NG-PON2 is the industry’s first multiple wavelength (per direction), standards-based passive optical network system that is compatible with power-split optical distribution networks. The physical media dependent layer recommendation (ITU-T G.989.2) is the result of over three years of collaborative work by members of the FSAN and ITU-T Study Group 15, Question 2 groups. This two-part paper provides the technical insight and rationales behind the recently approved standard. The first part of the paper focuses on optical link design topics, including the optical distribution network characteristics, wavelength plan, Raman fiber nonlinearity related degradation, and interchannel cross-talk tolerance. It also describes the wavelength-tuning capability of optical network units and its impact on the physical media dependent layer specification.
© 2015 Optical Society of America
Network operators continue to see exponential growth of traffic carried over their networks, driven by higher-bandwidth-consuming applications and services. At the network edge, the increasing bandwidth demand is met by pushing fiber deeper into the access network toward subscribers and ultimately terminating at the subscriber premises in the case of fiber-to-the-home/premises. Additionally, new applications demanding deeper fiber architectures are emerging such as mobile backhaul/front-haul and high-speed Ethernet-leased lines.
To meet the demands for higher bandwidths, network operators are seeking new fiber access technologies. To this end, network operator members of FSAN  worked together to define their requirements for a new generation of passive optical network (PON) system with higher capacity . In turn, this drove further studies within the full FSAN group to identify candidate technologies to meet these requirements.
The system identified in FSAN as best meeting their requirements, in the timeframe necessary to serve the expected demand, was primarily based on a hybrid time and wavelength division multiplexing (TWDM) method. The target capacity was at least 40 Gb/s with a per wavelength channel line rate up to 10 Gb/s. In addition to this, wavelength overlay channels were incorporated to enable virtual point-to-point connectivity (PtP WDM) over the same PON infrastructure as TWDM.
Once the ideas developed within FSAN were sufficiently mature, with a high level of industry consensus emerging, contributions were taken by FSAN member companies into the ITU-T to initiate a formal standards development project. NG-PON2 is the new standards-based PON system  arising from the project undertaken by Study Group 15, Question 2 of the ITU-T.
The early development of NG-PON2 standards has been reviewed previously in  along with a high-level overview of the underlying physical layer characteristics. The objective of this two-part paper is to give in-depth insight into the physical layer requirements embodied in G.989.2 and background into the reasoning behind some new parameters specified for the NG-PON2 optical interfaces.
II. NG-PON2 System Overview
NG-PON2, as specified in the G.989 series of ITU-T recommendations, is a successor standard to G-PON  and XG-PON1 , which embody the 1 and 10 Gb/s ITU-T standards-based PON technologies, respectively. The physical media dependent (PMD) layer standard, subject of this paper, is covered by the G.989.2 recommendation. The key feature that sets the NG-PON2 system apart from its predecessors is the specification of the industry’s first PON standard supporting multiple wavelengths per direction and compatibility with power-splitter-based optical distribution networks (ODNs). Many of the existing G-PON and XG-PON1 features, such as the optical path loss classes, support of RF video, and service requirements, are retained in the NG-PON2 system to ensure maximal reuse of existing technology, installed optical fiber infrastructure and coexistence with deployed legacy PON systems.
Two technologies are specified in NG-PON2 standards: a TWDM PON  and a wavelength overlay PtP WDM PON. Due to the fundamental NG-PON2 requirement for compatibility with power-splitter-based ODNs, these PON systems need wavelength-tunable optics in the transmitter (Tx) and receiver (Rx) of the optical network unit (ONU). NG-PON2 systems cannot depend on wavelength filtering in the ODN to work, even though operation over wavelength splitters is not precluded. This makes PtP WDM PON in NG-PON2 unique compared with previous WDM PON system concepts  and results in a synergy in tunable transceiver technology with TWDM PON that offers the prospect of lower optical component cost.
NG-PON2 systems support a minimum aggregate capacity of 40 Gb/s in the DS and 10 Gb/s in the US directions. From the per-wavelength channel perspective, TWDM offers three DS/US line rate combinations: the 10/2.5 Gb/s base case and optional symmetric rates of 10/10 and 2.5/2.5 Gb/s. Three line rate classes (around 1, 2.5, and 10 Gb/s) for PtP WDM are specified to transport Ethernet, SDH/OTN, and CPRI services. Each NG-PON2 system supports a minimum of 256 addressable ONUs per ODN. The optical parameters are specified assuming four and eight bi-directional wavelength channels for TWDM and PtP WDM, respectively. However, the specification anticipates a future increase in the number of wavelength channels for both technologies.
