FTTH broadband fiber access

FTTH technology

Advantages of network access

From a global perspective, most telecom companies develop broadband access based on ADSL. However, ADSL is Broadband access technology based on copper wire. Copper is a global strategic resource. As the price of international copper cable continues to rise (an average annual increase of 20%-30% in recent years), the development of xDSL based on copper cable The cost of the circuit is getting higher and higher, and the raw material of the optical fiber is silica, which is inexhaustible and inexhaustible in nature. In fact, the current market price of optical fiber is already lower than that of ordinary copper wire, and its life span is much higher than the latter. In the replacement of new subscriber lines or old cables, optical fiber has become a more reasonable choice, especially in the backbone section and even the distribution section. Secondly, as an active device, xDSL electromagnetic interference is difficult to avoid, and maintenance costs are getting higher and higher. Optical fiber as a passive transmission medium can avoid such problems.

Finally, as the entire network's optical fiberization process continues to extend to the user side, the end-to-end broadband connection limitation is increasingly concentrated in the access section. The current ADSL uplink and downlink connection rate cannot satisfy high-end users. Long-term business needs. Although ADSL2+ and VDSL2 technologies are expected to alleviate this pressure, the continued substantial increase in their speed and transmission distance is limited, and a substantial breakthrough cannot be expected. Obviously, with the large number of applications of optical fiber in long-distance networks, metropolitan area networks, and even the backbone of the access network, the logical development trend is to continue to extend the optical fiber to the distribution section and lead-in part of the access network, and finally achieve fiber to the home. . The key question is: How fast is the advancing speed? This will depend on a variety of factors, including market demand, competition needs, application stimulus, technological progress, cost reduction, and the development of supporting operation and maintenance systems. my country's hosting of the Olympic Games in 2008 and the hosting of the World Expo in 2010 will also promote the development of FTTH to a certain extent.

Ethernet System

Historically, Ethernet technology has always been the most popular method in the application environment of enterprises and institutions, and it has become the second largest residence after the power supply socket. Interface with office utilities. The main reason is that there is a huge network foundation and long-term experience and knowledge. At present, all popular operating systems and applications are also compatible with Ethernet, with good performance-price ratio, scalability, easy installation and opening, and high reliability.

For the public network residential user application environment, the point-to-point active Ethernet system uses active service concentration points to replace the passive components of the passive point-to-multipoint system, so that the transmission distance can be extended to 120km Far away. The main advantage of this technology is dedicated access, bandwidth is guaranteed, each user can enjoy 100Mbit/s or even 1Gbit/s in the wiring section and the incoming line section; the central office equipment is simple and cheap; the transmission distance is long, and the service area is large; and the cost is It increases linearly with the actual increase in the number of users, is predictable, does not require planning, has low investment risk, and has high equipment port utilization. Therefore, the cost is low in low-density user distribution areas. The disadvantage is that the equipment and optical fiber facilities at both ends are dedicated, and users cannot share the central office equipment and optical fibers. When the demand increases rapidly and the users are dense, the number of optical fibers and equipment at both ends and their cost and space requirements also increase rapidly. Too suitable for high-density user areas. In addition, active Ethernet requires multi-point power supply and backup power supply. There are many network management components (including power supply), which increases the complexity of power supply and network management. Third, from the perspective of standardization, Active Ethernet does not have a unified standard, but uses multiple related standards to produce a variety of incompatible solutions. Finally, another factor that may affect the choice of Ethernet technology is the provision of traditional video services. For example, some American telecommunications companies (such as Verizon) promise to provide traditional analog radio frequency video programs of the same quality, while Ethernet technology is supporting traditional analog radio frequency. The transmission of video programs is more difficult.

