Introduction of various PON systems
1. APON technology
In the mid-1990s, some major network operators established the Full Service Access Network Alliance (FSAN), whose purpose is to formulate a unified standard for PON equipment so that equipment manufacturers and operators can enter the PON equipment market and compete together. The first result is the specification of the 155Mbit / s PON system standard in the ITU-T G.983 series of recommendations. Because ATM is used as the bearer protocol, this system is called APON system, and it is often misunderstood as only providing ATM services. Therefore, it is renamed Broadband Passive Optical Network (BPON) system to show that this system can provide Ethernet Broadband services such as network access, video distribution, and high-speed leased lines. However, for this generation of FSAN systems, the most commonly used name is APON. Later, the APON standard was enhanced, and it began to support downlink 622 Mbit / s rates, and new features were added in protection methods, dynamic bandwidth allocation (DBA), and other aspects.
APON uses ATM as the bearer protocol. Downstream transmission is a continuous ATM stream with a bit rate of 155.52Mbit / s or 622.08Mbit / s. A special physical layer operation management and maintenance (PLOAM) cell is inserted into the data stream.Upstream transmission is ATM cells in burst form. In order to achieve burst transmission and reception, a 3-byte physical overhead is added in front of each 53-byte cell. For a basic rate of 155.52 Mbit / s, the transmission protocol is based on a downlink frame containing 56 ATM cells (53 bytes per cell); when the bit rate is increased to 622.08 Mbit / s, the downlink frame is expanded to 224 Cell. At the basic rate of 155.52 Mbit / s, the format of the uplink frame is 53 cells, each cell is 56 bytes (53 ATM cell bytes plus 3 bytes overhead). In addition to the 54 data cells in the downlink frame, there are two PLOAM cells, one at the beginning of the frame and the other in the middle of the frame. Each PLOAM cell contains the uplink transmission authorization for the specific cell in the upstream frame (53 Upstream frame cells have 53 grants mapped into PLOAM cells) and OAM & P information. APON provides very rich and complete OAM functions, including bit error rate monitoring, alarming, automatic discovery, and automatic search. As a security mechanism, it can scramble and encrypt downlink data.
From the perspective of data processing, in APON, user data must be transmitted under protocol conversion (AAL1 / 2 for TDM and AAL5 for data packet transmission). This conversion is difficult to adapt to high bandwidth, and the equipment that performs this function includes some related auxiliary equipment, such as cell memory, Glue Logic, etc., which also adds a lot to the system cost.
Now, whether it is a long-distance core transmission network or a metropolitan area access network convergence layer, digital communication technology has gradually shifted from ATM-centric to IP-based to provide video, audio, and data communications. Therefore, only the access network structure that can adapt to both current access and future network core technologies can make the future all-optical IP network a reality.
APON has gradually withdrawn from the market due to its complexity and low data transmission efficiency.
2. EPON
Almost at the same time as the APON system, the IEEE also established the first mile Ethernet (EFM) research group to launch Ethernet-based EPON (Ethernet Passive Optical Network) in terms of fiber access networks, showing a good market prospect. The study group belongs to the IEEE 802.3 group that developed the Ethernet standard. Similarly, its research scope is also limited to the architecture, and it must conform to the existing 802.3 media access control (MAC) layer functions. In April 2004, the research group introduced the IEEE 802.3ah standard for EPON, with an uplink and downlink rate of 1 Gbit / s (using 8B / 10B coding, and a line rate of 1.25 Gbit / s), ending the EPON manufacturers ’use of private protocols to develop equipment standard status.
EPON is a broadband access system based on Ethernet technology. It uses the PON topology to implement Ethernet access. The key technologies of the data link layer mainly include: Multiple Access Control Protocol (MPCP) for the uplink channel, the plug and play problem of the ONU, the ranging and delay compensation protocols of the OLT, and protocol compatibility issues.
The physical layer of IEEE 802.3ah includes both point-to-point (P2P) connected optical fibers and copper wires, as well as PON network scenarios for point-to-multipoint (P2MP). In order to facilitate network operation and fault repair, the OAM mechanism is also included. For P2MP network topology, EPON is based on a mechanism called Multipoint Control Protocol (MPCP), which is a function within the MAC sublayer. MPCP uses messages, state machines, and timers to control access to the P2MP network topology. Each optical network unit (ONU) in the P2MP network topology has an MPCP protocol entity that communicates with the MPCP protocol entity in the OLT. .
The basis of the EPON / MPCP protocol is a point-to-point simulation sublayer, which makes a P2MP network look like a collection of P2P links to higher protocol layers.
In order to reduce the cost of the ONU, the key technologies of the EPON physical layer are concentrated on the OLT, including fast synchronization of burst signals, network synchronization, power control of optical transceiver modules, and adaptive reception.