In the following sections, the notable physical layer aspects of the NG-PON2 system that distinguishes it from previous PON systems are discussed in more detail.
III. Optical Distribution Network
NG-PON2 systems have the capability to support power- and wavelength-split ODNs as well as hybrids of the two. It maintains the four optical path loss (OPL) classes previously specified for XG-PON1 (Table I), while not posing restrictions on number, geographical distribution (within the reach limits), or port count of the splitters.
The support of power-split-only ODN is a mandatory requirement for an NG-PON2 system to enable backward compatibility with the significant installed base of access fiber infrastructure. Although wavelength-split ODNs are permitted, ONUs that have no wavelength-selective capability and can, thus, only work within a fully wavelength-split ODN, are out of scope of NG-PON2 standards.
Two classes of PtP WDM PON ODN architectures are described in G.989.2 Annex A: 1) the wavelength-selected ODN (WS-ODN), which relies on tunable filters to provide a wavelength selection capability in the ONUs, and 2) the wavelength-routed ODN (WR-ODN), which has an intrinsic wavelength routing capability through wavelength splitters in the ODN.
The different types of ODNs and their design considerations are described in more detail as follows.
A. WS-ODN and WR-ODN
NG-PON2 systems support single-stage splitting as well as multi-stage splitting WS-ODNs with a requirement to support a maximum split ratio of at least . Lower split ratios may, of course, also be used. Thus, with the OPL classes listed in Table I, all legacy ODNs can be reused. For coexistence with other PON systems, e.g., G-PON and XG-PON1, a wavelength-band multiplexer called a coexistence element (CE) enables them to be combined on the same ODN. The respective insertion loss of the CE is lumped with the ODN and thus covered by the OPL class. Power splitting in practical devices leads to loss that scales approximately with 3.5 dB, where is the total split ratio and 0.5 dB excess loss is assumed per split. Therefore, maximum reach may be traded off against split ratio. A WS-ODN example is shown in the top part of Fig. 1.
NG-PON2 systems can also be based on purely WR-ODN, which only makes sense for PtP WDM deployments. WR-ODN enables longer reach or the use of transceivers with lower-budget classes. This is due to the lower insertion loss of the wavelength splitters, compared with power splitters.
NG-PON2 systems need a wavelength multiplexer (WM) to combine the multiple wavelength channels. The WM is considered part of the optical line terminal (OLT) equipment. The extra loss from WM is not part of the ODN loss, which is specified from the output of the WM (the S/R-CG reference point) to the input of the ONU (R/S reference point), as shown in Fig. 1. S/R-CG denotes the location where the OLT sends/receives a set of DS/US wavelength channel pairs, called channel group (CG), to/from ONUs. At the OLT, a logical channel termination (CT) function terminates each individual TWDM or PtP WDM channel.
The filters in the WM, and wavelength splitters for WR-ODNs, may be realized as arrayed waveguide gratings (AWG) or thin film filters. Wavelength splitters for 40 to 80 channels can have an insertion loss of 6 to 7 dB using AWGs. Single lumped wavelength splitters, such as cyclic AWGs, or cascaded filters can be used. A WR-ODN example with a single AWG is shown in the bottom part of Fig. 1.
A wavelength-agnostic fiber plant can simplify bandwidth and wavelength planning and provisioning. Full flexibility requires a WS-ODN. WR-ODNs can only support a certain degree of wavelength assignment flexibility when using cyclic AWGs with multiple input and output ports .
The total feeder-fiber reach capacity product of a WR-ODN can be higher than that of a WS-ODN. This is related to the lower splitting loss of WR-ODN. Hence, a WR-ODN may be better suited in cases where PtP WDM is used toward the aggregation part of a network, e.g., for combined backhaul and front-haul in fixed-mobile convergence scenarios and in green-field rural areas. WS-ODN is strictly mandatory in cases of legacy ODN support and typically preferred for residential access and in (dense) urban areas.
B. Hybrid ODN
An NG-PON2 ODN can be a hybrid of the WS and WR types. In comparison with WS-ODNs, such hybrid ODNs can provide higher split ratios for the same OPL class. For example, as shown in Fig. 2, to operate over a two-stage WS-ODN with a 1:256 split ratio, an NG-PON2 system requires a 26–28 dB loss budget. If the second-stage power splitter is replaced with an AWG, the resulting total loss is nominally 16–18 dB, thus allowing a reasonable power budget for system margin and fiber losses.
The hybrid approach can also be employed to separate segments of the ODN. For example, the first segment sends all signals to a power splitter, whereas the second segment routes certain wavelengths to specific end users via a wavelength splitter. In another example, the first segment groups wavelengths as wavelength sets by using a wavelength splitter, whereas the second segment sends all wavelengths in a single set to the ONUs via a power splitter. Numerous ODN combinations are possible to meet various operator and application needs.
