Abstract

This paper reviews the key factors in the discussion and selection process before the launch of the higher speed passive optical network (PON) standards project in the Full Service Access Network and ITU-T SG15/Q2. It reviews the requirements for such a system and the progress of the related ITU-T standards documents. The key technologies necessary for the physical and protocol layers of the 50G-PON are also discussed.

© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. EVOLUTION OF PON TECHNOLOGIES AND STANDARDIZATION STATUS

In recent years, work has been conducted within the main passive optical network (PON) standardization groups, i.e., the ITU-T and IEEE to increase the nominal line rate of PON systems beyond 10 Gbps. In the IEEE, the 802.3ca 50G Ethernet Passive Optical Network (EPON) Task Force has defined a system based on 25 Gbps nominal line rates. The ITU-T has followed a path to even higher nominal line rates of 50 Gbps.

Operators are already deploying PON at the 10 Gbps nominal line rate as an upgrade to gigabit-capable passive optical networks (GPONs) and EPON deployments made during the past 10 years or so. In the ITU-T discussions, it became clear that the next step in PON should offer line rates at least 4 times those being deployed now. This requirement has driven the work in the ITU-T, and the latest progress of ITU-T PON standards towards this goal is the subject of this paper. For those interested in a review of the IEEE PON standardization project, readers are directed to [1].

Even though this paper is focusing on the ITU-T PON standards, there will nevertheless be some reference to IEEE PON. This is because, as a general rule, the IEEE and ITU-T endeavor to ensure as much commonality between key aspects of their respective PON standards to minimize divergence and promote convergence. This is especially true for the physical layer to enable the use of common optical components, e.g.,  by adopting similar wavelength plans as much as possible.

A. FSAN Standards Roadmap 2.0

The Full Service Access Network (FSAN) group is a forum for the world’s leading telecommunications services providers, independent test labs, and equipment suppliers to work towards a common goal of truly broadband fiber access networks, with more than 70 member organizations. In the past, the FSAN group successfully collected and converged operator needs, promoted standardization projects and research into the appropriate standard bodies (e.g., ITU-T, Broadband Forum), and drove existing standards into industry products and services. In 2016, after long discussion and justification among the members, the FSAN group published the FSAN Standards Roadmap 2.0, as shown in Fig. 1 [2]. The FSAN roadmap serves as a guideline for the industry and sets expectations for what is needed from standardization bodies developing standards for PON systems, which serve as a stimulus for future technology development. The next steps for PON evolution are identified by the FSAN roadmap as XG(S)-PON+, i.e., a single-channel PON system beyond 10 Gbps line rates, and NG-PON2+, i.e., a multiple-wavelength channel PON system with a line rate increase beyond 10 Gbps. Standards for such systems are expected to be needed, according to the roadmap, around the year 2020.

 figure: Fig. 1.

Fig. 1. FSAN standards roadmap 2.0 [2].

Download Full Size | PPT Slide | PDF

B. ITU Standardization Direction Concerning Next-Generation Higher Speed PON

ITU-T SG15/Q2 launched the G.sup.HSP (where HSP = higher speed PON) project in 2016 (completed and released in February 2018 as ITU-T G.Sup.64 [3]), researching various key technologies and technical feasibility for PON networks providing higher speeds. This supplement describes the characteristics of optical transmission above 10 Gbps per wavelength between the optical line termination (OLT) and the optical network unit (ONU). The challenges of transmission above 10 Gbps in optical access were reviewed based on some assumed system requirements, signal modulation formats, optical transmitter/receiver design, coexistence requirements, and operating wavelength dependency, etc.

During the research for G.sup.HSP, one of the most important questions was what nominal line rate per wavelength should be selected for next-generation higher speed PON? From the beginning, 20, 25, 50, and even 100 Gbps nominal line rates per wavelength were considered as options for the next-generation PON after 10-gigabit passive optical network (XG-PON) [4], 10-gigabit-capable symmetric passive optical network (XGS-PON) [5], and 10 Gbps/ch time and wavelength division multiplexing (TWDM PON) [6]. Note that the nominal rate is the approximate (headline) rate, rather than the precisely defined system clock rate.

Given the past experience of PON deployment, the interval between the massive deployments of two adjacent generations of broadband access technologies is about 7–8 years. For example, large-scale group tenders for GPON were launched in the year 2011 in the China market, while large-scale group tenders for XG-PON were launched in the same market around the year 2018. Hence, following this experience, the reasonably expected time frame for massive deployment of the next-generation PON technology would be around the year 2025. Furthermore, a 4–5× increment of PON capacity also looks reasonable, based on the past history of PON development. Just like from GPON to XG-PON, the nominal line rate in the downstream direction increased from 2.5 Gbps to 10 Gbps, a 4× boost. Besides this, there are many salient factors that point to a target 4–5× increment in PON capacity, e.g., progress in technology maturity, equipment platform evolution, and service requirement increments.

XG(S)-PON (meaning XG-PON and/or XGS-PON) was identified by the ITU-T as the evolution path for GPON to provide a 4× downstream capacity upgrade. Therefore, a 25 Gbps upgrade step would be considered too close to 10 Gbps and so not sufficiently future proof and, hence, wasteful of the huge investment by network operators in fiber-to-the-home (FTTH) deployment. On the other hand, 100 Gbps would be nice to have, but this is currently considered too challenging when considered from the whole PON industry perspective.

From the perspective of system capacity, based on several assumptions including a 1:64 split ratio, a 50% service concurrency ratio, and an 80% payload ratio, a 50 Gbps nominal line rate can provide on average 1.25 Gbps service access capability simultaneously for 64 ONUs. Such a capacity meets the large bandwidth requirements brought by new services such as virtual reality (VR) and augmented reality (AR), which are becoming more and more popular in the residential market.

In 2018, as the leading international organization developing PON system standards, the ITU-T finally concluded around two years of study and discussion on the next generation of PON beyond the 10 Gbps nominal line rate. After much debate during the January 2018 Study Group 15 Plenary Meeting, it was decided to select 50 Gbps PON as the next generation of PON system technology and to establish a new G.HSP standard project series. The new series of standards projects will carry out corresponding technical research and standardization development work based on requirements, physical layer, and protocol layer aspects of HSP. This standard series will primarily consist of four main recommendations:

  • (1) ITU-T G.9804.1 (ex G.hsp.req) [7], Higher Speed Passive Optical Networks: Requirements includes overall system requirements, evolution and coexistence, and supported services and interfaces of high-speed PON systems. The standard achieved consent in July 2019 and was officially approved in November 2019.
  • (2) ITU-T G.hsp.50GPMD, Higher Speed Passive Optical Networks: 50G PMD: Physical Media Dependent (PMD) Layer Specification is a 50 Gbps time-division multiplexed PON (TDM PON) physical layer standard project addressing physical layer architecture, optical layer interface parameters, etc. This is still under development.
  • (3) ITU-T G.hsp.TWDMPMD, Higher Speed Passive Optical Networks: TWDM PMD includes 50 Gbps TWDM PON physical layer architecture, optical layer interface parameters, etc. This is in the early stages of development.
  • (4) ITU-T G.hsp.COMTC, Higher Speed Passive Optical Networks: Common Transmission Convergence (TC) Layer is a general TC layer standard project for HSP systems, including TC layer architecture, physical adaptation layer, business adaptation layer, management process, and message definition, etc.; this is also under development.

