Physical Layer Tests Of 100 Gb/s Communications Systems

2y ago
106 Views
2 Downloads
3.88 MB
30 Pages
Last View : 1m ago
Last Download : 2m ago
Upload by : Ellie Forte
Transcription

Physical Layer Tests of 100 Gb/sCommunications SystemsApplication Note

Application NoteTable of Contents1. Introduction. 34. Diagnostic Tests . 282. 100G and Related Standards. 44.1. What to do if the transmitter fails.284.2. What to do if the receiver fails.292.1. 100 GbE – IEEE Standards 802.3ba, 802.3bj,and 802.3bm.62.2. OIF-CEI.62.3. Fibre Channel 32GFC.63. Testing 100G Systems. 73.1. Testing optical transmitters.93.2. Testing optical receivers.113.3. Testing electrical transmitters.163.4. Testing PAM4 electrical transmitters.193.5. Testing electrical receivers.223.6. Testing PAM4 electrical receivers.262www.tektronix.com/opticalSummary. 29

Physical Layer Tests of 100 Gb/s Communications Systems1. IntroductionThe list of bandwidth intensive applications has exploded inthe last decade. Video-on-demand, voice-over-IP, cloudbased computing, and storage have created a ravenousbandwidth appetite that has rushed deployment of 100Gb/s technology and spurs development of 400G.The power of HSS (high speed serial) technology, withits noise-resistant differential signaling and jitter-resistantembedded clocking plus closed-eye equalization,enables 25 Gb/s on previously inconceivable lengthsof PCB (printed circuit board). Combining HSS links inparallel simplifies electrical 100G signal transmissionto optical transceivers, easing connectivity to the fiberoptic backbone. The result is that datacom and telecomtechnologies use 100 Gigabit Ethernet (100 GbE) fortransport, including SAS, Infiniband, even Fibre Channel.This note covers the transmitter and receiver testsnecessary to assemble 25-32 Gb/s single lane systemsand complete 100G systems. Since every 25 Gb/s HSStechnology shares common themes, we’ll follow 100 GbE,that is, the IEEE 802.3ba, 802.3bm, and 802.3bj compliancespecifications but point out differences between otherhigh rate systems like Fibre Channel’s 32GFC, and theOptical Internetworking Forum’s CEI (common electricalinterface) implementation agreements, called IAs rather thanstandards.As we work through the tests, we’ll encounter commonthemes in the interplay of jitter, noise, frequency response,and crosstalk. After the description of compliance tests,we’ll suggest strategies for diagnosing noncompliantcomponents and systems as well as techniques formeasuring performance margins.Technology specifications tend to be written in theengineering equivalent of legalese, so we composed thisnote as a supplement to clarify the tests themselves, theirrole, and how to perform them.www.tektronix.com/optical3

Application NoteStandard100 GbEOIF-CEIFibre ChannelGeometryReach or Max LossData Rate100GBASE-ER4100GBASE-LR44 SM fibers2 m to 30 km2 m to 10 km4 25.78125 Gb/s100GBASE-SR44 MM fibers 100 m100GBASE-CR44 sets of matched cables 5m100GBASE-KR4PCB 35 dB at 12.9 GHz100GBASE-KP4PCB 33 dB at 6.8 GHzPAM4 4 13.6 GBdCEI-25G-LRPCB0-68.6 cm 1 or 2 connectors19.90-25.80 Gb/sCEI-28G-MRPCB0-50 cm 1 connector19.90-28.10 Gb/sCEI-28G-SRPCB0-30 cm19.90-28.05 Gb/sCEI-28G-VSRPCB10 cm on host PCB 1 connector 5 cm on module PCB19.60-28.10 Gb/s32GFCN channels optical andelectricalMany optical and electricalvariations28.05 Gb/sTable 1. Summary of 100G and related standards.2. 100G and Related StandardsStandards recommend tests to assure componentinteroperability. In this section, we summarize thespecifications, Table 1. It’s important to keep in mind thattechnology standards documents constantly evolve. Thenumbers we quote should be considered typical of what toexpect, but for compliance testing, always check the actualstandards!4www.tektronix.com/opticalThe electrical technology shares the followingcharacteristics: balanced, unidirectional, 100 Ohmdifferential signaling with embedded clocks, low voltageswings, NRZ (non-return to zero) signals, and multiplechannels. As data rates increase and engineeringexperience grows, multi-level signaling, particularly PAM4(4-level pulse amplitude modulation), will replace NRZ incertain applications. By encoding two bits in each symbol,PAM4 reduces the bandwidth demand at a given data rateby about half, at the expense of signaling complexities.