C. Optical Path Loss Classes and Link Types
The four OPL classes specified for NG-PON2 systems are inherited from the XG-PON1 PMD recommendation, i.e., Nominal 1 (N1), Nominal 2 (N2), Extended 1 (E1), and Extended 2 (E2), as shown in Table I. These OPL classes enable reuse of an installed fiber infrastructure, based on WS-ODNs, for NG-PON2 systems. The Tx and Rx optical interface parameters for both TWDM and PtP WDM are specified to support these four OPL classes. The NG-PON2 N1 (14–29 dB) and E1 (18–33 dB) OPL classes are compatible with the G-PON B+ (13–28 dB) and C+ (17–32 dB) OPL classes, respectively, thus allowing coexistence of NG-PON2 and G-PON systems on the same ODN. The N1 class has similar optical interface power levels to those of the IEEE EPON PX30 and PR/PRX30 classes. The E1 class is specified with optical interface power levels similar to those of IEEE EPON PX40 and PR/PRX40 classes. Such commonality in optical specifications enables a higher volume for common optical submodules in two PON system types, which is expected to help minimize the costs.
As an implementation option, NG-PON2 includes two upstream TWDM link types, which differ in their optical interface parameters. The Type A link parameters are derived through an implicit assumption of not using an optical preamplifier at the OLT Rx. Conversely, the underlying assumption for a Type B link is for parameters commensurate with an optically preamplified OLT Rx. Therefore, the Type A link requires a more powerful ONU Tx, and the Type B link requires a higher OLT Rx sensitivity at the S/R-CG point. This allows more technology options for the NG-PON2 components. It should be noted that, although there are underlying assumptions about the OLT Rx in defining the interface parameters, the NG-PON2 PMD standard does not mandate any particular implementation.
IV. Wavelength Plan and Tuning Capability
The selection of the NG-PON2 wavelength plan was a result of a thorough evaluation of many options and a compromise between two seemingly divergent requirements. While it must allow for coexistence with legacy PON generations, it also has to be flexible enough to accommodate different deployment scenarios and future expansion.
The final agreed-upon wavelength plan is shown in Table II. TWDM DS spectrum is allocated to avoid interference with the RF video signals and XG-PON1, while allowing for ODN monitoring at 1650 nm. The choice of C-band for the TWDM US is to enable a low-cost ONU Tx through the reuse of existing optical technologies, which are presently shipping in large volumes.
The three TWDM US spectral options (wide band, reduced band, and narrow band) are determined by the wavelength control capabilities of the ONU Tx, with better control allowing the use of a narrower operating band. For PtP WDM, two spectrum options are provided: expanded spectrum to support greenfield deployment or flexibly reuse spectrum nominally allocated for other PON systems but unoccupied; and shared spectrum for full co-existence with legacy PON systems and TWDM. Unlike TWDM, PtP WDM allows DS and US wavelength channels to be in the same band to permit some implementation flexibility.
A key and distinctive feature of an NG-PON2 system is its multiwavelength capability; both TWDM and PtP WDM may use more than one wavelength channel to increase the total system capacity. ONUs with the capability of tuning to a specific channel, both in the DS (ONU Rx) and in the US (ONU Tx), are thus required to enable colorless operation. This wavelength tunability of the ONUs is a new feature not found in previous generations of PON standards. It offers several new capabilities (as discussed in the next subsection), but, at the same time, requires some important challenges to be solved. These challenges are discussed in Part 2 of this paper.
A. Tuning Time Classes
A specified characteristic of the tunable device (optical Rx or Tx) in an NG-PON2 ONU is its tuning time, which is defined as the elapsed time from the moment the tunable device leaves the source wavelength channel to the moment the tunable device reaches the target wavelength channel. Note that the tuning time thus defined is related only to the tunable optical device characteristics. It does not include contributions of other subsystems nor of protocols and mechanisms that control a complete tuning operation at the system level. G.989.2 Appendix VIII provides an example method to measure the tuning time, based on the observed power variation through a reference WM when an optical Tx tunes from one wavelength channel to another.
G.989.2 specifies three tuning time classes, as shown in Table III. Tunable devices supporting the different classes may be based on a variety of technologies having potentially dissimilar costs in order to allow for a range of capabilities of the NG-PON2 system.
The slowest (Class 3) tunable devices could be based on thermal effects to change their operating wavelength and are suited to applications in which tuning operations are infrequent or when a short service interruption is tolerable. Channel-based protection mechanisms can be adopted, but a typical sub-50 ms protection objective cannot be achieved. Semi-static load-sharing and power-saving mechanisms based on channel changes can also be of interest with Class 3 devices.
