:`NCHRP\Phase:.rp ~NCH~P 3-51 Phase 2 final Report A1-58

OCR for page 63
b. Attenuation One of the most important optical parameters is attenuation. Attenuation is the loss of optical power as the light pulse travels down the optical fiber. The power coupled into the fiber at the transmitter loses some energy as it transmits down He fiber. There are two main types of attenuation, intrinsic and extrinsic. Intrinsic attenuation is We result of inherent effects within the optical fiber that either absorb or scatter the light pulse. Absorption occurs when nature impurities in He fiber absorb energy from the transmitted light. This effect is minimized during manufactunng of the optical fiber. Scattering of light occurs when light traveling in the fiber collides wad small imperfections in the glass, scattering light In all directions. This form of attenuation is He source of more than 90% of optical fiber attenuation. Again, this effect is minimized by the optical fiber manufacturer. There are two types of extrinsic attenuation: macrobending and m~crobending. Macrobending consists of large scale bends induced on the optical fiber. This bending effects He light guiding properties of He fiber, and a portion of He light traveling in the fiber's core is lost into He cladding. This loss is due to localized change in the critical angle. The percentage of loss depends on the severity of the bend. The energy loss associated with macrobends is reversible with removal of He bend. Microbends are small scale distortions of the fiber Hat alter He index of refraction at He point of stress. This altered index changes the light guiding properties of He fiber at Hat point, and causes light to escape into He cladding due to He change in index of refraction. These bends are typically caused by temperature, or by tensile- or crush-related effects on the optical fiber cable or the unprotected optical fiber. Optical fiber attenuation is typically stated in terms of decibels (dB) and normalized to decibels per kilometer (dBA`m). The end-to-end attenuation of a fiber link will typically be stated in dB, while the cabled fiber attenuation will be specified in dB/km. Specifying attenuation is usually done in He windows of operation for He optical fiber. A multimode fiber will have He attenuation specified at 850 rim and 1300 nm, while a single-mode fiber will have attenuation specified at 1310 nm and 1550 nm. Typical values for multimode optical fibers are 3.5 dB/km .~.= NCH"3-51 P~2F - ~n A1-59

OCR for page 63
at 850 nm and I.5 dB/km at 1300 nm. For single-mode optical fiber, typical values are 0.4 dBA`m at 1310 nm and 0.3 dB/km at 1550 nm. An optical fiber system can be limited by the attenuation and/or the dispersion. The attenuation limit comes from loss in optical power. If the system is too long, Me signal wait not have enough energy to be detectable when it reaches We receiver. It should be noted that this can be a function of the end equipment as well as the optical fiber. The transmitter may not have enough power, the receiver threshold may be too high, or the optical fiber link may have too much attenuation. One must consider splice and connector loss as weB as the fiber loss with distance. (Dispersion limits are discussed in Me next section.) c. Dispersion Data rates are limited by dispersion which causes the input pulse to spread in time as it travels through a fiber. This spreading increases the potential for interference between sequential pulses. Pulses launched closely together High data rates) that spread too much High dispersion) result in bit errors at the receiver. For example: if He input pulse is 5 ns long, the output pulse may be 6 ns long after 5 km. There are three types of dispersion; modal and chromatic are He two most common and have the most impact on typical telecommunications systems. The Bird Thorpe is polarization mode dispersion. Modal dispersion occurs only In multimode fiber while chromatic dispersion occurs in bow single-mode and multimode fiber. Modal Dispersion: Each mode of light in a multimode fiber follows a different path through the core of the fiber. These individual paths have different associated lengths relative to the fiber length. The modes following the shortest paths will arrive at He end of the fiber before the modes following the longer paths. A pulse of light consisting of hundreds of modes in a multimode fiber will, therefore, broaden In time as it travels through He fiber. Chromatic Dispersion: Chromatic dispersion is determined by He fiber's composition, structure, and design, and by He operating wavelength and spectral width of the light source. Chromatic dispersion is measured in units of (ps/nm x km): picoseconds of light pulse spread L:~h~.~t N~P3-51 e P~2F~Re"n A1-60