In FTTH applications, point-to-point Ethernet is mainly used for multi-household unit access, which is divided into single-fiber system and dual-fiber system. The uplink and downlink of the single-fiber system use different wavelengths. , The typical upstream wavelength is 1310nm, the downstream wavelength is 1550nm, and the transmission distance is 15km. It follows the standard TS-1000 formulated by the Japan Telecommunications Technical Committee, so the interoperability is good and the network complexity is low. The dual-fiber system uses two optical fibers, follows the IEEE802.3ub standard, uses multimode optical fibers, and the transmission distance is only 2km. In order to extend the transmission distance and enhance its management functions, different operators have formulated many private standards, which makes the interoperability of the system very poor.

Passive optical network

Introduction

Passive optical network is a pure medium network, its main feature is to remove the active Equipment, thereby avoiding the influence of electromagnetic interference and lightning, reducing the failure rate of lines and external equipment, simplifying the complexity of power supply configuration and network management, and reducing operation and maintenance costs. Secondly, PON has good service transparency and wide bandwidth, can be applied to signals of any standard and rate, can support analog broadcast and television services economically, and has a triple-play function. Third, the central office equipment and optical fiber (from the feeder section to the lead-in line) are shared by users, so the length of the optical fiber line and the number of transceiver equipment are small, and the corresponding cost is lower than other point-to-point communication methods, and the civil construction cost can also be significant. reduce. Especially with the increasing advancement of optical fiber to users, its comprehensive advantages have become more and more obvious. The cost per user of PON decreases rapidly with the increase in the number of users sharing the OLT, so it is most suitable for scattered small businesses and residential users, especially those where the user area is relatively dispersed, and the users in each area are relatively concentrated in a small area. User areas, especially new areas. Finally, the degree of standardization of passive optical networks is good, and they are basically divided into two categories: ITUFSAN (Full Service Access Network) and IEEE, both of which can provide independent and feasible single compatible solutions. Therefore, most large US telecommunications companies tend to choose PON instead of optical Ethernet technology.

The main disadvantage of PON is the high one-time investment cost, because the central office optical line terminal (OLT) is very expensive, and the passive infrastructure such as optical fiber and splitter must be in place again. When there are few users or the distribution of users exceeds a certain limited distance, the cost per user is very high, which will generate a lot of sunk costs. In addition, its tree-shaped branch topology prevents users from having protection functions or the cost of protection functions is high, which affects large-scale development.

From the analysis of the network structure, no matter which PON can have two different structures, namely centralized and distributed, the former has only one fiber between the central office OLT and the service flexibility point (FP) Connected, the splitter is centrally placed at the FP (that is, the traditional transfer box), and a dedicated optical fiber is connected from the splitter to the user's optical network terminal. In the distributed structure, splitters are placed at both the flexible point and the distribution point (DP) to form a two-level shunt. Analysis shows that the distributed structure has a cost advantage when the user penetration rate is close to 100%, but the actual situation is mostly not the case, especially for the case where the user penetration rate is not high, the centralized structure has obvious cost advantages, and the cost can be With the growth of the actual number of users, there is no problem of the large initial sunk cost of the distributed structure, and it will not need to be redeployed with the advancement of technology (such as the emergence and application of GPON).

An important trend of passive optical network technology is to provide a variety of voice processing methods, which can be connected to the PSTN through the V5 interface at the central office to provide traditional PSTN voice services, and the control module can be built in the central office. , Support H.248/H.323 protocol, flexibly adapt to H.248 protocol-based softswitch VoIP network or H.323 protocol-based traditional VoIP network, its main development trend is to focus on supporting softswitch network.

APON and BPON

The early narrowband passive optical networks were based on TDM, and their performance/price ratio was not good, and they have naturally died out. ATM-based passive optical networks (APON/BPON) can use ATM's centralized and statistical multiplexing, combined with the sharing function of passive splitters on optical fiber and optical line terminals, greatly improving the performance-price ratio. Currently in the United States and Japan Some 1.5 million lines have been laid in other countries.