In addition, the ONU management for HSP systems is expected to be realized by revising the ITU-T G.988 standard [8].

2. REFERENCE ARCHITECTURE AND SYSTEM REQUIREMENTS FOR HSP SYSTEMS

Higher speed PON systems [7] include three subtypes: 50G TDM PON, 50 Gbps/ch TWDM PON, and higher speed 50 Gbps/ch point-to-point wavelength division multiplexed PON (PtP WDM PON).

The 50 Gbps TDM subtype is an evolution of conventional TDM/TDMA (time division multiple access) PON with a single-wavelength channel per direction, similar to that used for GPON and XG(S)-PON. TWDM PON is a multiwavelength PON whereby each wavelength channel pair is shared by multiple ONUs using TDM/TDMA. Finally, the PtP WDM subtype offers dedicated wavelength channel pairs to each ONU end point. In the case of TWDM and PtP WDM the ONUs need wavelength-tunable transmitters and receivers controlled to dense wavelength division multiplexing (DWDM) accuracy. For TDM PON, the ONU and OLT transmitters are nominally fixed wavelength but may drift within well-defined wavelength bands.

 figure: Fig. 2.

Fig. 2. 50 Gbps/ch TWDM PON reference architecture. UNI = user network interface, SNI = service network interface, ${{R}}$ = receive and ${{S}}$ = send when defining bidirectional PON interfaces, CG = channel group, CP = channel pair, CT = channel termination, WM = wavelength multiplexer.

Download Full Size | PPT Slide | PDF

 figure: Fig. 3.

Fig. 3. 50G TDM PON reference architecture.

Download Full Size | PPT Slide | PDF

The reference architectures [7] for both the 50 Gbps/ch TWDM PON and the 50G TDM PON systems are shown in Fig. 2 and Fig. 3, respectively. All HSP systems operate over a legacy power splitting-based optical distribution network (ODN). The 50 Gbps TWDM PON OLT is generally composed of multiple OLT channel terminations (CTs) connected via a wavelength multiplexer (WM). Each OLT CT represents a logical function that resides at the OLT equipment and terminates at a single TWDM channel, which refers to a pair of one downstream wavelength channel and one upstream wavelength channel providing point-to-multipoint access and connectivity by using TDM/TDMA mechanisms. In a specific implementation, the WM can be either integrated with multiple CTs or used externally as an independent device; each approach comes with its own advantages and disadvantages. A 50G TDM PON OLT can be considered as a special case of a 50 Gbps/ch TWDM PON system, which operates over a single wavelength channel. Hence, a WM is not needed for 50G TDM PON.

Table 1 summarizes the general technical requirements for HSP systems in terms of nominal line rates, fiber distance, and ODN characteristics. One of the most important features of the next-generation PON is that the nominal line rate in the downstream of each channel in the system works at a 50 Gbps line rate. For 50G TDM PON, this means that the system capacity is approximately 50 Gbps in the downstream direction and up to a nominal 50 Gbps in the upstream direction. The capacity in the upstream direction is expressed as “up to,” as there are different requirements for upstream capacity in different applications. Therefore, lower cost ONUs with around 10 Gbps or 25 Gbps may be used in residential services, where upstream data generation is generally lower than downstream data consumption.

Tables Icon

Table 1. General Technical Requirements for HSP Systems

After accounting for all reasonable overheads of forward error correction (FEC), operation, administration and maintenance (OAM), etc., a 50G TDM PON ONU is able to support the maximum service rate of approximately 40 Gbps, i.e., the actual capacity available for user services traffic after deducting overheads. Therefore, such a 50G TDM PON ONU is able to offer multiple 10 GbE interfaces or a single 25 GbE interface to the subscriber. It is common in TDM/TDMA PON systems for network operators to permit subscribers to have access up to the peak rate of the PON during periods in which the PON loading is low. To do this, the ONU interfaces to the user must be capable of operating at the peak data rate for the service offering. This does not mean that all subscribers will be able to access this data rate simultaneously.

For a 50 Gbps/ch TWDM PON, the total system capacity is 50 Gbps multiplied by the number of OLT CTs supported. Typically, each CT/ONU shall have a similar service capability as for 50G TDM PON. However, in some special cases that exploit the multiwavelength characteristics, a 50 Gbps/ch TWDM PON ONU can be enabled to offer higher service access capabilities through a wavelength channel bonding function.

3. EVOLUTION PATH AND COEXISTENCE TECHNOLOGY

An important characteristic for high-speed PON in standardization studies is to support the requirement for smooth coexistence evolution, specifically, to support the coexistence of legacy PON systems and high-speed PON systems in the same fiber infrastructure (ODN); the service interruption of ONUs during evolution and upgrade should be avoided or minimized as much as possible; an HSP system should continue to support services running on the legacy PON system after evolution and upgrade. The legacy PON systems in the above requirements include the previous PON systems defined in ITU-T, such as GPON, XG(S)-PON, and TWDM PON (${{4} {\sim} {8}}$ channels × 10 Gbps). The wavelength plans of the relevant PON systems as shown in Figs. 4(a) and 4(b) show how GPON and XG(S)-PON users can be smoothly upgraded to 50G TDM PON, using three PON generation coexistence in one ODN, based on the option UW1 50G TDM PON upstream wavelength (for further details, see Section 4.C). At the same time, in order to support PON convergence [11], it was agreed that ITU-T 50G TDM PON will also support IEEE 10G EPON coexistence and a smooth evolution.

 figure: Fig. 4.

Fig. 4. (a) ITU-T PON wavelength plan and (b) coexistence and upgrade evolution diagram. GPON US: ${{N}}$ = narrow and ${{R}}$ = reduced; XGS-PON: only the basic wavelength set is shown; 50G TDM PON: ${{W}}$ = wide and ${{N}}$ = narrow.

Download Full Size | PPT Slide | PDF

HSP provides various upgrade paths, supporting legacy PON generations, such as GPON, XG-PON, XGS-PON, and 10G EPON, to evolve to higher system capabilities. Specific coexistence and upgrade evolution paths are shown in Fig. 5. XG(S)-PON systems can be upgraded to 50G TDM PON smoothly with a 5× system capacity increase. This is the mainstream evolution path supporting the coexistence of two PON generations. In the ITU-T G.9804.1 standard [7], three generations of PON system coexistence are also described, and this is understood to come with higher technical difficulty on wavelength band selection and optical PMD parameters.

 figure: Fig. 5.

Fig. 5. PON system coexistence and upgrade evolution paths.

Download Full Size | PPT Slide | PDF

One feasible evolution path for 10G EPON to be upgraded with 50G TDM PON is under the premise that the wavelength band of the upstream 1 Gbps line rate signal is narrowed to 1260–1280 nm. This is especially meaningful for those operators with both XG(S)-PON and 10G EPON in their networks, as they may achieve technology convergence in the next PON generation and so greatly simplify network management and reduce costs.

In the case of GPON, the narrow and reduced upstream wavelength options, as defined in G.984.5 [12], are able to coexist with HSP. This is not a significant restriction, as the 1260–1360 nm wavelength option has not been deployed in any significant volume.