Physical Layer Tests of 100 Gb/s Communications SystemsFigure 1. Reed Solomon forward error correction scheme. RS-FEC(n, k) is capableof correcting up to t errors in each word.The capacity for FEC to correct errors relaxes the BERcompliance requirement at the expense of latency. Toaccount for FEC limitations, both the BER and FLR (frameloss rate) are specified. Where BER is defined as the ratioof the number of bits in error to the total number of bitstransmitted, the FLR is defined as the ratio of the numberof invalid, lost frames, to the total number of framestransmitted.To meet the operational net BER 10-12 requirement,100 GbE specifies both:FEC (forward error correction) is required in some of the100G specifications and optional in others. Reed-SolomonFEC is like a cyclic redundancy check but with the ability tocorrect as well as count errors. The RS-FEC(n, k) schemeuses an algebraic technique to encode k symbols of datawith 2t parity symbols into an n symbol word in such a waythat t symbol errors can be corrected, Figure 1.For example, all of the electrical (100GBASE-CR4 andKR4) and one of the optical (100GBASE-SR4) 100 GbEspecifications covered in this document use RS-FEC(528,514). The PAM4 100GBASE-KP4 specification uses RSFEC(544, 514).Here’s how it works: each symbol consists of ten bits. Eachcodeword consists of 528 symbols, totaling 5280 bits. RSFEC(528, 514) can correct at most 7 symbols, a maximumof 70 bits out of each 5280, the maximum depends on theordering of the errors. The ability of RS-FEC(528,514) toidentify errors isn’t so restricted. Similarly, 100GBASE-KP4’sRS-FEC(544, 514) can correct at most 150 errors.BER 5 10-5 or 10-6 prior to or independent of FEC andFLR 6.2 10-10 for 64 octet (i.e., 512 bit) frames,including FEC.these requirements all but assure operation at BER 10-15even in the presence of burst errors.The two electrical applications for which FEC is notrequired, 100GBASE-ER4 and LR4, require BER 10-12.Since terminology varies among the specifications, let’sclear up potential misunderstandings immediately. In thisdocument, we distinguish the data rate and payload rate;the data rate is the rate at which raw data propagates.The payload rate does not include overhead from FEC andcoding, hence, payload rate data rate. In discussions ofNRZ signaling (more formally called PAM2 for 2-level pulseamplitude modulation), we use Gb/s and for PAM4, GBd.www.tektronix.com/optical5

Application Note2.1. 100 GbE – IEEE Standards 802.3ba,802.3bj, and 802.3bmThree IEEE standards cover the 100 GbE systems listed inTable 1. All of them achieve the nominal 100 Gb/s data rateby combining four separate 25.78125 Gb/s lanes.The IEEE Std 802.3ba standard covers the long reach,100GBASE-LR4 and extended reach, 100GBASE-ER4, fourlane single mode fiber specifications over 10 and 40 km,respectively. The differences between them are primarilyat the receive end. The ER4 receiver has greater sensitivityand has to pass a more difficult stress tolerance test thanthe LR4 receiver.The IEEE Std 802.3bm covers the short reach 100GBASESR4 MM (multimode) fiber specification which has a rangeof at least 100 m. Where RS-FEC is optional for 100GBASELR4 and 100GBASE-ER4, it’s mandatory 100GBASE-SR4.The IEEE std 802.3bj covers the electrical specifications.100GBASE-CR4 consists of four lanes of shielded balancedcables with reach up to 5 m and 100GBASE-KR4 and100GBASE-KP4 consist of four differential traces onbackplanes. Rather than specify a minimum reach, KR4and KP4 specify the maximum allowed insertion loss at theNyquist frequency which provides greater system designflexibility. RS-FEC is mandatory for 100GBASE-CR4,100GBASE-KR4, and 100GBASE-KP4.2.2. OIF-CEIImplementation Agreements (IAs) from OIF-CEI do notprescribe compliance tests the way that IEEE’s 802.3100 GbE or Fibre Channel specifications do. Instead, theemphasis is on informative and normative tests that attemptto assure component interoperability across standards.“Normative” tests are like compliance tests in the sense thatthe committee prescribes them to assure interoperability.“Informative” tests are recommended to develop a morethorough understanding of performance and margin.In this note, we draw from the OIF-CEI-03.1 IAs summarizedin Table 1.6www.tektronix.com/opticalThe long reach CEI-25G-LR consists of multiple lanesat 19.90-25.8 Gb/s of differential pairs over up to 686mm of PCB trace with up to two connectors. While thePCB channel is specified to a length up to 686 mm, thephysical length is less important than the channel frequencyresponse. The medium reach CEI-28G-MR consistsof multiple lanes at 19.90-28.10 Gb/s for signaling ofdifferential pairs over up to 500 mm of PCB trace, vias, andup to one connector. The short reach CEI-28G-SR consistsof multiple lanes at 19.90-28.05 Gb/s of differential pairsover 300 mm of PCB trace, vias, and up to one connector.The very short reach CEI-28G-VSR consists of multipleelectrical lanes at 19.60-28.05 Gb/s for signaling betweenserdes (called hosts in the IA) and transceivers (modulesin the IA). The serdes and transceiver can be separated bysome 100 mm of host PCB trace to a mated connector pairplus an additional 50 mm or so of transceiver trace.The OIF-CEI provides independent IAs for each level ofthe protocol stack in order to permit mixing and matchingof applications, like use of a SONET framer in an opticalmodule connected to a CEI electrical interface. Since FECoperates just above the media level, FEC is optional in OIFCEI.2.3. Fibre Channel 32GFCThe high rate Fibre Channel standard, 32GFC, has a datarate of 28.05 Gb/s. The confusing name scheme, 32GFC for28.05 Gb/s technology, comes from the desire for the nameof each generation to demonstrate that the payload rate,as opposed to the data rate, is double that of the previousgeneration. The confusion began with a large decrease inoverhead in the transition from 8GFC to 16GFC when thedata rate advanced from 8.5 to 14.025 Gb/s but the payloadrate doubled from 6.4 to 12.8 Gb/s. The payload rate for32GFC is 25.6 Gb/s, twice that of 16GFC, but the datarate, 28.05 Gb/s, is well short of that implied by the 32GFCabbreviation.

