Passive Optical Networks YH Previous A passive optical network (PON) is a point-to-multipoint, fiber to the premises network architecture in which unpowered optical splitters are used to enable a single optical fiber to serve multiple premises, typically 32-128. A PON consists of an Optical Line Termination (OLT) at the service provider's central office and a number of Optical Network Units (ONUs) near end users. A PON configuration reduces the amount of fiber and central office equipment required compared with point to point architectures. Downstream signals are broadcast to each premises sharing a fiber. Encryption is used to prevent eavesdropping. Upstream signals are combined using a multiple access protocol, invariably time division multiple access (TDMA). The OLTs "range" the ONUs in order to provide time slot assignments for upstream communication. A PON is a shared network, in that the OLT sends a single stream of downstream traffic that is seen by all ONTs. Each ONT only reads the content of those packets that are addressed to it. Optical Line Terminal (OLT) The OLT provides the interface between the PON and the backbone network. These typically include: Internet Protocol (IP) traffic over Gigabit, 10G, or 100 Mbit/s Ethernet standard time division multiplexed (TDM) interfaces such as SONET/SDH or PDH at various rates ATM UNI at 155-622 Mbit/s The OLT is connected to the switched network via standardized interfaces. On the distribution side, it presents optical access interfaces according to this and other G-PON standards, in terms of bit rate, power budget, jitter, etc. The OLT consists of three major parts: service port Interface Function; cross-connect function; Optical Distribution Network (ODN) interface. OLT functional block diagram 1) PON core shell This block consists of two parts, the ODN interface function and the PON TC function. PON TC function includes framing, media access control, OAM, DBA, and delineation of Protocol Data Unit (PDU) for the cross-connect function, and ONU management. Each PON TC selects one mode of ATM, GEM and Dual. 2) Cross-connect shell The Cross-connect shell provides a communication path between the PON core shell and the Service shell. Technologies for connecting this path depends on services, internal architecture in OLT and other factors. OLT provides cross-connect functionality according to selected modes, such as GEM, ATM or Dual. 3) Service shell This shell provides translation between service interfaces and TC frame interface of the PON section. Optical Network Terminals (ONT) The ONT terminates the PON and presents the native service interfaces to the user. These services can include voice (plain old telephone service (POTS) or voice over IP (VoIP)), data (typically Ethernet or V.35), video, and/or telemetry (TTL, ECL, RS530, etc.). Often, the ONT functions are separated into two parts: the ONU, which terminates the PON and presents a converged interface – such as xDSL or multiservice Ethernet – toward the user, and network termination equipment (NTE), which provides the separate, native service interfaces directly to the user The functional building blocks of the G-PON ONT are mostly similar to the functional building blocks of the OLT. Since the ONT operates with only a single PON Interface (or maximum 2 interfaces for protection purposes), the cross-connect function can be omitted. However, instead of this function, service MUX and DMUX function is specified to handle traffic. Each PON TC selects one mode of ATM, GEM and Dual. ONT functional block diagram Upstream bandwidth allocation The OLT is responsible for allocating upstream bandwidth to the ONTs. Because the optical distribution network (ODN) is shared, ONT upstream transmissions could collide if they were transmitted at random times. ONTs can lie at varying distances from the OLT, meaning that the transmission delay from each ONT is unique. The OLT measures delay and sets a register in each ONT via PLOAM (physical layer operations and maintenance) messages to equalize its delay with respect to all of the other ONTs on the PON. Once the delay of all ONTs has been set, the OLT transmits so-called grants to the individual ONTs. A grant is permission to use a defined interval of time for upstream transmission. The grant map is dynamically re-calculated every few milliseconds. The map allocates bandwidth to all ONTs, such that each ONT receives timely bandwidth for its service needs. Some services – POTS, for example – require essentially constant upstream bandwidth, and the OLT may provide a fixed bandwidth allocation to each such service that has been provisioned. DS1 and some classes of data service may also require constant upstream bit rate. But much data traffic – internet surfing, for example – is burst and highly variable. Through dynamic bandwidth allocation (DBA), a PON can be oversubscribed for upstream traffic, according to the traffic engineering concepts of