For TWDM PON, the overall migration path is clear and migration directly to the 50 Gbps/ch TWDM PON is the obvious evolution step. Even so, the full details (e.g.,  number of channels, wavelength plan) of this migration path are still under study in the ITU-T.

Because the HSP system needs to support technology evolution by coexistence, the HSP system should be able to support operation on the legacy PON ODN infrastructure and, under ideal circumstances, work in a wavelength band that has not been occupied by any legacy PON systems. However, if a particular legacy PON system is not deployed and is not required in the future to coexist with HSP, then the HSP system can use the wavelength band of that particular legacy PON system.

There are two main methods to achieve PON evolution and upgrade: by external coexistence elements (CEx) (WDM1r, CEx, etc.) defined in ITU-T G.984.5 [12] and by a multi-PON module (MPM), as shown in Fig. 6 [7]. The main difference between the two methods is whether the WM and demultiplexer for different PON system signals is inside the OLT optical module or in the ODN link. This difference will first result in different values for the PMD layer optical parameters in the standard, such as the number of optical path loss classes, OLT transmitter launch power, and receiver sensitivity. The practical difference between these two approaches is realized in the network operator engineering and construction procedures in actual deployments.

 figure: Fig. 6.

Fig. 6. PON evolution and upgrade methods: (a) by external coexistence elements and (b) by MPM.

Download Full Size | PPT Slide | PDF

Tables Icon

Table 2. Coexistence Methods and Scenarios for 50G TDM PONa

With the external CEx method, the operators need to deploy the CEx devices in advance, and add a new OLT line card when upgrading to a next-generation PON system. On the other hand, the MPM method has certain technical and cost advantages in the large-scale deployment stage. In large-scale deployments, when the MPM method is adopted, operators can greatly reduce costs, such as reducing the demand for central office space, saving OLT equipment frame occupation, reducing power consumption, and so on. In addition, the MPM method also helps to simplify engineering operations in large-scale network upgrades; it simplifies the technical requirements of constructors and reduces failure rates. With the MPM method, the operators can just remove the previous PON OLT line card and replace it with the new generation OLT line card, thus reusing the previous line card slot, with no need to add a new OLT mainframe and perform fiber rearrangement. Due to the replacement of the OLT line card in the MPM case, the operation process will involve a short business interruption (normally less than 5 min). This short business interruption could be avoided in the external CEx case assuming it is preinstalled. These experiences have also promoted the definition of new technical solutions in the development of the HSP standards series, thereby formulating even better and industry-leading technical standards. Recently, the MPM method has been growing in acceptance in many network operators around the world and is widely used for upgrading GPON networks to XG(S)-PON networks.

The ITU-T G.9804.1 standard adopts the above two coexistence technologies, and the current version specifies the specific coexistence scheme for legacy PON systems to evolve to 50G TDM PON via two methods under different scenarios.

As shown in Table 2, the current version of the ITU-T G.9804.1 standard [7] describes three types of coexistence scenarios: XG(S)-PON and 50G TDM PON coexist (two PON generations); two generations of 10G EPON coexist with 50G TDM PON (two PON generations); and GPON, XG(S)-PON, and 50G TDM PON coexist (three PON generations). It also describes coexistence and upgrade options based on either (1) external multiplexers and demultiplexers (coexistence elements) or (2) MPM methods for each coexistence scenario. The features and main points of each scheme have been briefly listed in Table 2. At present, during the standards study process, PMD layer parameters and TC layer functions are being developed to support various evolution and upgrade technology approaches. Operators can choose the appropriate upgrade path and technical solution according to the actual situation of their own network.

4. SUMMARY OF SOME OF THE KEY STUDY ITEMS FOR 50G TDM PON

A. New Interfaces Supported by HSP

HSP will, as a baseline, fully support the existing service requirements of the traditional legacy PON systems (NG-PON2, XGS-PON, and GPON) as defined in their corresponding standards. Besides these, at the interface level, HSP newly supports 25GBASE [13] physical layer UNI interfaces, which can carry 25 Gbps Ethernet services and Ethernet-based enhanced Common Public Radio Interface (eCPRI) services defined in [14]; new CPRI option 10 UNI interfaces, which can carry 24,330.24 Mbps wireless fronthaul services; and 200 GbE/New SNI interfaces such as 400 GbE, which improve the convergence capability of PON equipment uplink interfaces.

B. Modulation

The 50G TDM PON physical layer specification and common TC specification are currently under active discussion and development in the standardization study phase. After more than a year of study, discussion, and some debate, the current PMD standard (draft) has determined that 50G TDM PON uses the non-return-to-zero (NRZ) on–off keying (OOK) modulation method for both uplink and downlink. The main reason for this is that it needs to operate with high optical path loss (OPL) in the ODN with, at least, all the loss budget classes up to 32–33 dB being supported. NRZ has a sensitivity advantage compared to 4-level pulse amplitude modulation (PAM4), for example, and can meet the needs of high OPL without the need for unreasonably high transmitter launch power.

The purpose of standards is primarily to enable interoperability between the OLT and the ONU. Therefore, the standard only defines the modulation format at the transmitter side and does not specify or limit the demodulation method at the receiver side. Manufacturers have multiple technical options when actually implementing standards-compliant solutions. For example, as illustrated in Fig. 7, the 50 Gbps PON can receive using either an avalanche photodiode (APD) receiver capable of direct NRZ reception at 50 Gbps or lower bandwidth devices such as APDs used for 25 Gbps NRZ combined with digital signal processing (DSP) technology [15,16]. High-speed DSP technology can compensate for insufficient system bandwidth, chromatic dispersion-induced penalties, and eliminate intersymbol interference caused by other factors. This can improve the sensitivity of the receiver, reduce the optical device bandwidth requirements, and reduce the cost of the entire system. In addition, it is also possible to realize the reception of 50 Gbps high-speed signals based on low-bandwidth optical devices, by using the electrical duobinary or three-level demodulation method on the receiving side [17,18]. In standards for HSP systems, it is not intended that DSP or a specific reception method should be part of any recommendation. This is not needed to ensure interoperability at defined optical interfaces, and vendors need some freedom to differentiate their products.

 figure: Fig. 7.

Fig. 7. Implementation options for 50 Gbps NRZ signal reception.

Download Full Size | PPT Slide | PDF

C. Wavelength Band Plan

The wavelength plan of 50G TDM PON is close to being finalized, as shown in Fig. 8. So far, the downstream wavelength of 50G TDM PON has been agreed to be 1340–1344 nm. The main reason for this choice is that the fiber chromatic dispersion in the O-band is relatively low, which is beneficial in overcoming the dispersion-induced intersymbol interference in high-speed 50 Gbps NRZ signal transmission. The 4 nm wavelength range is selected to provide sufficient fabrication tolerance on the center wavelength of the transmitter laser and so reduce the cost of the optical device. Such a wavelength tolerance can be met by temperature-controlled transmitter modules.

 figure: Fig. 8.

Fig. 8. Wavelength plan of 50G TDM PON.

Download Full Size | PPT Slide | PDF

The upstream wavelength of 50G TDM PON has been allocated multiple options. The main reason for this is that different operators globally have different coexistence requirements. For example, some operators are more interested in coexistence with 10 Gbps PON, and GPON coexistence is not considered to be necessary when 50G TDM PON starts to be deployed, while for other operators, GPON coexistence is considered very important, as they expect that the GPON users will exist in their networks for a very long time.