OCR for page 63
per nanometer of spectral width and per kilometer of fiber length. Chromatic dispersion occurs because different wavelengths of light experience a change In index of refraction and, therefore, transmit at different speeds. There are two components of chromatic dispersion: material and wavegu~de. The amount of material dispersion vanes web the composition of He glass and is a function of the wavelength of the light source. No transmitter produces a pure light source of only one wavelength. Instead, sources produce a range of wavelengths, resulting In pulse spreading that increases with distance. Wavegaide dispersion results from a portion of the light traveling In Me cladding as weR as He core of the fiber. Since He mode field diameter is a function of wavelength and each pulse consists of more than one wavelength, the pulse spreads out. This also increases with distance. Both material and wavegu~de dispersion can be decreased by using a better light source (smalder spectral we. Modal dispersion is the dominant factor in multimode fiber but does not occur in single-mode fiber because there is only one mode. Chromatic dispersion is present in both fiber types. A maximum dispersion value is specified for single-mode fiber in most industry standards to aid in calculating He data rate capacity for the fiber. This must be used in conjunction with transmitter and receiver specifications. Industry standards for multimode fiber, however, specify bandwidth instead of dispersion. Polarization mode dispersion: (PMD) is an optical effect which can spread or disperse an optical signal, occulting in commonly used single-mode fibers. It originates web the source, typically a narrow-linewidth laser diode. Light which is restricted to a narrow spectral width is polarized when it is emitted. The light is then coupled into He single-mode fiber where it is resolved into two orthogonaNy-polanzed components which make up the fundamental mode. PI occurs when He two components which comprise the fundamental mode beg n to propagate at different speeds. The result is a distortion of the originally transmitted signal. This difference in propagation velocity represents a difference in the index of refraction for the two components, referred to as He fiber's birefringence. Since both of the polarized components carry a portion of the transmitted power, any offset in the arnval time of He components acts to distort the lingual signal. In digital commun~cabon systems, the result is dual imaging of the signal at the receiver. This dual imaging is destructive to the signal integrity, liming operational transmission capacity. Additionally, it is important to note that PMr) increases at the square root of the length, while chromatic dispersion increases linearly win length. Therefore, c;WCHRmPhasc:.rpr NCHRP 3-51 Phase 2 final Report A1-61

OCR for page 63
Be impact of PI diminishes wig longer length systems which is not the case with chromatic dispersion. PMO may be an issue with long haul applications. Test methods and specifications are still under development. d. Bandwidth Bandwidth is a measure of Be data rate capacity for multimode fiber. It is typically measured by Be fiber manufacturer on the bare fiber and repotted in units of MHz-km. Due to the fact that bandwidth is not linear, special care must be taken when estimating the data rate capacity for an installed system since a number of fibers may be concatenated together to comprise the system. Bandwidths of fibers Cat are concatenated together forming a long span cannot be simply added together then divided by the length to get the band of Be system. The formulas for calculating system bandwidth are beyond the scope of this article but can be obtained from the fiber or cable manufacturer. Suffice it to say, Be end equipment manufacturer should state the fiber bandwidth necessary for proper operation of a system over a given length. 3) Fiber Types There are two major classifications of fiber. In general, multimode fiber is Lest suited for local network use, where links are short (less Man 2000 m) and there are many connectors. The higher numencal aperture of multimode fiber allows the use of relatively inexpensive Light Emitting Diode (LED) transmitters and lower cost connectors. Single-mode fiber is best suited for long distance applications or applications requiting extremely high data rates. Given these two optical fiber types and typical trar~sm~tter and receiver specifications, which fiber type is best suited for Be applications listed above for a given distance? The following tables summarize all of these parameters. utNC~Phasc2~pt NClIRP3-51a Phase2FmalReport A1-62