Physical Layer Tests of 100 Gb/s Communications Systems3. Testing 100G SystemsFigure 2 is a diagram of the components of typical 100Gsystems. A serdes serializes a signal and transmits four25 Gb/s differential pairs. The serdes may consist of eitherseparate components for each output or a single integratedcomponent. The 25 Gb/s electrical signals are transmittedfrom the serdes to an optical transceiver. The transceiverretimes the signals and transmits optical versions on eitherSM (single mode) or MM (multimode) optic fibers. A secondtransceiver receives the optical signals, converts them toelectrical signals and transmits them to another serdes fordeserializing. The purely electrical version follows the samescheme without the intermediate transceiver-driven opticalsignaling.Figure 2. Diagram of (a) 4 25 Gb/s 100G serdes-transceiver WDM optical system,(b) 4 25 Gb/s 100G serdes-transceiver optical system, and (c) a 4 25 Gb/s 100Gserdes to serdes electrical system. The figure does not show the symmetric returnpaths.Whether for transmitter or receiver testing, optical orelectrical, we need test patterns that put every aspect of acomponent and every component of a system to the test.The PRBSn (pseudo-random binary sequences of length2n-1) are standardized patterns with every permutation ofn bits. The CEI CID jitter tolerance pattern is designed tohave the most aggressive elements of the PRBS31, plus72 CID (consecutive identical) bit sequences but at a moremanageable total length than PRBS31.www.tektronix.com/optical7