However, the service adaptation of APON/BPON is very complicated, the service provision capacity is limited, the data transmission rate and efficiency are not high, and the cost is high. Its market prospects are bleak due to the fading of ATM. Finally, from the perspective of business development trends, the available bandwidth of APON is still insufficient. Take FTTC as an example. Although the typical backbone downlink rate can reach 622Mbit/s, the bandwidth that can actually be allocated to each user will be greatly reduced after branching. According to the calculation of 32 channels, the available bandwidth of each branch is only 19.5Mbit/s, and then shared by 10 users, each user can only be allocated about 2Mbit/s of bandwidth. Obviously, such a price-performance ratio cannot meet the development needs of networks and services.

EPON

With the rise and development of IP, someone proposed the concept of EPON, that is, on the basis of a similar structure to APON and G.983, trying to retain its essence -Physical layer PON, and replace ATM with Ethernet as the link layer protocol, forming a new combination-EPON that can provide greater bandwidth, lower cost and stronger business capabilities. This idea has received a positive response in the Ethernet community, and the EPON standard has been formed under the banner of IEEE802.3ah. In order to distinguish it from the previous non-standard EPON based on 100Mbit/s in Japan, it is called GEPON. In view of the fact that the non-standard EPON based on 100Mbit/s has died out, the current EPON actually refers to GEPON, and no special distinction is made anymore.

EPON is mainly based on the IEEE802.3ah standard. The main difference from traditional point-to-point Ethernet is that it uses point-to-multipoint communication. Its downstream direction works in TDM mode, and the data stream is broadcast to ONUs in variable-length Ethernet frames, and each ONU decides whether to choose according to the MAC address of the Ethernet frame. The upstream direction works in TDMA mode, and ONU data streams from different time slots are converged to public optical fiber facilities and OLT. In addition, traditional Ethernet works in continuous optical transmission mode, and both directions are continuous bit streams, so the timing and decision of the receiving end are easy to implement. While the upstream bit stream of EPON is a burst of data packets sent in turn, the OLT's receiving timing recovery, decision threshold setting, ranging and delay compensation are more complicated.

From the perspective of the structure of EPON, its key advantage is that it greatly simplifies the traditional multilayer overlay network structure. The main features are:

●eliminates the ATM and SDH layers, thereby Reduced initial cost and operating cost;

●Downlink service rate can reach 1Gbit/s, allowing support for more users and higher bandwidth;

●Simple hardware, no outdoor electronic equipment , So that the installation and deployment can be simplified;

●A large number of mature Ethernet technology chips can be used, which is simpler and lower in cost;

●Improved flexible circuit assignment and business The ability to provide and reconfigure;

●Provides multi-layer security mechanisms, such as VLAN, closed user group, and VPN support.

The uplink and downlink wavelengths of the IEEE802.3ah standard EPON technology are 1310nm and 1490nm, the uplink and downlink rates are both 1.25Gbit/s, the transmission distance is 10/20km, and the split ratio is 32/16. The main business It is data and voice. After adding a 1550nm TV broadcasting wavelength, it becomes the so-called triple service bundled service of voice, data and TV. For transmitting a single Ethernet service, EPON is a good solution.

The main disadvantage of EPON is that IEEE802.3ah only specifies the MAC layer and the physical layer, and the standards above the MAC layer are developed by manufacturers themselves, which brings flexibility and also causes poor device interoperability. Shortcomings. Secondly, the overall efficiency of EPON is low, mainly due to the use of 8B/10B line coding, the introduction of 20% bandwidth loss, plus other additional overhead, the available load is only about 50%, while APON and GPON both use NRZ interference The code is a line code, and there is no bandwidth loss. GPON's GFP encapsulates 4-65535 bytes per frame, which is much larger than the Ethernet frame load of 46-1500 bytes, and the average overhead is small. Coupled with the efficiency of the carrier layer, the efficiency of the transmission convergence layer, and the efficiency of service adaptation, the overall transmission efficiency of EPON is Lower, about half of GPON. Third, since EPON started as a standard driven by Ethernet equipment manufacturers, it did not fully consider the operational needs of network operators, and its management functions were not rich enough, but it was significantly improved compared to ordinary Ethernet. It could provide remote fault indication and remote Basic management functions such as end loopback control and link monitoring can also meet the basic management functions. Finally, the design of EPON did not consider directly supporting services other than Ethernet, which is an important shortcoming for traditional operators who advocate multi-service support capabilities. FTTH broadband optical fiber access-