So, in order to support various bit rates and coexistence requirements, respectively, covering 10 Gbps and 25 Gbps rates, the following upstream wavelength options were adopted:

  • Option 1: for coexisting with XG(S)-PON/10G EPON using WDM: wideband, ${{1290}}{\sim}{{1310}}\;{\rm{nm}}$; narrowband, ${{1298}}{\sim}{{1302}}\;{\rm{nm}}$;
  • Option 2: for coexisting with GPON supporting both narrowband (1300–1320 nm) and reduced (1290–1330 nm) upstream wavelength options: wideband, ${{1260}}{\sim}{{1280}}\;{\rm{nm}}$.
  • (The strict necessity for a narrowband option for HSP in this Option 2 use case is still under discussion.)

The inclusion of wideband (20 nm) options is mainly to facilitate the use of uncooled transmitters (e.g.,  directly modulated lasers, DMLs) to enable low-cost ONU implementations. As there are already some existing WDM1r deployed for the narrow GPON upstream wavelength band in the current ODNs, the narrowband option (${{1298}}{\sim}{{1302}}\;{\rm{nm}}$) was agreed on for HSP to enable better compatibility with these deployed WDM1r (and so enabling reuse of these for 50G TDM PON upgrades). The narrow option will necessitate cooling on the ONU side, which can lead to increased costs, but also can enable higher transmit power and facilitate higher OPL.

The downstream and upstream wavelength gap for Option 2 is up to 60 nm, which is favorable for a bidirectional optical subassembly (BOSA) design to separate the downstream and upstream signal in OLT/ONU modules. However, the gap is only 30 nm for Option 1 wideband, and this presents some difficulty for a typical 45-deg filter BOSA design due to the limited wavelength source. There are several methods to separate the downstream/upstream signal with a 30 nm gap using BOSA designs in the industry already, such as using collimated beams, small-angle reflecting filter structures, and depolarization filters [19,20].

The above wavelength plans are those agreed upon for 10 Gbps and 25 Gbps upstream line rates. The upstream wavelength plan for the 50 Gbps line rate has yet to be decided. The main reason for this is that, from the perspective of network operator deployment, the 50 Gbps symmetric rate combination has a lower priority according to the business demand time frame compared to the asymmetric upstream and downstream rate combination options. At the same time, the feasibility at 50 Gbps of the burst-mode receiving technology necessary for the upstream traffic has not been fully verified yet. How to realize reliable 50 Gbps burst-mode reception with a 33 dB ODN power budget is under active study, with initial proposals concerning this subject reviewed in recent ITU-T meetings.

The key discussion points and technical items under consideration with respect to the 50 Gbps wavelength plan generally relate to the importance of different coexistence/evolution scenarios envisaged by the operators and the technical feasibility of the high optical budget. In PON systems, there is a strong guideline to put the challenging technical requirements on the OLT side of the link, as the costs of that transceiver are shared among all the subscribers. Costs at the ONU side scale with the number of subscribers in the FTTH scenario. However, there is still a balance to be struck, as the volume of ONUs is also much higher, so costs can be reduced due to the volume effect. Furthermore, the OLT requirements need to be feasible in mass production with a high yield so they cannot be unreasonably hard. Therefore, at 50 Gbps, the tendency would be for the OLT receiver to have a high sensitivity requirement to relax the launch power of the ONU. Some contributors in the ITU debate argue that a high sensitivity requires an optically preamplified receiver at the OLT unless better APD technologies can be developed. This would then drive for an ONU with narrow wavelength tolerance for the transmitter to allow the optical amplifier noise to be filtered at the OLT. Such a narrow wavelength range will come with some extra cost for the ONU transmitter. Furthermore, these costs might also be added to ONUs with 10 Gbps and 25 Gbps rates if a triple-rate receiver is considered to enable TDM coexistence. The effect and importance of all of these cost and performance trade-offs need to be fully studied to determine the right choices. The use of other wavelength bands is also part of the debate in the case of three-generation (GPON, 10G-PON, and 50G TDM PON) coexistence enabled by WDM. At the time of writing this paper, the wavelength plan for 50 Gbps in the upstream direction is still, therefore, a topic needing much further study within the ITU-T.

The full specifications of optical interface parameters for 50G TDM PON are still in the development process. As of February 2020, the Q2/15 group in the ITU-T has reached an agreement on the sensitivity of the 25 Gbps upstream, i.e., ${-}{{25}}\;{\rm{dBm}}$ at ${{1}}{{{0}}^{- 2}}$ bit-error ratio (BER), which is the same as that in IEEE P802.3ca [21] and can be supported by the industry already [22,23].The major open items left are the following:

  • (a) the downstream PMD specifications, including OLT launch power and ONU receiver sensitivity requirement;
  • (b) the upstream specifications for different line rates, like 10 Gbps upstream, 25 Gbps upstream, and 50 Gbps upstream;
  • (c) the detailed method to define PMD parameters, such as using the optical modulation amplitude (OMA) [21] or average power (AVP) method to define launch power and receiver sensitivity, using transmitter dispersion eye closure (TDEC) [24,25] or optical path penalty (OPP) [5] to define the transmitter compliance, how to consider the balance between testing convenience for operators and the specification flexibility to allow versatile technology options from the vendors, and so on.

During the two years of discussion in the ITU-T since the project started, there have been many meeting contributions suggesting parameter specifications for consideration. The standardization plan of ITU-T SG15 is to complete and consent to the 50G TDM PON physical layer (PMD) standard in October 2020. Given the progress to date, it is expected that agreement can be achieved in the near term and that the target completion time scale can be met, but it is also possible to slip by one plenary meeting cycle.

D. FEC Code

With the increasing of the transmission rate, it is more and more difficult to realize the high OPL classes of the ODN, such as class C+ (32 dB) and class D (35 dB). In order to relax the requirements for the optical transceivers, a more powerful FEC encoding method will be introduced to 50G TDM PON at the data link layer. Compared with the Reed–Solomon (RS) code used in XG(S)-PON, the low-density parity check (LDPC) code [2629] introduced in HSP has a higher error rate threshold at the approximate code rate.

In order to achieve the output BER target of ${{1}}{{{0}}^{- 12}}$, the allowable input BER is relaxed from ${{1}}{{{0}}^{- 3}}$ (RS [255, 223]) to ${{1}}{{{0}}^{- 2}}$ (LDPC [17280, 14592] with hard decision), which will bring about 2 dB coding gain, as shown in Table 3 [30]. Furthermore, the code rate of the LDPC can be similar to the RS code, and the decoding can be implemented using either hard or soft decision approaches. By combining LDPC with DSP technology and using soft decision, the input BER could be relaxed to even ${2.1} \times {{1}}{{{0}}^{- 2}}$, which will bring an additional 1.0 dB coding gain referring to the example simulation shown in Fig. 9 [31].

Tables Icon

Table 3. Performances of Different FEC Codes

 figure: Fig. 9.

Fig. 9. Simulated BER performance improvement from soft decision versus hard decision LDPC.

Download Full Size | PPT Slide | PDF

The latency of LDPC decoding depends on the code word length, average iteration times, and degree of implementation parallelism. According to our estimation, the average decoding time of the LDPC is less than 4 µs, which is relatively small compared with the DBA latency in a TDM PON system (${\sim}$ several tens of microseconds) and the quiet window latency for new ONU registration (up to 250 µs).