Class 2 tunable devices allow faster channel tuning such that sub-50 ms protection becomes possible. They also enable dynamic load-sharing and dynamic power-saving features.
Class 1 tunable devices, characterized by the shortest tuning time, may enable a future dynamic wavelength and bandwidth allocation feature in the system. The OLT also could control dynamically, in addition to the transmission time and duration, the transmission wavelengths of ONUs to allow wavelength hopping between the transmission periods.
V. Optical Path Penalty
An essential aspect of optical system design is accounting for optical path penalties (OPP). For NG-PON2, in addition to the usual chromatic-dispersion-related penalties, the relatively high optical power and multiwavelength nature bring another degradation mechanism to bear, namely, Raman nonlinearity. This can result in nonlinear cross-talk and signal depletion for certain wavelengths. The following will describe, by way of example, the 2.5 Gb/s US analysis for TWDM to illustrate the process of determining the total permitted OPP.
A. Chromatic Dispersion
The chromatic-dispersion-related portion of OPP draws from existing standards, along with some underlying assumptions regarding Tx type. For instance, at 2.5 Gb/s, the use of a directly modulated DFB laser (DML) is assumed with its associated spectral characteristics and dispersion-related penalties. At 10 Gb/s, both DML and externally modulated laser (EML) Tx are considered. Note that G.989.2 assumes the use of EMLs in deriving the OPP values. When using DML at 10 Gb/s in the C/L-band, and just 20 km of fiber, some form of dispersion compensation is necessary to achieve the OPP values in G.989.2, e.g., dispersion compensating fiber at the ONU or burst-mode-capable electronic dispersion compensation at the OLT. The OPP values assumed for the chromatic-dispersion-only aspect are shown in rows 1 and 2 of Table IV. The values for the Raman-induced OPP are derived as follows.
B. Raman Nonlinearity
The impacts of Raman nonlinearity in fiber in an NG-PON2 system appear in two distinct ways. First, US TWDM channels are depleted by the counterpropagating DS TWDM channels and, second, through modulation cross-talk between any co-propagating optical signals separated in optical frequency by 1–40 THz. In the case of power depletion, this is accounted for in the NG-PON2 PMD recommendation by an increase in OPP.
The mean channel launch power maximum for 10 Gb/s DS TWDM in each OPL class of the NG-PON2 PMD recommendation is shown in Table V along with the maximum total launch power into the ODN.
With a worst-case wavelength separation of () between DS and US TWDM channels, the Raman gain coefficient is still significant at about 60% of the peak gain. Even with the effective reduction in net Raman interaction through polarization averaging effects in counterpropagation, the impact is still large enough to cause significant power depletion. Figure 3 shows example simulation results for Raman depletion provided during the development of the G.989.2 standard. Such results can be obtained, for example, by numerically solving the coupled nonlinear Raman equations . Figure 3 shows the worst-case Raman depletion of US TWDM channels for all OPL classes, and both four and eight channel cases as a function of fiber distance. The solid curves represent the eight-channel case, whereas the dotted curves represent the four-channel case (with 3 dB less total power than for eight channels).
C. Combined OPP
Taking power depletion values from Fig. 3, and adding the chromatic dispersion penalty for 20 or 40 km fiber length, respectively, with the associated number of TWDM channels, the combined optical path penalty is calculated. For example, in the case of the N1 OPL class at 20 km, 0.5 dB is the OPP value considering dispersion-only penalty. Figure 3 shows that the apparent additional penalty from Raman depletion is 0.25 dB for the four-channel case, which gives a combined OPP of 0.75 dB. The NG-PON2 standard rounded to the nearest 0.5 dB for the OPP, which leads to 1.0 dB as the final value in G.989.2. The methodology described above can also be applied to the 10 Gb/s US for TWDM and PtP WDM channels to define permitted OPP values.
Note that the line code for TWDM 10 Gb/s, 40 km US is left for further study in G.989.2, which is mainly due to the OPP consideration. Three implementation options are currently under discussion: (1) EML with NRZ; (2) directly modulated DFB with NRZ, which needs burst mode electronic dispersion compensation at the OLT; and (3) directly modulated DFB with dispersion-friendly line coding, which is not yet fully explored. Each of these implementations has a different degree of impact on the OPP. The specific line code and resulting OPP values can only be specified after the implementation assumptions are agreed upon.
VI. Cross-Talk Analysis
The inherent multiwavelength characteristics of NG-PON2 systems in both US and DS have implications for optical link design, as cross-talk impairments must be accounted for and constrained through appropriate specification in the NG-PON2 PMD standard.