OCR for page 63
Table A.~.2.~.~-1 Applications Summary for 62.5/125 Am Multimode Optical Fiber Application/ Maximum Data Rate (MBiVs} Bandwiclth Attenuation Wavelength (nary) Specification Distance (m) (MHz-km) (dBllcm) 568-A 2000/90 _ ~ 60/500 3.75/1 .5 e50/'300 Ethernet 2000 10 160 3.75 850 10Base F Ethernet 2000 100 500 2 5 Note 2 1300 100Base F Ethernet ~Note 3 ~ GO ~Note 3 ~50/' 300 VGAnvLAN . _ _ ._ Token Ring 2000 4 160 3.75 850 Token Ring 2000 16 160 3 75 850 FDDI-PMD 2000 100 500 2 5 Note 2 1300 . . . FDDI-LCF 500 100 500 N= ~1300 FDDi-850 1 Note 3 1 00 1 Note 3 1 Note 3 1 8~50 __ Fiber Channel 1500 133 160/500 1.0 1300 _ Fiber Channel 1500 266 160/500 1.0 1300 ATM l Nots 3 ~52 ~Note 3 ~Note 3 ~ 300 ATM 2000 155 500 1 .5 1300 ATM | 300 622 | Note 3 Note 3 | Note 3 ] Siecor 1 _ 1 200/500 1 3.5/1 .0 1 850/1300 Note 1: Backbone~onzontal distances Note 2: Typical (end-to-end attenuation is specified at 11 dB) Note 3: Standard is son under development Note 4: End-to-end attenuation is specified at 7 dB; typical normalized attenuation is not specified. Table A.~.2.~.~-2 Applications Summary for Single-Mode Optical Fiber Application/ Maximum Data Rate Attenuation Wavelength (nary) Specification Distance (m) (MBiVs) {dB/km} 568-A 60,000 ~1. 0/1.0 13 10/1530 FDDI-SMF ~60,000 ~100 ~0 5 Note 3 ~t 300 Fiber Channel 10,000 531 0.5 1300 . Fiber Channel 10,000 1063 0.5 1300 Siecor _ 0.5/0.4 1 310/1 550 Note 1: Beyond Me scope of ~A/ElA-568-A. L:~h~.~t NCH~3-51 a P0e2F~Re"n A1-63

OCR for page 63
Note 2: Maximum attenuation for inside optical fiber cable; maximum attenuation of outside optical fiber cable is 0.5/0.5 dBAcn at 1310/1550 nm. Note 3: Typical (end-to-end attenuation is specified up to 32 dB) a. Multimode Optical Fiber Wavelength of operation is another important parameter that must be considered in Me design of this system. Multimode fiber usually operates in two different windows: 850 nm and 1300 nm. The Milt signal will behave differently at each wavelength. The attenuation at 850 nm will tend to be higher than at 1300 nm. In addition, the bandwidth at 850 nm tends to be lower than at 1300 nm. Quite simply, multimode fiber performance is better at 1300 rim than at 850 nm, making longer system lengths possible at 1300 nm; however, the cost of the end equipment (primarily LEDs) may be less at 850 nm, lowering Me overall system cost. b. S=gle-mode Optical Fiber Standard single-mode fiber is designed to operate at 1310 rim and 1550 nm; however, Mere are tradeoffs at bow wavelengths. The attenuation at 1550 rim is generally lower Man at 1310 nm. The dispersion, however, is much higher at 1550 nm than at 1310 nm, since the zero dispersion wavelength is closer to 1310 nm. If the system designer desires bow low dispersion and low attenuation at 1550 nm, dispersion- shifted (DS) single-mode fiber is available. DS s~ngle-mode fiber is generally used for long distance and high data rate applications where the lowest possible loss is desired without A - sacr~nc~ng low dispersion. Regardless of Me fiber type chosen, the designer should specify Me attenuation and bandwidth (dispersion for single-mode fiber) at bow performance windows: 850 rim and 1300 nm for multimode fiber, and 1310 nm and 1550 nm for single-mode fiber. Specifying these values gives Me design more flexibility to accommodate future upgrades. L;WCHRP\Phase2.rpt NCHRP 3-51 Phase 2 Final Report A1-64