E. Low Latency Requirements and Corresponding TC Layer Functions

As HSP will be used to carry different types of services, the delay requirements are designed and standardized according to the characteristics of these services, as shown in Table 4.

Tables Icon

Table 4. Latency Requirement

TDM PON systems use the TDMA mechanism for communicating with multiple ONUs connected to the same OLT PON port. Therefore, DBA mechanisms and ONU activation mechanisms are needed in the TC layer, which periodically introduce additional delay to the data transmission in the upstream direction. In real implementations, the data transmission delay of the PON link is considered to fluctuate within about 1.5 ms between service node interface (SNI) and user–network interface (UNI). And when the OLT creates a quiet window as part of the ONU activation mechanism, the data transmission delay is significantly increased as a consequence.

Use as a wireless bearer is a potential application scenario of HSP, but wireless transport has stricter requirements on link delay than conventional PON applications. Therefore, for wireless bearer scenarios and considering issues in the PON operation method, better optimized approaches are needed. In order to meet the delay requirements of the wireless bearer, the HSP system allows an ONU to activate and range through an additional wavelength channel, which can avoid the introduction of a 250 µs delay in the data channel from the opening of the quiet window in PON. Combined with new features such as single-frame multiburst and cooperative dynamic bandwidth allocation (CoDBA) [7], HSP can better support low-latency services such as wireless transport.

F. Generalization of the TC Layer to Achieve Broad Commonality and Future Extensibility

The intention in naming the TC layer recommendation in the HSP series as ComTC is that experts in the ITU-T Q2/SG15 group wanted to make a line-rate-extensible HSP TC layer. Such a TC layer can adapt to HSP and future PON systems without fundamental or architectural change. Upon analyzing previous PON TC layer recommendations, a key observation is that most of the TC specification elements are rate-independent. They are generic in their current state and capable of supporting various line rates in PON systems. The two possible places where line-rate-dependent parameterization is beneficial are in the downstream physical interface (PHY) frame length and the GrantSize granularity. During the recent development of the ComTC project, line rate (Parameter: ${{R}}$) was accepted as a scaled parametric variable for those places associated with line rate in the ComTC recommendation. A supported line rate would be set as a parameter ${{R}}$ value used in the ComTC recommendation for one of following purposes: (a) an index for some TC functions related to line rate or (b) a parametric variable to express certain values in the ComTC document. For example, the PHY frame payload in the downstream direction varies with the downstream line rate. Now, the PHY frame payload in the downstream direction can be expressed as a simple equation,

$${{R \times 125 \times 125}} - {{L_{\rm PSBd}}},$$
where ${{{L}}_{\rm PSBd}}$ (unit: byte) is the length of the physical synchronization block, downstream (PSBd) field and ${{R}}$ (unit: Gbps) is the line rate in the downstream direction. In this way, the specific line rate related TC layer features can be replaced by expressions of parameter ${{R}}$, which improves the ComTC document commonality and future extensibility.

5. SUMMARY AND CONCLUSION

Over the course of 2018–2020, the PON industry has been, and remains, focused on the ITU-T standardization effort concerning requirements, PMD, and ComTC for HSP systems. Prototype systems can be expected by the end of 2020. Experts in the industry are conducting active technical research, innovations, and industry analysis to determine the most appropriate technical solutions, including PMD parameters and TC layer functions driving technology maturity, and ensuring acceptable costs for future massive PON deployment.

It can be expected that higher-speed PON will benefit from advances in other application spaces, such as DSP technology from the wireless field, LDPC encoding from high-speed optical transmission, and high-speed components from Ethernet and data centers. Based on the favorable architecture and functions provided by the G.9804.x series of standards, the HSP system can provide flexible 50 Gbps system capabilities, multiple upstream and downstream rate access, and flexible latency control and other capabilities on one single PON infrastructure for residential customers, small and medium enterprises, industrial parks, wireless bearers, and other scenarios, which can open new commercial opportunities for future optical access networks.

Acknowledgment

We thank the coeditors of G.9804.1, G.hsp.50GPMD, and G.hsp.COMTC for the information provided during the writing of this paper and the contributors to the HSP study in the ITU-T standards group and industry researchers for information offered through their contributions and presentations.

REFERENCES

1. J. S. Wey, “The outlook for PON standardization: a tutorial,” J. Lightwave Technol.38, 31–42 (2020). [CrossRef]  

2. “FSAN standards roadmap 2.0,” 2016, https://www.fsan.org/roadmap/.

3. “PON transmission technologies above 10 Gb/s per wavelength,” ITU-T Recommendation G.Sup64, 2018, https://www.itu.int/rec/T-REC-G.Sup64/en.

4. “10-gigabit-capable passive optical networks (XG-PON): general requirements,” ITU-T Recommendation G.987.1, 2016, https://www.itu.int/rec/T-REC-G.987.1/en.

5. “10-gigabit-capable symmetric passive optical network (XGS-PON),” ITU-T Recommendation G.9807.1, 2017, https://www.itu.int/rec/T-REC-G.9807.1/en.

6. “40-gigabit-capable passive optical networks (NG-PON2): general requirements,” ITU-T Recommendation G.989.1, 2015, https://www.itu.int/rec/T-REC-G.989.1/en.

7. “Higher speed passive optical networks: requirements,” ITU-T Recommendation G.9804.1, 2019, https://www.itu.int/rec/T-REC-G.9804.1-201911-I/en.

8. “ONU management and control interface (OMCI) specification,” ITU-T Recommendation G.988, 2019, https://www.itu.int/rec/T-REC-G.988/en.

9. “Characteristics of a single-mode optical fibre and cable,” ITU-T Recommendation G.652, 2016, https://www.itu.int/rec/T-REC-G.652-201611-I/en.

10. “Characteristics of a bending-loss insensitive single-mode optical fibre and cable,” ITU-T Recommendation G.657, 2016, https://www.itu.int/rec/T-REC-G.657-201611-I/en.

11. “SDOs team up on PON convergence,” LightReading, 2017, https://www.lightreading.com/gigabit/fttx/sdos-team-up-on-pon-convergence/d/d-id/731234.

12. “Gigabit-capable passive optical networks (G-PON): enhancement band,” ITU-T Recommendation G.984.5, 2018, https://www.itu.int/rec/T-REC-G.984.5/en.

13. “IEEE standard for Ethernet—Amendment 2: media access control parameters, physical layers, and management parameters for 25 Gb/s operation,” IEEE P802.3by, 2016, https://ieeexplore.ieee.org/document/7457590.

14. “Common Public Radio Interface: eCPRI interface specification,” eCPRI specification V2.0, 2019, http://www.cpri.info/downloads/eCPRI_v_2.0_2019_05_10c.pdf.

15. D. Liu and M. Tao, “50G single wavelength PON analysis and comparison,” in IEEE P802.3ca 50G-EPON Task Force Meeting, November 2017.

16. V. Houtsma, D. van Veen, and E. Harstead, “Recent progress on standardization of next-generation 25, 50, and 100G EPON,” J. Lightwave Technol.35, 1228–1234 (2017). [CrossRef]  

17. D. van Veen and V. Houtsma, “50 Gbps low complex burst mode coherent detection for time-division multiplexed passive optical networks,” in European Conference on Optical Communication (ECOC), September 2018.