GPON

In 2001, while the IEEE was actively formulating the EPON standard, the FSAN organization began to initiate the formulation of a PON network standard with a rate exceeding 1Gbit/s-Kyrgyzstan Bit Ethernet Passive Optical Network (GPON). Subsequently, ITU-T also intervened in the formulation of this new standard and passed two new GPON standards-G.984.1 and G.984.2 in January 2003.

According to the regulations of this latest standard, GPON can provide downstream rates of 1.244Gbit/s and 2.488Gbit/s and a variety of standard upstream rates specified by ITU, that is, it can provide symmetric and asymmetric rates flexibly . The transmission distance is at least 20km, and the system split ratio can be 1:16, 1:32, 1:64 or even 1:128, while EPON only provides 1.25Gbit/s symmetrical rate, and the split ratio is up to 1:32. That is, GPON has advantages in speed, speed flexibility, transmission distance, and split ratio. Secondly, GPON adopts two adaptation methods. In addition to the traditional ATM, it also adopts a new SDH-based standard universal framing procedure (GFP) at the transmission convergence layer. This kind of data signal is encapsulated into the general standard signal adaptation mapping technology of the existing SDH network, which can adapt to any user signal format and any transmission network standard, without additional ATM or IP encapsulation layer, high encapsulation efficiency, and flexible service provision, while APON/BPON EPON and EPON need to provide specific adaptation methods for each specific service. Third, because GPON uses GFP mapping, its transmission convergence layer is essentially synchronous, and it also uses standard SDH 125μs frames, so that GPON can support end-to-end timing and other quasi-synchronous services, especially direct high-quality, flexible Support real-time TDM voice service, with good delay and jitter performance. However, EPON has no specific regulations for carrying TDM services, which leads manufacturers to adopt different methods to carry them, including one, two and three layers. The interoperability is poor and the performance is difficult to ensure. Fourth, GPON has rich functions in network management, including bandwidth authorization allocation, dynamic bandwidth allocation, link monitoring, protection switching, key exchange, and various alarm functions, which are more thoughtful than EPON. However, EPON has significantly improved network management functions over ordinary Ethernet. It can provide basic management functions such as remote fault indication, remote loopback control, and link monitoring, and can also meet basic management functions. Fifth, in terms of QoS, GPON can use pointers to adjust the authorized bandwidth and authorized period of the ONU to ensure the bandwidth and delay requirements of the service. EPON mainly uses priority queues combined with DBA algorithms to ensure bandwidth and delay, and can basically meet the QoS requirements of different services.

From a technical point of view, GPON is the inheritance and development of BPON. GPON inherits many basic features of BPON. For example, both use the same OLT core technology, including ONU activation and ranging, use the same physical fiber facilities and optical power budget, and the same management software stack. On the other hand, GPON has adopted some of the latest technological achievements, in addition to the most important GFP packaging technology, it also includes new technologies such as forward error correction.

From the perspective of the services provided, GPON can not only provide 10/100Mbit/s, 1Gbit/s services, but also provide VLAN services and voice services. In fact, it can adapt to any existing services and new future services. Adaptation requirements. In general, GPON is not a manufacturer-driven technical standard, but an operator-driven standard. Therefore, it has more thoughtful operational benefits, higher speed, and greater speed flexibility; it has a common mapping format and can be Adapt to any new and old services; have rich OAM&P functions; have high transmission efficiency for various services, and can transmit flexibly and efficiently even for TDM services. It can help operators complete the smooth transition from traditional TDM voice circuits to all-IP networks.