OCR for page 63
As data rates increase and the system designer considers new applications, certain key end equipment (transmitter and receiver) parameters can be changed to improve system performance on the instaHed optical fiber. While optical fiber has the necessary bandwidth to handle applications wed into the future, the end equipment may have to be replaced to achieve higher system performance. As discussed above, changing the center wavelength may improve attenuation and/or bandwidth. Another key factor is spectral width. Tightening He spectral width will reduce chromatic dispersion and thus allow for a higher banded. Rise and fall tunes of the transmitter and receiver can also be lowered to achieve a higher bandwidth. Cable manufacturers and/or end equipment manufacturers can assist in evaluating a system with different light source characteristics (center wavelength and spectral width). 4) Comparison Between Fiber, Copper, ant Coax A number of considerations justify the transition from copper to fiber at such an accelerated pace. Fiber-based communication networks provide gigahertz bandwidth as compared win megahertz bandwidth available win copper twisted pair communications. Single-mode fiber supports link distances up to 80 kilometers without degrading its bandwidth capacity, while maintaining a lower noise floor or larger signal-to-noise ratio Han copper cable. In addidon, advanced multimedia commun~cadons services such as real-time interactive video-require the significant banded Hat optical fiber can deliver. Optical fiber systems support much greater distances between electronics without the necessity of line amplifiers and repeaters, simplifying He network, due to lower attenuation. Compared with copper twisted pair, optical communication systems exhibit a much lower bit error rate (BER) while operating at much higher data or bit rates. As a result, communication of Intelligent Transportation Systems FITS) data is both faster and more reliable over optical fiber systems. At high data rates (typically in the Gbps range) copper twisted pair communications can exhibit a bit error rate of lo-6. At 2.5 Gbps fiber optic networks typically exhibit a bit error rate of lO~9 and are tested at 10~~. Networks carrying critical data, particularly Hose involving safety and finance, demand He reliability that is possible with optical fiber. c:`NCHR~Phasc2.'p' NCHRP 3-51 Phase 2 Final Report A1-65

OCR for page 63
AddidonaBy, SONET offers advantages over Me T! digital hierarchy Eat include: SONET is synchronous and can support add/drop capabilities of lower rate signals to the DS-O 64 kbps level. Thus, expensive multiplexing equipment is not required to insert/extract lower rate signals at intermediate network locations. Figures A.~.2.3-la and b iBus~ate the concept. Thus, networks can be more cost-effectively implemented (40 50%). The SONET standard provides nearly 5% of the bit rate for support of advanced network management, operations, administration, maintenance, and provising functions. This is in addition to We fundamental payload. These capabilities are not consistently defined (mostly not available) in He T! hierarchy. SONET was conceived as a fiber standard, OC-ns. Although STS-n defines electrical standards, electrical physical level transmission is not deployed. In contrast, the DS-n, T1 digital hierarchy evolved to support wire and radio and does not define optical physical transmission standards. SONET is an international standard [called Synchronous Digital Hierarchy (SDH) internationally] . N SONET can transport the legacy' widely deployed, DS-n, T1 hierarchy signals but has the flexibility to transport emerging signal types such as LAN/WAN data, videos etc. Ability to implement/integrate fault tolerant architectures/topologies in an integrated network design. The standards and equipment have integrated support for this capability. The elements of a SONET network are similar to elements of a T} network (Section A.~.2.3) and generally consist of: Network controllers (e.g., network management); Digital cross-connect systems; Digital multiplexing equipment/terminals; and c:WCHRP\Phase:.rpr NCHRP 3-51 Phase 2 Final Report A1-1 13

OCR for page 63
Q o 1 In C] _ X ~ 1- 0 1 ~O 1 en 0 N Q _ 1 ~1 to Lo Lo Z X Z L,~J o ol 1~ (O ~ O _ ~ a) 1 (a A .< to O _ ~ ~ _ Z o \ ~ ~ O O ~ D oh ~ ~ ~ I ~ ill ~ i~ at_ cn ~ it- c) o 1 ~ ~ O - > (:7 ~ I Q' c' ~ .- ~Mu ~1 ~ O ,O,u, I