Application NoteTest Patterns0x00ff square wave8 bits low, 8 bits highPRBS9511 bitsPRBS1532,767 bitsPRBS312.1 GbitsRS-FEC encoded scrambled idleCEI CID jitter tolerance pattern(72 CID bits 10328 from PRBS31 seed) complementTable 2. Test Patterns.All PatternPro and BERTScope pattern generators provideall test patterns used in 100G communications, includingPRBS31, RS-FEC scrambled idle, or for that matter, everycommon test pattern as well as any that you devise up to 2MB long.All transmitter tests, both electrical and optical, shouldbe performed with every system channel active in bothdirections to include all reasonable sources of crosstalkinterference. To prevent unrealistic data-dependentinterference, patterns on the crosstalk channels shoulddiffer from the test signal pattern. If it’s not possible foreach aggressor to transmit a unique pattern, introducesufficient delay between them so that the patterns aren’tsynchronized.Since the frequency response of PCB punishes highfrequency content, electrical signaling between serdes chipsor between serdes and transceivers across centimetersof PCB requires signal conditioning: pre-emphasis at thetransmitter and/or equalization at the receiver.The big difference between the 100 GbE electricalspecifications and previous HSS data specifications,in addition to the introduction of PAM4 signaling for100GBASE-KP4, is extensive characterization of channelsin terms of the frequency dependence of insertion andreturn loss and the introduction of COM (channel operatingmargin), a way to combine the signal impairments fromthe transmitter and channel into one signal-to-noise-likeparameter. Since frequency response determines ISI(inter-symbol interference), compliant channels restrict ISI8www.tektronix.com/opticalto levels that can be accommodated by the combinationof transmitter equalization—usually three taps of pre- orde-emphasis, a form of FFE (feed-forward equalization) asit is referred to in some of the standards documents—andreceiver equalization that includes a CTLE (continuoustime linear equalization) filter usually combined with a DFE(decision feedback equalizer). Since most 100G technologyrequires both transmitter and receiver equalization, channelinsertion loss must meet both minimum and maximumcriteria.COM is derived by measuring channel S-parameters andthen modeling the effect of crosstalk from other channels,and random noise and jitter, to calculate the ratio in dB ofsignal amplitude to aggregate noise amplitude at the targetdetector error ratio.By specifying a maximum compliant COM, the standardsgrant designers latitude in budgeting the combinationof ISI, random noise, random jitter, and crosstalk signalimpairments as they see fit in a way that still assuresinteroperability.Other differences include more careful parameter definitionlike UUGJ (unbounded uncorrelated Gaussian jitter) in placeof RJ (random jitter) and introduction of UBHPJ (unboundedhigh probability jitter) which combines PJ (periodic jitter) andcrosstalk. EBUJ (effective bounded uncorrelated jitter) andETUJ (effective total uncorrelated jitter) are terms derivedfrom the dual-Dirac model applied to the components of thejitter distribution that are uncorrelated to the data.

Physical Layer Tests of 100 Gb/s Communications SystemsSummary of Typical 100 GBE Optical Transmitter erage launch power-2.9 to 2.9 dBm-4.3 to 4.5 dBm-9 to 2.4 dBmOMA-1.3 to 4.5 dBm0.1 to 4.5 dBm-7 to 3 dBmExtinction ratio 8 dB 4 dB 2 dBEye mask {X1, X2, X3, Y1, Y2, Y3}{0.25, 0.4, 0.45, 0.25, 0.28, 0.4}{0.3, 0.38, 0.45, 0.35, 0.41, 0.5}Table 3. Summary of transmitter specifications.3.1. Testing optical transmittersTypical transmitter requirements are summarized in Table 3.Figure 3 defines the eye diagram, average launch power,OMA (optical modulation amplitude), and, together withthe values in Table 3, the eye mask criteria. The normalizedlogic 0 and 1 levels used in the eye-mask are defined by theaverages of the lower and upper halves of the central 0.2 UIof the eye. Eye masks are the do-not-enter regions of theeye diagram specified by six parameters, as shown in bothTable 3 and Figure 3.Eye mask tests can be performed on a DSA8300 low-noiseequivalent-time sampling oscilloscope or BERTScope. Inboth cases, wide bandwidth optical-to-electrical receiversand clock recovery units are necessary. The 3 dB clockrecovery bandwidths differ among specifications and canbe accommodated by CR286A, a digital-based, secondorder Phase-Locked Loop (PLL) with user-specified cornerfrequencies capable of tracking jitter to 23 MHz. The clockrecovery bandwidth for optical 100 GbE systems has acorner frequency of 10 MHz with slope of 20 dB/decade.Apply crosstalk by engaging pattern generators onthree other channels or however many parallel lanes yoursystem supports. Each crosstalk signal should meet therequirements for a compliant transmitter and their testpatterns should differ from the test signal pattern. If it’s notFigure 3. Definitions of the transmitter eye mask and OMA.possible for each aggressor to transmit a unique pattern,introduce sufficient delay between them so that the patternsaren’t synchronized.Optical signals should be analyzed with a measurementbandwidth slightly higher than the symbol rate.Unfortunately some specifications refer to this requirementas 0.75 fB because early oscilloscopes used independentoptical-electrical converters that had square-law detectors.The filter is required so that different test platforms canoperate under uniform measurement conditions.www.tektronix.com/optical9

Application NoteFigure 4. Eye mask measurement of a 100 GbE optical 25 Gb/s signal.Figure 5. Eye mask testing with BER contours. The BER 10-6 contour, the outer bluecontour, corresponds to a 5 10-5 hit ratio.The random nature of a mask test is addressed by requiringa minimum “hit ratio.” The hit ratio is defined as the ratioof the number of mask violations to the total number ofsamples acquired per unit interval. Since this is a statisticalmeasurement, it’s worthwhile to keep in mind that accuracyimproves with more hits.Alternatively, it’s both more statistically reliable and easier tomeasure the BER Contour on a PatternPro Error Detector,BERTScope, or on a DSA8300 oscilloscope equippedwith 80SJNB jitter and noise analysis software. As longas the BER 10-6 contour is outside the mask, Figure 5,the transmitter passes the 5 10-5 hit ratio eye test. Thistechnique also makes it easier to see the margin with whicha transmitter passes.A transmitter is compliant if it achieves a hit ratio less than5 10-5, Figure 4.10www.tektronix.com/optical