statistical multiplexing. (Downstream traffic can also be oversubscribed, in the same way that any LAN can be oversubscribed. The only special feature in the PON architecture for downstream oversubscription is the fact that the ONT must be able to accept completely arbitrary downstream time slots, both in time and in size.) There are basically two forms of DBA, status-reporting (SR) and non-status reporting (NSR). In NSR DBA, the OLT continuously allocates a small amount of extra bandwidth to each ONT. If the ONT has no traffic to send, it transmits idle frames during its excess allocation. If the OLT observes that a given ONT is not sending idle frames, it increases the bandwidth allocation to that ONT. Once the ONT's burst has been transferred, the OLT observes a large number of idle frames from the given ONT, and reduces its allocation accordingly. NSR DBA has the advantage that it imposes no requirements on the ONT, and the disadvantage that there is no way for the OLT to know how best to assign bandwidth across several ONTs that need more. In SR DBA, the OLT polls ONTs for their backlogs. A given ONT may have several so-called traffic containers (T-CONTs), each with its own priority or traffic class. The ONT reports each T-CONT separately to the OLT. The report message contains a logarithmic measure of the backlog in the T-CONT queue. By knowledge of the service level agreement for each T-CONT across the entire PON, as well as the size of each T-CONT's backlog, the OLT can optimize allocation of the spare bandwidth on the PON. PON classification and Standards ITU-T G.983 APON (ATM Passive Optical Network). This was the first Passive optical network standard. It was used primarily for business applications, and was based on ATM. BPON (Broadband PON) is a standard based on APON. It adds support for WDM, dynamic and higher upstream bandwidth allocation, and survivability. It also created a standard management interface, called OMCI, between the OLT and ONU/ONT, enabling mixed-vendor networks. ITU-T G.984 GPON (Gigabit PON) is an evolution of the BPON standard. It supports higher rates, enhanced security, and choice of Layer 2 protocol (ATM, GEM, Ethernet). In early 2008, Verizon is in the process of installing millions of lines, while British Telecom and AT&T are in advanced trials. IEEE 802.3ah EPON or GEPON (Ethernet PON) is an IEEE/EFM standard for using Ethernet for packet data. IEEE 802.3av 10GEPON (10 Gigabit Ethernet PON) is an IEEE Task Force for 10Gbit/s backwards compatible with 802.3ah EPON. 10GigEPON will likely multi-lamda downstream and continue to use a single lamda with ATDMA for upstream. It will also be WDM-PON compatible. History of PON Early work on efficient fiber to the home architectures was done in the 1990s by the Full Service Access Network (FSAN) working group, formed by major telecommunications service providers and system vendors. The International Telecommunications Union (ITU) did further work, and has since standardized on two generations of PON. The older ITU-T G.983 standard is based on asynchronous transfer mode (ATM), and has therefore been referred to as APON (ATM PON). Further improvements to the original APON standard – as well as the gradual falling out of favor of ATM as a protocol – led to the full, final version of ITU-T G.983 being referred to more often as broadband PON, or BPON. A typical APON/BPON provides 622 megabits per second (Mbit/s) of downstream bandwidth and 155 Mbit/s of upstream traffic, although the standard accommodates higher rates. The ITU-T G.984 (GPON) standard represents a boost in both the total bandwidth and bandwidth efficiency through the use of larger, variable-length packets. Again, the standards permit several choices of bit rate, but the industry has converged on 2.488 gigabits per second (Gbit/s) of downstream bandwidth, and 1.244 Gbit/s of upstream bandwidth. GPON Encapsulation Method (GEM) allows very efficient packaging of user traffic, with frame segmentation to allow for higher Quality of Service (QoS) for delay-sensitive traffic such as voice and video communications. The IEEE 802.3 Ethernet PON (EPON or GEPON) standard was completed in 2004 (http://www.ieee802.org/3/), as part of the Ethernet First Mile project. EPON uses standard 802.3 Ethernet frames with symmetric 1 gigabit per second upstream and downstream rates. EPON is applicable for data-centric networks, as well as full-service voice, data and video networks. Recently, starting in early 2006, work began on a very high-speed 10 Gbit/s EPON (XEPON or 10-GEPON) standard (http://www.ieee802.org/3/av/). A PON takes advantage of wavelength division multiplexing (WDM), using one wavelength for downstream traffic and another for upstream traffic on a single Nonzero dispersion shifted fiber (ITU-T G.652). The specification calls for downstream traffic to be transmitted on the 1490 nanometer (nm) wavelength and upstream traffic to be transmitted at 1310 nm. The 