The optical cross-talk terms that are the subject of the NG-PON2 PMD relate to the in-band interferometric cross-talk resulting from optical power straying into the spectral region of other channels, as shown in Fig. 4. Impairments resulting from power addition of signal and cross-talk terms due to imperfect demultiplexers at OLTs and imperfect tunable filters at ONUs are not the subject of the NG-PON2 PMD standard.
In general, the most stringent requirements on Tx spectral characteristics arise in the US. This results from the wavelength transparency of the WS-ODN and from the large difference in power levels at the OLT Rx for signals from different ONUs. To define the spectral properties of PON Tx, the worst-case operational conditions permitted by the standard must be considered to ensure proper operation of the PON system, as illustrated in Fig. 5. These conditions occur where an ONU impacted by cross-talk is situated at the maximum ODN loss and is emitting at the low end of the Tx power range. The interfering signals arise from ONUs, which are situated at the minimum ODN loss. It can be shown that, for ODN split ratios of (where M is an integer), a split ratio is the maximum for which a 15 dB differential loss is possible within the E2 OPL class. This sets a limit for the maximum number of potential interferers that need to be considered.
Two cross-talk terms are important to characterize an enabled channel: the interference caused by similar signals, i.e., TWDM-to-TWDM and PtP WDM-to-PtP WDM, and the interference caused by dissimilar signals, i.e., TWDM-to-PtP WDM and PtP WDM-to-TWDM cross-talk. NG-PON2 recommendations use two terms to describe these distinct cases: (1) out-of-channel power spectral density and (2) out-of-band power spectral density. These two terms are described in detail as follows. In addition, the permissible interference caused by the off-state power of a Tx when it is not enabled is also explained.
Given the penalty attributed to each interference term, it is possible to derive the relative cross-talk using Eq. (1)  for an optically preamplified Rx and an average power decision threshold setting:2):
The BER of the victim channel is given by the reference BER value for the link, as defined in G.989.2, and largely dictated by whether or not FEC is used.
To account for the reduction in extinction ratio after transmission, is replaced with an effective extinction ratio given by Eq. (3):
In the US direction, a 1 dB relaxation of the permitted relative cross-talk is included to account for the randomized polarization states of the ONU Tx outputs and, consequently, reduced coherent beat noise.
A. Out-of-Channel Optical Power Spectral Density
The out-of-channel optical power spectral density (OOC-PSD) defines the maximum power spectral density an NG-PON2 Tx is permitted to emit outside the spectral interval corresponding to its current operating wavelength channel. This PSD is specified at the appropriate reference point (S/R-CG for DS direction, R/S for US direction).
In the US direction, a penalty of 1 dB due to the OOC interference has been accounted for in the NG-PON2 PMD standard. In the DS direction, this has been tightened to 0.1 dB, as a low OOC-PSD can be achieved at the S/R-CG reference point by virtue of the bandpass filter function of the WM.
For the calculation of the OOC-PSD, the worst-case ODN is assumed as previously described and is given by Eq. (4):
By way of example, and to illustrate the challenging requirements for this parameter, we shall derive the maximum OOC-PSD specification for an ONU transmitter in four-channel 10 Gb/s TWDM.
The minimum launch power of a 10 Gb/s ONU is with a minimum of 6 dB extinction ratio and a pre-FEC BER at the OLT Rx of [, from Eq. (2)]. Using the allowed OPP of 2 dB and Eq. (3), the effective extinction ratio is therefore 3.45 dB. Equation (1) gives the allowed relative cross-talk for a 1 dB penalty on a 10 Gb/s US TWDM channel as . With the ODN differential loss of 15 dB and three interfering TWDM channels, we obtain a (in 15 GHz) OOC-PSD requirement (including a 1 dB polarization averaging relaxation). With a maximum permitted ONU output power of , this results in a transmitter worst-case optical signal-to-noise-ratio (OSNR) requirement of 54.7 dB (in 15 GHz). With typical fiber-access transmitters, based on DFB lasers, able to provide around 45 dB side-mode suppression ratio (SMSR), the OOC-PSD requirement presents a significant challenge for a low-cost optical component design.
B. Out-of-Band Optical Power Spectral Density
The out-of-band optical power spectral density (OOB-PSD) defines the maximum power spectral density an NG-PON2 Tx is permitted to emit outside the specified operating wavelength band. This PSD is also specified at the appropriate reference point, i.e., S/R-CG for DS direction, R/S for US direction.