18. V. E. Houtsma and D. T. Van Veen, “Investigation of modulation schemes for flexible line-rate high-speed TDM-PON,” J. Lightwave Technol.38, 3261–3267 (2020). [CrossRef]  

19. T. Funada and T. Kihara, “Consideration on US/DS WDM filter for ONU,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January 2017.

20. Z. Liao and D. Liu, “NG-EPON diplexer filter analysis,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January 2017.

21. IEEE P802.3ca 50G-EPON Task Force, “Physical layer specifications and management parameters for 25 Gb/s and 50 Gb/s passive optical networks,” http://www.ieee802.org/3/ca/.

22. N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

23. F. J. Effenberger, H. Zeng, and X. Liu, “Burst-mode error distribution and mitigation in DSP-assisted high-speed PONs,” J. Lightwave Technol.38, 754–760 (2020). [CrossRef]  

24. F. Chang, ed., Datacenter Connectivity Technologies: Principles and Practice (River, 2018).

25. J. Petrilla, P. Dawe, and G. LeCheminant, “New metric offers more accurate estimate of optical transmitter’s impact on multimode fiber-optic links,” in DesignCon, Santa Clara, California, January 2015.

26. I. B. Djordjevic, O. Milenkovic, and B. Vasic, “Generalized low-density parity-check codes for optical communication systems,” J. Lightwave Technol.23, 1939–1946 (2005). [CrossRef]  

27. I. B. Djordjevic and B. Vasic, “Iteratively decodable codes from orthogonal arrays for optical communication systems,” IEEE Commun. Lett.9, 924–926 (2005). [CrossRef]  

28. S. Sankaranarayanan, I. B. Djordjevic, and B. Vasic, “Iteratively decodable codes on m-flats for WDM high-speed long-haul transmission,” J. Lightwave Technol.23, 3696–3701 (2005). [CrossRef]  

29. I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett.16, 338–340 (2004). [CrossRef]  

30. F. Effenberger, “Enhanced FEC consideration for 100G EPON,” in IEEE P802.3ca 50G-EPON Task Force Meeting, March 2016.

31. M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

Dezhi Zhang received a bachelor’s degree in engineering from Southeast University in 1998 and has been working in the field of communications. Mr. Zhang currently works with China Telecom Research Institute, is a senior expert in the field of optical access of China Telecom, mainly engaged in technical research, innovation, and application in the field of optical access, and actively participates in the standardization research of ITU-T, FSAN, and other international standard organizations. He has been heavily involved in the R&D and standardization effort of optical access networks, such as the XG-PON field trial and massive deployment, XG(S)-PON interoperability, XG-PON protocol design, and the WDM-PON trail for 5G fronthaul. Mr. Zhang is a co-editor of ITU-T Recommendations G.9807.1, G.989.3, G.9804.1, and G.984.5. He holds more than 10 Chinese patents.

Dekun Liu received a doctoral degree in integrated electronic and optical components from Zhejiang University in 2010 and has been working in the field of optical communications. Dr. Liu currently works at Huawei Technologies Co., Ltd., mainly engaged in research on advanced technology of optical access networks; he is also intensively involved in international standards on optical access systems. Dr. Liu is currently the associate rapporteur of ITU-T SG15 Question 1 and the co-editor of ITU-T Recommendations G.9807.1, G.989.2, G.hsp.50Gpmd, and L.oehc. He holds more than 30 patents and has authored/coauthored more than 10 papers in international journals.

Xuming Wu was born in Nanchang, Jiangxi, China, in 1986. Dr. Wu received his bachelor’s degree in electrical science and technology from Tianjin University in 2008. He studied at the China University of the Chinese Academy of Sciences from 2008 to 2013 and received his doctoral degree in microelectronics and solid-state electronics in 2013. He has been working on technical research of optical access networks in Huawei Technologies Co., Ltd., as a senior engineer since 2013. During this period, he has written 4 publications and holds more than 10 patents.

Derek Nesset received a bachelor’s degree with honors in physics from Birmingham University, UK, in 1989 and a master’s degree in telecommunications engineering from the University of London in 1995. He joined BT in 1989 and spent several years developing photonic components for fiber-optic communication systems. Following this, he worked on advanced fiber-optic system technologies up to 100 Gbps. In 2000, Prof. Nesset joined Marconi, where he was responsible for the ROADM subsystem development for ultra-long-haul DWDM. He returned to BT in 2003 to pursue research interests in enhanced PON systems for fiber access. He has spent over 15 years on next-generation PON technologies and standards and actively contributed to making progress on these topics in both FSAN and ITU-T. He chaired the NG-PON task group in FSAN from 2011 to 2016 and was an editor of ITU-T G.9807.1 (XGS-PON). Prof. Nesset now works for Huawei Technologies. His current research focus is optical transceiver components for next-generation PON systems. He was awarded an honorary professorship at Bangor University in 2014 and has contributed over 90 journal and conference publications and has 10 patents. He has participated in OFC, ECOC, OECC, and ACP conference subcommittees and chaired the Optical Access Subcommittee for OFC 2014 and the Fiber Based Networks Subcommittee for ECOC 2019.