1550 nm band is allocated for optional overlay services, typically RF (analog) video. As with bit rate, the standards describe several optical budgets, most common is 28 dB of loss budget for both BPON and GPONm but products have been announced using less expensive optics as well. 28 dB corresponds to about 20 km with a 32-way split. Forward error correction (FEC) may provide another 2-3 dB of loss budget on GPON systems. As optics improve, the 28 dB budget will likely increase. The GPON address space accommodates service to up to 128 ONUs on a single fiber. A PON consists of a central office node, called an optical line terminal (OLT), one or more user nodes, called optical network units (ONUs) or optical network terminals (ONTs), and the fibers and splitters between them, called the optical distribution network (ODN). In apartment buildings, the ONT often separately connects via VDSL or Ethernet to the apartments, at speeds up to 100 megabits. An ONT is a single integrated electronics unit that terminates the PON and presents native service interfaces to the user. An ONU is the PON-side half of the ONT, terminating the PON, and may present one or more converged interfaces, such as xDSL or Ethernet, toward the user. An ONU typically requires a separate subscriber unit to provide native user services such as telephony, Ethernet data, or video. In practice, the difference between an ONT and ONU is frequently ignored, and either term is used generically to refer to both classes of equipment. Current Status of PON TDM-PON Both APON/BPON and EPON/GEPON have been deployed widely, but most networks designed in 2008 use GPON or GEPON. GPON is most common in the West, GEPON more common in Asia. In china, Carriers looking to deliver the latest broadband services over fiber access have a choice of PON technologies. Gigabit PON offers greater bandwidth and functionality, but Ethernet PON benefits from maturity. The currently deployed PON (e.g., APON, GPON) is named as TDM-PON, in which the bandwidth is shared among the users in the manner of time domain multiplexing. This is different from the multiplexing manner discussed below, e.g., WDM-PON. WDM-PON Wavelength Division Multiplexing PON, or WDM-PON, is a type of passive optical networking, being pioneered by several companies, that uses multiple optical wavelengths to increase the upstream and/or downstream bandwidth available to end users. This technology looks forward to a day when optical technology is cheaper and easier to deploy, and end users demand higher bandwidth. WDM-PON can provide more bandwidth over longer distances by devoting more raw optical bandwidth to each user, and by increasing the link loss budget of each wavelength, making it less sensitive to the optical losses incurred at each optical splitter. The multiple wavelengths of a WDM PON can be used to separate Optical Network Units (ONUs) into several virtual PONs co-existing on the same physical infrastructure. Alternatively the wavelengths can be used collectively through statistical multiplexing to provide efficient wavelength utilization and lower delays experienced by the ONUs. Corecess and Korea Telecom have done important research. Novera Optics Inc. has introduced first products, which supports the WDM PON deployment in Korea. DOCSIS-PON Data Over Cable Service Interface Specification (DOCSIS) PON, or DPON/DPON, is a type of passive optical networking, being proposed by several companies, that implements the DOCSIS service layer interface on existing Ethernet PON (EPON or GEPON) Media Access Control (MAC) and Physical layer (PHY) standards. In short it implements the DOCSIS Operations Administration Maintenance and Provisioning (OAMP) functionality on existing EPON equipment. It makes the EPON OLT look and act like a DOCSIS Cable Modem Termination Systems (CMTS) platform. Some DPON systems may optionally support the Metro Ethernet Forum (MEF) 9 and 14 specifications for the delivery of Ethernet Transport services including Ethernet LANs (ELAN), Ethernet Virtual Private Line (EVPL), and point to point Ethernet Transport (ELINE) services. In these instances the DPON system also acts as an IP/MPLS Provider Edge (PE) Router. RF-PON or RFOG Radio Frequency PON (RF-PON) or Radio Frequency over Glass (RFOG) or Hybrid-Fiber-Coax PON (HFC-PON) or Cable PON, is a type of passive optical networking, that proposes to transport RF signals that are now transported over copper (principally over a hybrid fiber and coaxial cable) over PON. In the forward direction RF-PON is an optical overlay for existing PON such as GPON or GEPON/EPON. The overlay for RF-PON works in the same way that some CWDM PON or potential WDM-PON overlays work. Reverse RF support may be provided by digitizing and transporting the upstream or return RF into the digital PON upstream overhead, or by the use a of a CWDM overlay. Implementations vary by vendor and there are no standards for the return portion