OCR for page 63
Transport systems. The significant new capability is Me Add-drop Multiplexer (ADM) Mat permits cost-effective adding/dropping of lower level DS-N/SONET signals to DS-0, 64 kbps signals. The SONET signal format consists of payload (application data) and multiple levels of overhead for flexible implementation of network management, provising, monitonng, and control. These overhead functions, if implemented, were accomplished USA non-standard methods in T1 networks. SONET standards define these capabilities, or win define Hem (not aD standards are complete). The SONET OC-! signal format consists of 49.536 Mbps payload and 2.304 Mbps Overhead as depicted in Figure A.~.2.3-2. Based on years of %1 experience, Be SONET standard developers placed a great deal of emphasis on operations, administration, maintenance, and provising (OAM&P) requirements Mat win become more essential in multivendor system implementations. Thus, the overhead assignments presented in Figure A.~.2.3-2 are generally assigned for implementation as presented in Figure A.~.2.3-3 A path represents repeaterIess links; a line represents node to node links; and a path represents end-to-end application/user links. Spare communications capacity available for these functions is allocated independent of Me payload. The fundamental OC-! payload was conceived for transporting a DS-3, 44.736 Mbps, signal consisting of 28 DS-Is or 672 DS-0s (see Section A.~.2.3~. The 49.54 Mbps OC-1 payload includes extra byte allocations for synchronizing aBocations and paw overhead avocations. The SONET data actually framed is a 125 M sec (U sec = 10~9) frame which is Me standard T} 8000 frames/second rate. Pointers are employed in the line overhead bytes to allow floating payload frames. To avoid traditional T! buffering requirements and resulting slippage. The above SONET features allow simple synchronous bit-interleaving multiplexing techniques Mat permit cost-effective equipment development and simple interconnect of network elements of varying functionality. The SONET standards define: . Optical interface Specifications :`NCHRP`Phase2.rp: NCHRP 3-51 Phase 2 Fmal Report Al-1 15

OCR for page 63
l con - in Q a) sdq~ 9~ S CV3HS3AO HlVd cn ~ Q O LLI D I Y J > O cn Q ~ D LLJ I > o Cr) o CY an' - cn o in cn Q lO D N c 00 - o ~ ~ a, o 1 ~ ~ cn D o - 11 o C] it C] to LLI Cam CD o it to An At: (d - - c~ Hi . a: ~J

OCR for page 63
LLJ of I1J J ~ ct Q X LO x J J Q Y ~ ~_ LLJ 2 . J CL ~ -2 X ELI LO Z L1J to ~ Q o i) _ Z ~ _ Z ~ LL Z O_ - . ZZ LO Z Z ~ Jim C, 1.~_1 .: Cr L~J ~ / ~- i) ~ C) / \ Z Z Z 'C ~ Jew, m L. T o ZZ _ I ~I: ~ ~ 7 \ _ _ o llJ .: o L-I ~n z o LLJ ~n L . ~ I CL 1 C~ ~n ~: z y ~: o z z o o ~n z z - ~n 1 C~ ec :e o z o ~n 'td llJ

OCR for page 63
Rate and Fonnat Specifications Operational Specifications SONET Standards Relevant U.S. standards dealing with SONET include: ANS! T].101 Synchronization Interface Standards for Digital Networks. ANS T].106, Digital Hierarchy-Optical Interface Specifications (Single-modeJ. ANS T].102, Digital Hierarchy Electrical Interfaces. SONETA~-drop Multiplex Equipment (SONETADMJ Generic Criteria TR-TSY-000496, Issue 2 (Be~core, September 19891. Integrated Digital Loop Carrier System Generic Requirements, Objectives, and Interface. TR-TSY-000303, Issue ~ (Bet/core, September 1986) plus Revisions and Supplements. Digital Synchronization Network Plan. TA-NPL-000436, Issue l (Bet/core, November 19861. Synchronous Optical Network (SONETJ Transport Systems: Common Generic Criteria. TR TSY-000253, Issue ~ (Bellcore, September 1989~. (A module of TSGR, FR-NWT 000440,. Transport Systems Generic Requirements (TSGRJ: Common Requirements. TR-TSY-000499, Issue 3 (BeNcore, December 1989~. (A module of TSGR, FR-NW1-000440~. Synchronous Optical Network (SONET) Transport Systems: Common Generic Criteria TA NWT-000253, Issue 6 (BelIcore, September 1990), plus Bullets No. I, August 1991. Generic Reliability Assurance Requirements for Fiber Optic Transport Systems. TA-NWT 00418, Issue (Be~core, to be issued). Relevant, ITU-T international recommendations swim essentially the effect of a standard) on SONET/SDH include: G.702 Digital hierarchy bit rates G.703 Physical/elec~ical characteristics of hierarchical digital interfaces G.707 Synchronous digital hierarchy bit rates G.708 Network node interface for the synchronous digital hierarchy G.709 Synchronous multiplexing structure G.773 Protocol suites for Q interfaces for management of transmission systems c:\NCHRP\Phase2~p ~NCHRP3-51 Phase2FmalReport A1-118