References

  • View by:
  • |
  • |
  • |

  1. J. S. Wey, “The outlook for PON standardization: a tutorial,” J. Lightwave Technol. 38, 31–42 (2020).
    [Crossref]
  2. “FSAN standards roadmap 2.0,” 2016, https://www.fsan.org/roadmap/ .
  3. “PON transmission technologies above 10 Gb/s per wavelength,” ITU-T Recommendation G.Sup64, 2018, https://www.itu.int/rec/T-REC-G.Sup64/en .
  4. “10-gigabit-capable passive optical networks (XG-PON): general requirements,” ITU-T Recommendation G.987.1, 2016, https://www.itu.int/rec/T-REC-G.987.1/en .
  5. “10-gigabit-capable symmetric passive optical network (XGS-PON),” ITU-T Recommendation G.9807.1, 2017, https://www.itu.int/rec/T-REC-G.9807.1/en .
  6. “40-gigabit-capable passive optical networks (NG-PON2): general requirements,” ITU-T Recommendation G.989.1, 2015, https://www.itu.int/rec/T-REC-G.989.1/en .
  7. “Higher speed passive optical networks: requirements,” ITU-T Recommendation G.9804.1, 2019, https://www.itu.int/rec/T-REC-G.9804.1-201911-I/en .
  8. “ONU management and control interface (OMCI) specification,” ITU-T Recommendation G.988, 2019, https://www.itu.int/rec/T-REC-G.988/en .
  9. “Characteristics of a single-mode optical fibre and cable,” ITU-T Recommendation G.652, 2016, https://www.itu.int/rec/T-REC-G.652-201611-I/en .
  10. “Characteristics of a bending-loss insensitive single-mode optical fibre and cable,” ITU-T Recommendation G.657, 2016, https://www.itu.int/rec/T-REC-G.657-201611-I/en .
  11. “SDOs team up on PON convergence,” LightReading, 2017, https://www.lightreading.com/gigabit/fttx/sdos-team-up-on-pon-convergence/d/d-id/731234 .
  12. “Gigabit-capable passive optical networks (G-PON): enhancement band,” ITU-T Recommendation G.984.5, 2018, https://www.itu.int/rec/T-REC-G.984.5/en .
  13. “IEEE standard for Ethernet—Amendment 2: media access control parameters, physical layers, and management parameters for 25 Gb/s operation,” IEEE P802.3by, 2016, https://ieeexplore.ieee.org/document/7457590 .
  14. “Common Public Radio Interface: eCPRI interface specification,” eCPRI specification V2.0, 2019, http://www.cpri.info/downloads/eCPRI_v_2.0_2019_05_10c.pdf .
  15. D. Liu and M. Tao, “50G single wavelength PON analysis and comparison,” in IEEE P802.3ca 50G-EPON Task Force Meeting, November2017.
  16. V. Houtsma, D. van Veen, and E. Harstead, “Recent progress on standardization of next-generation 25, 50, and 100G EPON,” J. Lightwave Technol. 35, 1228–1234 (2017).
    [Crossref]
  17. D. van Veen and V. Houtsma, “50 Gbps low complex burst mode coherent detection for time-division multiplexed passive optical networks,” in European Conference on Optical Communication (ECOC), September2018.
  18. V. E. Houtsma and D. T. Van Veen, “Investigation of modulation schemes for flexible line-rate high-speed TDM-PON,” J. Lightwave Technol. 38, 3261–3267 (2020).
    [Crossref]
  19. T. Funada and T. Kihara, “Consideration on US/DS WDM filter for ONU,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.
  20. Z. Liao and D. Liu, “NG-EPON diplexer filter analysis,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.
  21. IEEE P802.3ca 50G-EPON Task Force, “Physical layer specifications and management parameters for 25 Gb/s and 50 Gb/s passive optical networks,” http://www.ieee802.org/3/ca/ .
  22. N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.
  23. F. J. Effenberger, H. Zeng, and X. Liu, “Burst-mode error distribution and mitigation in DSP-assisted high-speed PONs,” J. Lightwave Technol. 38, 754–760 (2020).
    [Crossref]
  24. F. Chang, ed., Datacenter Connectivity Technologies: Principles and Practice (River, 2018).
  25. J. Petrilla, P. Dawe, and G. LeCheminant, “New metric offers more accurate estimate of optical transmitter’s impact on multimode fiber-optic links,” in DesignCon, Santa Clara, California, January2015.
  26. I. B. Djordjevic, O. Milenkovic, and B. Vasic, “Generalized low-density parity-check codes for optical communication systems,” J. Lightwave Technol. 23, 1939–1946 (2005).
    [Crossref]
  27. I. B. Djordjevic and B. Vasic, “Iteratively decodable codes from orthogonal arrays for optical communication systems,” IEEE Commun. Lett. 9, 924–926 (2005).
    [Crossref]
  28. S. Sankaranarayanan, I. B. Djordjevic, and B. Vasic, “Iteratively decodable codes on m-flats for WDM high-speed long-haul transmission,” J. Lightwave Technol. 23, 3696–3701 (2005).
    [Crossref]
  29. I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett. 16, 338–340 (2004).
    [Crossref]
  30. F. Effenberger, “Enhanced FEC consideration for 100G EPON,” in IEEE P802.3ca 50G-EPON Task Force Meeting, March2016.
  31. M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

2020 (3)

2017 (1)

2005 (3)

2004 (1)

I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett. 16, 338–340 (2004).
[Crossref]

Dawe, P.

J. Petrilla, P. Dawe, and G. LeCheminant, “New metric offers more accurate estimate of optical transmitter’s impact on multimode fiber-optic links,” in DesignCon, Santa Clara, California, January2015.

Djordjevic, I. B.

I. B. Djordjevic, O. Milenkovic, and B. Vasic, “Generalized low-density parity-check codes for optical communication systems,” J. Lightwave Technol. 23, 1939–1946 (2005).
[Crossref]

I. B. Djordjevic and B. Vasic, “Iteratively decodable codes from orthogonal arrays for optical communication systems,” IEEE Commun. Lett. 9, 924–926 (2005).
[Crossref]

S. Sankaranarayanan, I. B. Djordjevic, and B. Vasic, “Iteratively decodable codes on m-flats for WDM high-speed long-haul transmission,” J. Lightwave Technol. 23, 3696–3701 (2005).
[Crossref]

I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett. 16, 338–340 (2004).
[Crossref]

M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

Effenberger, F.

F. Effenberger, “Enhanced FEC consideration for 100G EPON,” in IEEE P802.3ca 50G-EPON Task Force Meeting, March2016.

Effenberger, F. J.

Funada, T.

T. Funada and T. Kihara, “Consideration on US/DS WDM filter for ONU,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

Harstead, E.

Houtsma, V.

V. Houtsma, D. van Veen, and E. Harstead, “Recent progress on standardization of next-generation 25, 50, and 100G EPON,” J. Lightwave Technol. 35, 1228–1234 (2017).
[Crossref]

D. van Veen and V. Houtsma, “50 Gbps low complex burst mode coherent detection for time-division multiplexed passive optical networks,” in European Conference on Optical Communication (ECOC), September2018.

Houtsma, V. E.

Kihara, T.

T. Funada and T. Kihara, “Consideration on US/DS WDM filter for ONU,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.

LeCheminant, G.

J. Petrilla, P. Dawe, and G. LeCheminant, “New metric offers more accurate estimate of optical transmitter’s impact on multimode fiber-optic links,” in DesignCon, Santa Clara, California, January2015.

Li, L.

M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

Liao, Z.

Z. Liao and D. Liu, “NG-EPON diplexer filter analysis,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.

Liu, D.

Z. Liao and D. Liu, “NG-EPON diplexer filter analysis,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.

D. Liu and M. Tao, “50G single wavelength PON analysis and comparison,” in IEEE P802.3ca 50G-EPON Task Force Meeting, November2017.

Liu, X.

F. J. Effenberger, H. Zeng, and X. Liu, “Burst-mode error distribution and mitigation in DSP-assisted high-speed PONs,” J. Lightwave Technol. 38, 754–760 (2020).
[Crossref]

M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

Milenkovic, O.

Ogita, S.

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

Petrilla, J.

J. Petrilla, P. Dawe, and G. LeCheminant, “New metric offers more accurate estimate of optical transmitter’s impact on multimode fiber-optic links,” in DesignCon, Santa Clara, California, January2015.

Sankaranarayanan, S.

S. Sankaranarayanan, I. B. Djordjevic, and B. Vasic, “Iteratively decodable codes on m-flats for WDM high-speed long-haul transmission,” J. Lightwave Technol. 23, 3696–3701 (2005).
[Crossref]

I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett. 16, 338–340 (2004).
[Crossref]

Sugimoto, Y.

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

Tanaka, K.

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

Tanaka, N.

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

Tao, M.

D. Liu and M. Tao, “50G single wavelength PON analysis and comparison,” in IEEE P802.3ca 50G-EPON Task Force Meeting, November2017.

Umeda, D.

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

van Veen, D.

V. Houtsma, D. van Veen, and E. Harstead, “Recent progress on standardization of next-generation 25, 50, and 100G EPON,” J. Lightwave Technol. 35, 1228–1234 (2017).
[Crossref]

D. van Veen and V. Houtsma, “50 Gbps low complex burst mode coherent detection for time-division multiplexed passive optical networks,” in European Conference on Optical Communication (ECOC), September2018.