In both transmission directions, a penalty of 0.1 dB due to the OOB interference has been accounted for in the NG-PON2 standard. A low penalty is expected to be readily achievable, as an optical bandpass filter function can be imposed on any Tx before launching into the ODN. For example, a TWDM ONU transceiver may implement a bandpass filter as part of the diplexer and so suppress optical emission outside the TWDM US wavelength bands.
For the calculation of the OOB-PSD, the worst-case ODN is again assumed and is given by Eq. (5):
Here, is the number of interfering channels and equivalent to all the wavelength channels in the relevant direction and operating wavelength band.
In some instances, the OOB-PSD specification for a Tx is determined by the impact of this OOB emission, as reflected by the ODN, and impacting channels propagating in the opposite direction. In particular, this determines the specification for the DS OOB-PSD for both TWDM and PtP WDM OLT Tx, respectively.
C. Optical Power Spectral Density When Not Enabled
The optical power spectral density when not enabled (WNE-PSD) defines the maximum power spectral density an NG-PON2 Tx is permitted to emit, at any wavelength inside or outside the operating wavelength band, when the Tx is not enabled, i.e., nominally in an off-state. This PSD is also specified at the same reference points as for the OOC/OOB-PSDs. A similar formula, Eq. (6), is used to derive the WNE-PSD as for OOC and OOB:
In the case of TWDM in the US direction, the maximum number of ONUs () in the not-enabled state and having an impact on a victim ONU is 63.
To define the WNE-PSD in the DS direction, the worst-case ODN, as shown in Fig. 6, is assumed. In Fig. 6, an ODN is shown with two OLTs connected and offering protection in the event of a primary OLT or feeder fiber failure. The same wavelength channel(s) may be provisioned for the primary and secondary links, and, in this instance, we should be concerned about the power within this wavelength channel from the not-enabled DS Tx relative to that in the enabled Tx on the other OLT.
D. In-Band Cross-Talk Tolerance
In order to define a qualification test criteria for an NG-PON2 Rx, the ratio of the sum of all in-band cross-talk terms, as described above, to that of the received signal is used. The term in-band cross-talk tolerance is used to express the minimum value of this ratio that must be tolerated while maintaining compliance with the specified Rx sensitivity. Furthermore, we assume the cross-talk is polarized and aligned with the signal polarization. It is envisaged that a simple Rx qualification test may consist of a broadband light source coupled with the transmitted signal to set the signal to cross-talk ratio accordingly.
By ensuring the Rx sensitivity must be met in the presence of an effective OSNR degradation, the penalty due to in-band cross-talk terms therefore does not need to be accounted for as part of the OPP.
E. Mask for Out-of-Channel Power Spectral Density
The out-of-channel power spectral density (OOC-PSD) mask, in relation to the channel spacing (CS), is shown in Fig. 7. The purpose of the mask is to define the transmitted power allowed outside of the transmitter’s maximum spectral excursion, which is explained in more detail in Part 2 of this paper. In some sense, this is analogous to the term SMSR often used in other applications. The reason for the change in terminology is that the OOC-PSD is not a ratio but the absolute power level allowed in a given spectral interval. The level of impact of this OOC power on other channels is dependent on whether the power interacts interferometrically or as power addition. Interferometric interference is more damaging and is the reason for the two levels in the OOC-PSD mask, OOC1 and OOC2.
It is possible to design lasers so that the largest side modes fall into the more harmless OOC1 regions. Hence, the OOC-PSD mask is a guide to laser manufacturers on where to place the largest side modes for minimal impact. Compliance with this mask is required under all conditions, during ONU initialization, registration, and tuning.
The ITU-T G.989 series of recommendations specifies NG-PON2 as the industry’s first PON system with multiple wavelengths per direction (downstream and upstream) and compatibility with a power-splitter-based ODN. An NG-PON2 system is composed of a set of TWDM channels and/or a set of PtP WDM channels, both of which require wavelength tunable Tx and Rx at the ONU.
In Part 1 of this paper, we have focused on topics related to optical link design and by what rationale the physical layer requirements were developed.
NG-PON2 standards accommodate operation over legacy power-splitter-based ODNs with optical path loss specifications aligned with XG-PON1. The wavelength plan accommodates requirements for co-existence with G-PON, XG-PON1, and RF-video. Flexibility in the wavelength plan enables any unoccupied spectrum to be used for capacity expansion depending on the specific operator use case.
NG-PON2 technology supports both wavelength-selected and wavelength-routed ODNs, each intended for different scenarios. WS-ODN can be supported by TWDM and PtP WDM while WR-ODN is generally considered for PtP WDM deployments only. Hybrid ODNs combining WS and WR are also possible for both TWDM and PtP WDM.