of the RF overlay for PONs. RF-PON offers backwards compatibility with existing RF modulation technology, but offers no additional bandwidth for RF based services. It offers a means to support RF technologies in locations where only fiber is available or where copper is not permitted or feasible. The Future of PON On paths to next-gen PON: There are two general approaches, both in the standards bodies and among most vendors: the higher line rate, similar to current GPON but run at faster speeds like 10 Gb/s; the other is to run multiple wavelengths on the same fiber and use wavelength-specific filters at the [optical network terminals]. That could be either using DWDM or CWDM wavelengths, depending on how many wavelengths you want to run and the performance [you need]. Two standards bodies are working on 10G: The IEEE is working on an Ethernet or EPON flavor, and [Full Service Access Network] has started on a successor to 2.5G GPON at 10 Gb/s. There’s been some talk about trying to unify those standards, but that hasn’t progressed very far. Right now at least the optics or lasers used could be used for both technologies, even if the framing mechanisms and protocols are different. It’s still early, but one general trend we see is more interest in 10 Gb/s GPON for business or enterprise services. There’s an expectation that GPON, because of the way it was standardized by FSAN and the ITU, is a little more stringent in meeting [service level agreements]. Also, GPON can support an embedded synchronization mechanism if you want to do something that requires synchronization like TDM-type emulation. We see the 10G flavor of EPON [more among] CLECs or [in] developing countries like China, where the service provider may not have lots of existing infrastructure or their main target is to get high-speed Internet to as many customers as possible. On 10G PON: There’s a number of alternatives on how to do 10G. It depends on your confidence level in your optics and your assumptions about span design. If you want to maintain 20 km and 32-way splits, you’ve got to increase your launch power on 10G or have very good receiver designs that can handle signals with low receive levels. There’s going to be varying points of view from technology companies: increase power, reduce split ratios, shorten distances, have really good receiver designs (or combinations of those things) or go to parallel modulations. A lot of different approaches will be suggested, and it will boil down to: What’s your assumption of [outside plant] span design? Do you still think you have to meet 20 km and 32-way splits? That was driven very much by a North American model — typical suburbia. If you think your main markets are going to be high-density big cities in Asia or Europe, 32-way splits with 20 km may not make sense. On hybrids: For example, running 2.5 Gb/s PON on, say, four CWDM wavelengths on the same fiber at the same time — technically there’s 10 Gb/s running on the fiber, but it’s four times 2.5. Most people in the standards bodies wouldn’t consider that true next-gen PON. NG PON is either 10G or DWDM PON. A lot of people consider running 2.5G on four wavelengths an interim step between current GPON and NG GPON. It’s a hybrid. You’d need to put wavelength filters at the ONTs. Whether you can add a filter to your current ONTs or will need to replace them depends on the design of your current ONTs. On RF video: That’s one of the issues being debated in FSAN as they look at 10G. If your solution to 10G is to increase the transmission power at the [optical line terminal], you run into interference issues with [radio frequency] wavelengths. Will RF overlay still be needed at the point 10G GPON comes on the scene? I don’t know. Customers like to have options. Getting a customer to say, ‘I’ll never do another RF wavelength’ — customers tend not to be telling us that. But it has major implications for how you approach the cost of the technology. Verizon is still not very comfortable with IP video. How quickly will that change over time is the crystal-ball question. In a hybrid approach, RF overlay impacts the ONT and splitter designs; one of the wavelengths has to be reserved for RF. The IEEE is not considering [RF video] at all for 10G EPON. On the timing of commercial NG PON gear: In 2010 you might start to see early products, with mass deployment in 2011 or later. WDM PON will probably be the same, if not further out. Hybrid 2.5 Gb/s gear is theoretically feasible now. You might see 2.5G hybrid approaches in the next year or two for niche applications — business parks or residential high-rises. On DOCSIS 3.0 and NG GPON: I don’t think 3.0 is really driving interest in NG PON. What comes after 3.0 might drive a move to WDM. 3.0 creates at best parity with GPON, and that’s debatable. On upstream bandwidth: There’s no clear consensus yet on what you do upstream but lots of discussion on the trade-offs and costs. The End. Thank you!