OCR for page 63
G.781 Structure of recommendations on multiplexing equipment for the synchronous digital hierarchy G.782 Types and general characteristics of synchronous digital hierarchy multiplexing equipment G.783 Charactenstics of synchronous digital hierarchy multiplexing equipment G.784 Synchronous digital hierarchy management G.955 Digital line systems based on Me I.544-Mbps hierarchy on optical-fiber cables G.956 Digital line systems based on Me 2.048-Mbps hierarchy on optical-fiber cables G.957 Optical interfaces for equnpment and systems relating to Me synchronous digital hierarchy G.958 Digital line systems based on Me synchronous digital hierarchy for use on op~cal-fiber cables G.652 Character~shcs of a single-mode optical-fiber cable G.653 Charactenshcs of a dispersion-shifted, s~ngle-mode, optical-fiber cable G.654 Charactenstics of a I,500-nm wavelength, loss-m~zed s~ngle-mode, optical-fiber cable M.30 Telecommunications management network A.~.2.4 Asynchronous Transfer Mode (ATM) As discussed in Section A.2.4 voice and video require constant delay traditionally served by c~rcuit-svv~tched communication technology. conversely, random arrival, bursty, data has been traditionally served by packet data communication technology. In reality, video has been served by analog technology and has not histonca}Ry represented a digital network load. The emerging "video-on~emand" and very popular Internet are creating demand for broadband services. LANs can provide local broadband data services; however, w~de-area broadband services have not been cost-effectively available except to large bulb users. Fur~elmore, LANs have not served video and voice well, due to random delays and out-of-order packet amvals.. ATM is Me industry solution to provide Integrated multimedia communication services on a common public network infrastructure. ATM is the switching technology and SONET is the generally preferred transmission technology. The concept Is depicted in Figure A.~.2.4-~. ~:\NCHRP`Phasc2.rpr NCHRP3-51 Phase2FmalReport Al-1 19

OCR for page 63
/ / 1 o D Q en ~ ._ . _ ~ 1 to _ 1 - 1 C) o a) - _ _ a) Z ~ a) In _ D Q FEZ _ CL In C . _ ._ > C ~ ..~ ~ _ D :] Q a) \ \ \ ~\ (L) en O A o en o In In _ U) C:, /1 ' en _ Us ~ .c ,- ~ ~ ~ DO ~ ~ ~ ~ ~ In _ Q u) cn ~ - ~. - > -. - ~ -= ~ D ~ ~ O _ cn ~ 1 , 1 Z ~ ~ \ _ D O ~ ~ +, O ~ - ~ C) C~ _ V =' a) ~= =, \ m/ m C. 0 1 N ~n ~ _ _ ~: ~ c I ,~ ~ \ Z I..J _ ~ ~ 3 0 z O z o cn z ~n ~: o Z - - m- r