Van Veen, D. T.

Vasic, B.

S. Sankaranarayanan, I. B. Djordjevic, and B. Vasic, “Iteratively decodable codes on m-flats for WDM high-speed long-haul transmission,” J. Lightwave Technol. 23, 3696–3701 (2005).
[Crossref]

I. B. Djordjevic and B. Vasic, “Iteratively decodable codes from orthogonal arrays for optical communication systems,” IEEE Commun. Lett. 9, 924–926 (2005).
[Crossref]

I. B. Djordjevic, O. Milenkovic, and B. Vasic, “Generalized low-density parity-check codes for optical communication systems,” J. Lightwave Technol. 23, 1939–1946 (2005).
[Crossref]

I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett. 16, 338–340 (2004).
[Crossref]

Wey, J. S.

Yang, M.

M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

Zeng, H.

IEEE Commun. Lett. (1)

I. B. Djordjevic and B. Vasic, “Iteratively decodable codes from orthogonal arrays for optical communication systems,” IEEE Commun. Lett. 9, 924–926 (2005).
[Crossref]

IEEE Photon. Technol. Lett. (1)

I. B. Djordjevic, S. Sankaranarayanan, and B. Vasic, “Irregular low-density parity-check codes for long–haul optical communications,” IEEE Photon. Technol. Lett. 16, 338–340 (2004).
[Crossref]

J. Lightwave Technol. (6)

Other (23)

F. Chang, ed., Datacenter Connectivity Technologies: Principles and Practice (River, 2018).

J. Petrilla, P. Dawe, and G. LeCheminant, “New metric offers more accurate estimate of optical transmitter’s impact on multimode fiber-optic links,” in DesignCon, Santa Clara, California, January2015.

D. van Veen and V. Houtsma, “50 Gbps low complex burst mode coherent detection for time-division multiplexed passive optical networks,” in European Conference on Optical Communication (ECOC), September2018.

T. Funada and T. Kihara, “Consideration on US/DS WDM filter for ONU,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.

Z. Liao and D. Liu, “NG-EPON diplexer filter analysis,” in IEEE P802.3ca 50G-EPON Task Force Meeting, January2017.

IEEE P802.3ca 50G-EPON Task Force, “Physical layer specifications and management parameters for 25 Gb/s and 50 Gb/s passive optical networks,” http://www.ieee802.org/3/ca/ .

N. Tanaka, D. Umeda, Y. Sugimoto, T. Funada, K. Tanaka, and S. Ogita, “25.78-Gbit/s burst-mode receiver for 50G-EPON OLT,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper M1F.5.

“FSAN standards roadmap 2.0,” 2016, https://www.fsan.org/roadmap/ .

“PON transmission technologies above 10 Gb/s per wavelength,” ITU-T Recommendation G.Sup64, 2018, https://www.itu.int/rec/T-REC-G.Sup64/en .

“10-gigabit-capable passive optical networks (XG-PON): general requirements,” ITU-T Recommendation G.987.1, 2016, https://www.itu.int/rec/T-REC-G.987.1/en .

“10-gigabit-capable symmetric passive optical network (XGS-PON),” ITU-T Recommendation G.9807.1, 2017, https://www.itu.int/rec/T-REC-G.9807.1/en .

“40-gigabit-capable passive optical networks (NG-PON2): general requirements,” ITU-T Recommendation G.989.1, 2015, https://www.itu.int/rec/T-REC-G.989.1/en .

“Higher speed passive optical networks: requirements,” ITU-T Recommendation G.9804.1, 2019, https://www.itu.int/rec/T-REC-G.9804.1-201911-I/en .

“ONU management and control interface (OMCI) specification,” ITU-T Recommendation G.988, 2019, https://www.itu.int/rec/T-REC-G.988/en .

“Characteristics of a single-mode optical fibre and cable,” ITU-T Recommendation G.652, 2016, https://www.itu.int/rec/T-REC-G.652-201611-I/en .

“Characteristics of a bending-loss insensitive single-mode optical fibre and cable,” ITU-T Recommendation G.657, 2016, https://www.itu.int/rec/T-REC-G.657-201611-I/en .

“SDOs team up on PON convergence,” LightReading, 2017, https://www.lightreading.com/gigabit/fttx/sdos-team-up-on-pon-convergence/d/d-id/731234 .

“Gigabit-capable passive optical networks (G-PON): enhancement band,” ITU-T Recommendation G.984.5, 2018, https://www.itu.int/rec/T-REC-G.984.5/en .

“IEEE standard for Ethernet—Amendment 2: media access control parameters, physical layers, and management parameters for 25 Gb/s operation,” IEEE P802.3by, 2016, https://ieeexplore.ieee.org/document/7457590 .

“Common Public Radio Interface: eCPRI interface specification,” eCPRI specification V2.0, 2019, http://www.cpri.info/downloads/eCPRI_v_2.0_2019_05_10c.pdf .

D. Liu and M. Tao, “50G single wavelength PON analysis and comparison,” in IEEE P802.3ca 50G-EPON Task Force Meeting, November2017.

F. Effenberger, “Enhanced FEC consideration for 100G EPON,” in IEEE P802.3ca 50G-EPON Task Force Meeting, March2016.

M. Yang, L. Li, X. Liu, and I. B. Djordjevic, “FPGA-based real-time soft-decision LDPC performance verification for 50G-PON,” in Optical Fiber Communications Conference and Exhibition (OFC) (2019), paper W3H.2.

Cited By

OSA participates in Crossref's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (9)

Fig. 1.
Fig. 1. FSAN standards roadmap 2.0 [2].
Fig. 2.
Fig. 2. 50 Gbps/ch TWDM PON reference architecture. UNI = user network interface, SNI = service network interface, ${{R}}$ = receive and ${{S}}$ = send when defining bidirectional PON interfaces, CG = channel group, CP = channel pair, CT = channel termination, WM = wavelength multiplexer.
Fig. 3.
Fig. 3. 50G TDM PON reference architecture.
Fig. 4.
Fig. 4. (a) ITU-T PON wavelength plan and (b) coexistence and upgrade evolution diagram. GPON US: ${{N}}$ = narrow and ${{R}}$ = reduced; XGS-PON: only the basic wavelength set is shown; 50G TDM PON: ${{W}}$ = wide and ${{N}}$ = narrow.
Fig. 5.
Fig. 5. PON system coexistence and upgrade evolution paths.
Fig. 6.
Fig. 6. PON evolution and upgrade methods: (a) by external coexistence elements and (b) by MPM.
Fig. 7.
Fig. 7. Implementation options for 50 Gbps NRZ signal reception.
Fig. 8.
Fig. 8. Wavelength plan of 50G TDM PON.
Fig. 9.
Fig. 9. Simulated BER performance improvement from soft decision versus hard decision LDPC.

Tables (4)

Tables Icon

Table 1. General Technical Requirements for HSP Systems

Tables Icon

Table 2. Coexistence Methods and Scenarios for 50G TDM PONa

Tables Icon

Table 3. Performances of Different FEC Codes

Tables Icon

Table 4. Latency Requirement

Equations (1)

Equations on this page are rendered with MathJax. Learn more.

R × 125 × 125 L P S B d ,