Physical Layer Tests of 100 Gb/s Communications SystemsSummary of 100 GBE Optical Receiver Test ConditionsLong and Extended Reach100GBASE-ER4100GBASE-LR4Average received power4.5 to -20.9 dBm4.5 to -10.6 dBmOptical Modulation Amplitude (OMA) -17.9 dBm -6.8 dBmVertical Eye Closure Penalty (VECP)3.5 dB1.8 dBSJAccording to the template in Figure 7Sinusoidal interference 0.1-2 GHzSum to J2 and J9Conditions for Stressed Receiver Tolerance Testing:RJJ2 jitter0.3 UIJ9 jitter0.47 UIBER requirement 10-12Short Reach100GBASE-SR4Average received power-10.9 to 2.4 dBmOptical Modulation Amplitude (OMA) 3 dBmStressed receiver sensitivity-5.6 dBmConditions for Stressed Receiver Sensitivity Testing

2.2. OIF-CEI Implementation Agreements (IAs) from OIF-CEI do not prescribe compliance tests the way that IEEE’s 802.3 100 GbE or Fibre Channel specifications do. Instead, the emphasis is on informative and normative tests that attem

Related Documents:

100%: 100%. 100%: 100%. 100%: 100%. 100%: 100%. 100%: 100%. 100%: 100%. 100%: 100%. 100%: 2. Plain Cement Concrete: 100%. 100%: 100%. 100%: 100%. 100%: 100%. 100% .

Office IP Phones Access Layer Distribution Layer Main Distribution Facility Core Switch Server Farm Call Servers Data Center Data/Voice/Video Pipe IDF / Wiring Closet VoIP and IP Telephony Layer 1 - Physical Layer IP Phones, Wi-Fi Access Points Layer 1 - Physical Layer IP Phones, W i-F Access Points Layer 2 - Distribution Layer Catalyst 1950 .

NFP121 5 Sommaire 1) Tests et tests unitaires - Outil : junit www.junit.org une présentation Tests d'une application - Une pile et son IHM Tests unitaires de la pile Tests de plusieurs implémentations de piles Tests d'une IHM Tests de sources java Invariant et fonction d'abstraction comme tests - Tests en boîte noire - Tests en boîte blanche

Physical Layer Specifications TS 36.201 E-UTRA Physical layer: General description . TS 36.211 E-UTRA Physical channels and modulation . TS 36.212 E-UTRA Multiplexing and channel coding . TS 36.213 E-UTRA Physical layer procedures . TS 36.214 E-UTRA Physical layer - Measurements The latest version of the specifications can be downloaded from:

Physical Layer Specifications TS 36.201 E-UTRA Physical layer: General description . TS 36.211 E-UTRA Physical channels and modulation . TS 36.212 E-UTRA Multiplexing and channel coding . TS 36.213 E-UTRA Physical layer procedures . TS 36.214 E-UTRA Physical layer - Measurements The latest version of the specifications can be downloaded from:

AP Biology Practice Tests 2 2020 2020 Practice Tests . AP Calculus AB Practice Tests ; 2 2020 . 2020 . Practice Tests . AP Calculus BC Practice Tests 2 2020 2020 . Practice Tests . AP Chemistry Practice Tests . 2 2020 . 2020 : Practice Tests AP Computer Science 2 2019 2020 Practice Tests . AP English Language and Composition Practice Tests : 2 2020

9. Build a sugar-cube pyramid as follows: First make a 5 5 1 bottom layer. Then center a 4 4 1 layer on the rst layer, center a 3 3 1 layer on the second layer, and center a 2 2 1 layer on the third layer. The fth layer is a single 1 1 1 cube. Express the volume of this pyramid as a percentage of the volume of a 5 5 5 cube. 10.

C. Rockwell hardness test LAMINATES RHN LAYER 1 95 LAYER 2 96 LAYER 3 97 LAYER 4 98 Table 4.2 Hardness number RHN rockwell hardness number D. Impact test LAMINATES ENERGY (J) DEGREE (ang) LAYER 1 1.505 105 B. LAYER 2 2.75 114 LAYER 3 3.50 124 LAYER 4 4.005 132 Table 4.3 Impact Test data E.