As far as cost analysis is concerned, the cost of PON optical module is about 20%-30% of the equipment cost. The main cost is various electrical interfaces and protocol processing conversion. In this respect, GPON and BPON are more EPON is much more complicated. Secondly, as far as optical modules are concerned, because GPON has to meet high burst synchronization indicators, it has high requirements for the module's drive circuit and front and rear amplifier chips, and also has to meet a higher power budget. Only distributed feedback lasers (DFBs) can be used. ) The cost of transmitter and avalanche photodiode (APD) receiver is higher than that of Fabry-Perot cavity (FP) transmitter and photodiode (PIN) receiver of EPON module. The yield is also lower, so the whole The cost of optical modules is relatively high. In addition, EPON has entered the stage of mass production, while GPON has not yet entered the stage of mass production, leading to a clear advantage in cost for EPON at present.

As far as transmission efficiency is concerned, GPON is the highest in terms of scrambling efficiency, transmission convergence layer efficiency, bearer protocol efficiency and service adaptation efficiency, so its total efficiency is the highest. For example, assuming that TDM services account for 10% and data services account for 90%, the total efficiency of GPON is 94%, while APON and EPON are 72% and 49%, respectively.

The main disadvantage of GPON is that although ONU only needs to support one of ATM and GFP adaptation, OLT must support both at the same time, that is, it must retain complex ATM layer functions, plus optical modules The technical difficulty is higher, which makes the equipment cost higher. In addition, the maturity of GPON is not as good as that of EPON. Currently, no professional chip manufacturer has launched truly commercial GPON core chips and optical modules. EPON has many providers, and the core chips have been developed to the third-generation system-on-chip (SoC) stage. The cost of optical modules has also been reduced to a level close to that of ordinary Gigabit Ethernet.

In general, in the current situation where the output is not large, the equipment cost of GPON and BPON is much higher than that of EPON. With the advancement of technology and the large-scale increase of output, the cost difference will be Gradually decreasing, the total cost will probably ultimately depend on the output size, that is, the market's choice.

The common challenges faced by GPON and EPON are: how to effectively carry TDM services on Ethernet/GFP and provide carrier-class service quality; secondly, because GPON and EPON are point-to-multipoint star Or tree network, need to realize the protection and recovery of carrier-class through a 1+1 optical network passing through different routes, the network cost will be very high; third, the current cost of GPON and EPON equipment is mainly limited by the burst optical transmission/ Receiving modules and core control modules/chips, these modules are either immature or too expensive to meet market needs; fourth, the one-time investment cost of GPON and EPON is relatively high, which is not suitable for traditional telecommunications with gradual investment and expansion The construction mode is most suitable for dense user areas that are completely newly built or rebuilt.

Thinking about development

FTTH is not a new concept. It has a history of 27 years and is a second development opportunity. The first time was France, Canada and Japan in 1978, and the second time around 1995 was mainly the United States and Japan. Both development opportunities died due to high costs and lack of market demand. Driven by technological progress, macro information society development policies and telecommunications regulatory policies in certain countries, FTTH has entered the third development opportunity period since 2004.

However, although the price of FTTH equipment has fallen sharply, it is still as high as US$300 per household, which is almost 10 times that of ADSL. The policy risk and market risk of operating video services are still very large. Therefore, my country It does not yet have the conditions for large-scale commercial use of FTTH, and is currently in the stage of field trials and commercial trials. In fact, the development of FTTH not only depends on technology, but more importantly, cost, business and business model.

For China Telecom, which is undergoing overall corporate transformation, FTTH has become an important area that affects future network and technology business transformation. China Telecom will carry out FTTH field trials in four provinces and cities in accordance with the guidelines of “strengthening leadership, active testing, timely summary, and orderly promotion” to accumulate technology, planning, business, commercial models, and operations for future large-scale applications. Comprehensive experience in maintenance. With the approach of the 2008 Olympic Games and the 2010 World Expo, the practical application of FTTH in our country is approaching. The ideal of fiber to the home is no longer an unattainable vision, but sufficient patience and comprehensive and solid preparation are not only impossible What's missing is also the only way to success.