The tunable feature of the ONU imposes new challenges for the transceiver design. The NG-PON2 PMD recommendation specifies three tuning time classes that facilitate various new system functions and network use cases.
Due to the relatively high optical power and multiwavelength nature, Raman nonlinearity needs to be accounted for in the optical path penalty. The resulting upstream channel power depletion, as well as the methodology used to derive the standardized OPP values, was reviewed.
Finally, interferometric cross-talk, from optical power straying into the spectral region of other channels or bands, has a major impact on the optical link design. The standard specifies the maximum power spectral densities an NG-PON2 Tx is permitted to emit to ensure proper system operation. Definitions and derivations used to specify the standardized parameter values were explained in detail.
The authors would like to thank members in the FSAN and ITU-T SG15/Q2 groups for the invaluable discussions and collaboration during the NG-PON2 standards development process.
1. Full Service Access Network (FSAN) [Online]. Available: http://www.fsan.org.
2. P. Chanclou, A. Cui, F. Geilhardt, H. Nakamura, and D. Nesset, “Network operator requirements for the next generation of optical access networks (Invited Paper),” IEEE Netw. Mag., vol. 26, no. 2, pp. 8–14, 2012.
3. “40-gigabit-capable passive optical networks (NG-PON2),” ITU-T G.989 Series Recommendations: G.989.1: General Requirements (2013); G.989.2: Physical Media Dependent Layer Specification (2014); G.989.3: Transmission Convergence Layer Specification (approved).
4. D. Nesset, “NG-PON2 technology and standards,” J. Lightwave Technol., vol. 33, no. 5, pp. 1136–1143, 2015. [CrossRef]
5. “Gigabit-capable passive optical networks (GPON),” ITU-T Series Recommendation G.984, 2008.
6. “10-gigabit-capable passive optical networks (XG-PON),” ITU-T Series Recommendation G.987, 2012.
7. Y. Luo, X. Zhou, F. Effenberger, X. Yan, G. Peng, Y. Qian, and Y. Ma, “Time and wavelength-division multiplexed passive optical network (TWDM-PON) for next-generation PON stage 2 (NG-PON2),” J. Lightwave Technol., vol. 31, no. 4, pp. 587–593, 2013. [CrossRef]
8. “Multichannel seeded DWDM applications with single-channel optical interfaces,” ITU-T Recommendation G.698.3, 2012.
9. H. Woesner and M. Maier, “Routing and resilience in AWG based WDM networks,” in TransiNet Workshop, HHI, 2002.
10. C. Headley and G. Agrawal, Eds., Raman Amplification in Fiber Optical Communication Systems. Academic, 2004.
11. F. Liu, C. J. Rasmussen, and R. J. S. Pedersen, “Experimental verification of a new model describing the influence of incomplete signal extinction ratio on the sensitivity degradation due to multiple interferometric crosstalk,” IEEE Photon. Technol. Lett., vol. 11, no. 1, pp. 137–139, 1999. [CrossRef]
Jun Shan Wey (M’95–SM’11) received a Ph.D. in electrical engineering from University of Maryland, College Park, in 1995. She began her career in the research faculty at the Lab for Physical Sciences, University of Maryland, focusing on high-speed optical communication and signal processing. She joined Siemens in 2004 to work on IPTV product development. Later, with Nokia Siemens Networks and, subsequently, Coriant, she became active in the standardization for IPTV and next-generation optical access networks. She participated in both FSAN and ITU-T on developing the NG-PON2 standard and was a co-editor of the G.989.2 recommendation. In 2014, she founded Lightnotes Consulting to advise clients on competitive and go-to-market strategies for emerging technologies. Dr. Wey co-chaired the IPTV Interoperability Task Force of the ATIS IIF and MSF groups. She was a co-editor in the Verizon VICAS Forum. She currently serves on the WDM-PON forum board and the OFC technical program N4 subcommittee.
Derek Nesset (M’01–SM’13) received a B.Sc. degree in physics from Birmingham University, UK, in 1989 and an M.Sc. 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 Gbit/s, including the first field demonstration of 40 Gbit/s transmission over BT’s fiber infrastructure. In 2000, Derek 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 on enhanced PON systems for fiber access. Most recently, he has focused on next-generation PON technologies and standards and actively contributes to progressing NG-PON2 in both FSAN and ITU-T. He leads research into future optical access networks within BT’s Research & Technology organization near Ipswich, UK. He also chairs the NG-PON task group in FSAN. Mr. Nesset was awarded an honorary professorship at Bangor University in 2014. He has contributed over 90 journal and conference publications and has eight patents. He has participated in OFC, OECC, and ACP conference subcommittees and chaired the optical access subcommittee for OFC 2014.