OCR for page 63
Table A.~.2.4-! presents an overview of the role SONET and ATM in emerging multimedia networks. Table A.~.2.4~1 Role of SONET and ATM in Multimedia Network SONET ATM Transmission/Multiplexing Services Services Operations, Administration, and User network interfaces Maintenance (OAM) functions Switching operations between SONET links In many private networks, switching is not required. Thus, ATM may not be needed in many ITS-related communication infrastructures, particularly ATMS systems that have limited switching requirements. ATM technology is based on fast packet technology Mat is intended for operation over low BER circuits (i.e., SONET). It employs a 53-byte cell structure consisting of 48 bytes of user data and 5 bytes of protocol overhead. In terms of the OSI protocol model, ATM defines physical (~) layer and link (2) layer protocols, although the physical layer definition generally references over standards such as SONET. The short fixed (53) packet cell structure and low overhead is a well conceived compromise to accommodate isochronous voice and video traffic as wed as packet data ATM is a connection-oriented technology Hat pemlits every ATM cell to travel over the same route that is defined during cad setup on switched services or by provisioning in private line services. ATM supports virtual networking in that Be route is fixed during a connection, but may be changed based on time-of-day, failure, reconfiguration, or over requirements. Virtual networks should allow public services providers to offer lower cost, more reliable, alternatives to dedicated leased lines as redundant facilities can be shared/reaBocated among many users. The ATM connection exists as sets of routing tables in each switch, based on the address contained in Be cell headers. ATM we define Public and Private User Network Interfaces ~s). ATM and SONET are generally Be technologies of the Broadband ISDN (BISDN) services Cat are being defined by :\NCHRP\Phasez.rp ~NCHRP3-51e Phase2F~nalReport A1-121

OCR for page 63
ITU-T for fixture public multimedia telecommunication services. ATM defines bow ~ infrastructure and end user interfaces and services. ATM standards (e.g., ATM 25) are defined for desktop interfaces that could eventually replace LANs in corporate networks and have weD- conceived MAN/WAN strategies. These wiN undoubtedly have ITS-related applications as Hey become established in the marketplace. The fundamental goal of ATM is to provide common communication infrastructure and interfaces for multimedia networks of video, graphics, voice, and data. BISDN, ATM, and SONET wig integrate the many disjointed LAN/MAN/WAN and T! services now available. ATM standards are still In development, but He key ATM current standards are presented in Table A.~.2.4-2. L.:`NC~Phase2.rp ~NCHRP 3-51 Phase 2 Fmal Report A1-122

OCR for page 63
Table 1.2.4~2 ATM Standards and Specifications [1] | ATM Forum, "ATM User-Network Interface Specification, Version 3.0, 1993, Prentice Hall [2] ATM Forum, ``Network Compatible ATM for Local Network Applications," Phase 1, Version 1.0, 1992 [3] ATM Forum, ATM UNI Specification. V.22, July 1993 [4] Bellcore, SR-NWT-002076 Report on the Broadband ISDN Protocols for Providing SMDS and Exchange Access SMDS," Issue 1, September 1991 _ [5] ITU-T Recommendation 1.1 13, "Vocabulary Terms for Broadband Aspects of ISDN,n 1992 (Rev.) _ . [6] ITU-T Recommendation 1.121, Broadband Aspects of ISDN," 1990 [7] ITU-T Recommendation 1.150, UBISDb1 ATM Functional Characteristics," 1992 (Rev.) [8] ITU-T Recommendation 1.21 1, BISON Service Aspects," 1992 (Rev.) [9] ITU-T Recommendation 1.31 1, UBISDN General Network Aspects," 1992 (Rev.) [10] ITU-T Recommendation 1.321, "BISDN Protocol Reference Model and its Application," 1990 [1 1] ITU-T Recommendation 1.327' aBISDN Functional Architecture Aspects," 1992 [12] ITU-T Recommendation 1.35B, "BISDN ATM Cell Transfer Performance," 1992 (Draft) [13] ITU-T Recommendation 1.361, "BISDN ATM Layer Specification, 1992 [14] ITU-T Recommendation 1.362, BISON ATM Adaptation Layer (ML) Functional Description," 1992 .. [15] ITU-T Recommendation 1.363, MISDO ATM Adaptation Layer (ML) Specification," 1992 L;~.NCHRP\P~.rpr NCHRP3-51a Phase2FmalReport A1-123