Maurizio Valvo received his M.Sc. degree in electronics engineering cum laude from the University of Naples, Italy, in 1991. In the same year, he joined CSELT, the Center for Research in Telecommunications, now Telecom Italia Lab–Turin, where he is currently active. He has focused since the beginning of his career on broadband access networks and in particular on PON systems, which he also contributed to specify, develop, and test during the 1990s in the framework of several European-funded projects. He has led research projects with the objective to specify and test, both in the laboratory and in the field, innovative access network technologies based on PON, xDSL, HFC, WiMAX, and free-space optics. Currently, Mr. Valvo leads the laboratory for fixed access network innovation and a research project for the innovation of the Telecom Italia optical access network. He holds four patents and is co-author of three books and several papers.
Klaus Grobe (M’94–SM’13) received Dipl.-Ing. and Dr.-Ing. degrees in electrical engineering from Leibniz University, Hannover, Germany, in 1990 and 1998, respectively. He has worked over 20 years in the fields of lightwave guides and WDM. This includes positions as a technical staff member at German and pan-European network operators. In 2000, he joined ADVA Optical Networking SE, where he now works as a senior principal engineer in the CTO office in Munich, Germany. He is one of the main authors of Wavelength Division Multiplexing—A Practical Engineering Guide (Hoboken, NJ, Wiley, 2014) and has authored and co-authored more than 100 technical publications as well as three further book chapters on WDM and PON technologies. His research interests include next-generation broadband access networks as well as high-speed WDM transport. He holds 25 (pending) patents. Dr. Grobe is a member of the German VDE/ITG and ITG Study Group 5.3.3 on photonic networks. He served on the OFC subcommittee 10 in 2009–2012 and works in FSAN and ITU T–SG15-Q.2.
Hal Roberts (M’12) received a B.A. with a major in physics from Colgate University, Hamilton, NY, USA, in 1976 and an M.S. from Thayer Engineering School of Dartmouth College, Hanover, NH, USA, in 1979. He joined ADC Telecommunications in 1985, first developing optical components and then a PON-based access system for last-mile telephony transport in the late 1980s/early 1990s called “Homeworx.” Subsequently, Hal was responsible for leading a team that developed a new RF access technology for the Homeworx system to allow operation on a hybrid fiber coax system. This resulted in the first OFDMA-based access solution. Mr. Roberts is listed as a key contributor to the DOCSIS 2.0 and 3.0 DRFI and went on to the development of digital DAS systems for mobile front-haul. He left ADC and joined Calix in 2009 and since then has worked on PON and wireless technologies and participated in FSAN and ITU-T [T15/Q2]. Mr. Roberts has worked on progressing NG-PON2 with over 30 contributions related to the PMD and was editor of G. supplement 51 on PON protection. He has over 40 patents in wireless, optical, and coax access technologies.
Yuanqiu Luo (S’02–M’06–SM’11) received a Ph.D. in electrical engineering from the New Jersey Institute of Technology, Newark, NJ, USA, in 2006. She is currently a senior staff engineer in the Optical Access Network Department of Huawei Technologies, Bridgewater, NJ, USA. Before joining Huawei, she was with NEC Laboratories America, Princeton, NJ, USA. She has been heavily involved in the R&D effort of optical access networks, such as the first XG-PON1 field trial, PON as wireless backhaul, XG-PON1 and NG-PON2 protocol design, and the first NG-PON2 prototype system. Dr. Luo is a co-editor of ITU-T Recommendations G.987.3, G.989.2, and G.9802. She is a clause editor of IEEE Standard 802.1AS. Dr. Luo is an editor of IEEE Communications Surveys and Tutorials and an editor of Springer Photonic Network Communications. She has authored more than 60 publications in journals and conferences. She serves on OFC, HPSR, AccessNets, and other conference subcommittees. She is a co-author of the book Passive Optical Networks: Flattening the Last Mile Access (IEEE Comsoc Pocket Guides to Communications Technologies). She holds 19 US patents. Dr. Luo was honored with an IEEE Standards Award in 2011.
Joe Smith has worked in the fiber-optic industry for nearly four decades. He joined ITT in 1976 and worked on initial fiber-optic concepts, manufacturing techniques, and validation of new fiber optic technologies. Subsequently, he worked on the development of the world’s first fiber optic system at 1.5 Mb/s, followed by systems for 45 Mb/s through 2.5 Gb/s in transport for ITT, Alcatel, and Alcatel-Lucent. Mr. Smith has participated in standards in various forums for 25 years in transport arenas for ANSI, Bellcore, FSAN, and ITU, where he currently holds co-editorship responsibility for NG-PON2 PMD Layer, G.989.2.