EPON combines the advantages of PON and Ethernet data products to form many unique advantages. The EPON system can provide uplink and downlink bandwidths of up to 1 Gbit / s, which can meet the needs of users in the future for a long time. EPON uses multiplexing technology to support more users, and each user can enjoy greater bandwidth. The EPON system does not use expensive ATM equipment and SONET equipment, and is compatible with the existing Ethernet, greatly simplifying the system structure, low cost, and easy to upgrade. Due to the long life of passive optical devices, the maintenance costs of outdoor lines are greatly reduced. At the same time, standard Ethernet interfaces can take advantage of existing low-cost Ethernet equipment and save costs. The PON structure itself determines that the network is highly scalable. As long as the terminal equipment is replaced, the network can be upgraded to 10 Gbit / s or higher. EPON can not only integrate the existing cable TV, data and voice services, but also be compatible with future services such as digital TV, VoIP, video conferencing and VOD, etc., to achieve integrated service access.
The comprehensive use of EPON bearer and other access technologies further enriches broadband access technology solutions.
Using EPON can make DSL break the traditional distance limitation and expand the coverage. When the ONU is integrated into the Digital Subscriber Line Access Multiplexer (DSLAM), the reachable range of the DSL and its potential user group will increase greatly.
Similarly, by integrating the CMTS (Cable Modem Termination System) of the ONU, EPON can provide bandwidth to existing Cable connections, and allow cable operators to implement truly interactive services while reducing construction and operating costs.
In both cases, operators can increase their user base based on their existing network structure and investment. EPON can also extend the point-to-point MSPP (Multiple Services Provisioning Platform) and IP / Ethernet.
In addition, EPON technology can also be used to solve the problem of the uplink data of the base station in the wireless access technology pooled to the core network.
3． GPON
In 2001, FSAN launched a new effort to standardize PON networks operating above 1 Gbit / s. In addition to supporting high rates, the entire protocol has been open in order to rethink and find the best and most effective solution in terms of supporting multi-service, OAM & P functions and scalability. As part of GPON’s work, FSAN first gathered the requirements of all its members (including major operators around the world), then based on this, wrote a document called Gigabit Service Requirements (GSR) and made it a formal recommendation ( G.GON.GSR) to ITU-T. The main GPON requirements described in the GSR file are as follows.
l Supports full services, including voice (TDM, SONET / SDH), Ethernet (10/100 Base-T), ATM, leased lines, etc.
l The physical distance covered is at least 20km, and the logical distance is limited to 60km.
l Supports various bit rates using the same protocol, including symmetrical 622 Mbit / s, symmetrical 1.25 Gbit / s, downstream 2.5 Gbit / s and upstream 1.25 Gbit / s, and other bit rates.
l OAM & P powerful functions that can provide end-to-end service management.
l Due to the broadcast characteristics of PON, the security of downlink services must be guaranteed at the protocol level.
FSAN proposed that the design of the GPON standard should meet the following goals.
l The frame structure can be expanded from 622Mbit / s to 2.5Gbit / s, and supports asymmetric bit rate.
l Guarantee high bandwidth utilization and high efficiency for any business.
l Encapsulate any service (TDM and packet) into a 125ms frame through GFP.
l Efficient and cost-free transmission of pure TDM services.
l Dynamic bandwidth allocation for each ONU through a bandwidth pointer.
Since GPON reconsidered the application and requirements of PON from the bottom up, it laid the foundation for the new solution and is no longer based on the previous APON standard, so some manufacturers call it native PON (natural mode PON). On the one hand, GPON retains many functions that are not directly related to PON, such as OAM messages, DBA, etc. On the other hand, GPON is based on a new TC (transmission convergence) layer. The GFP (general framing procedure) selected by FSAN is a frame-based protocol that adapts service information from high-level customers of the transport network through a general mechanism. The transport network can be any type of network, such as SONET / SDH and ITU-T G.709 (OTN), etc. The customer information can be packet-based (such as IP / PPP, ie IP / Point to Point protocal, or Ethernet MAC frames, etc. ), Can also be a constant bit rate stream or other types of business information. GFP has been officially standardized as ITU-T standard G.7041. Because GFP provides an efficient and simple way to transmit different services on the synchronous transmission network, it is ideal to use it as the basis of the GPON TC layer. In addition, when using GFP, the GPON TC is essentially synchronous and uses standard SONET / SDH 8kHz (125ms) frames, which enables GPON to directly support TDM services. In the officially released G.984.3 standard, FSAN’s proposal on GFP as the TC layer adaptation technology was adopted, and further simplified processing was done, named GPON encapsulation method (GEM, GPONEncapsulationMethod).
Application of EPON system
EPON, as a new broadband access technology, is a full-service provisioning platform that can support data services as well as real-time services such as voice and video.
The optical path design of EPON can use 3 wavelengths. If you do not consider supporting CATV or DWDM services, two wavelengths are generally used. When using 3 wavelengths, the upstream wavelength is 1310nm, the downstream wavelength is 1490nm, and an additional 1550nm wavelength is added. The increased 1550nm wavelength is used to directly transmit analog video signals. Because the current analog video signal is still dominated by radio and television services, it is estimated that it will not be completely replaced by digital video services until 2015. Therefore, the currently designed EPON system should support both digital video services and analog video services. The original 1490nm still carries downlink data, digital video and voice services, and 1310nm transmits uplink user voice signals, digital video on demand (VOD), and request information for downloading data.
Voice signals have strict requirements on delay and jitter, and Ethernet does not provide end-to-end packet delay, packet loss rate, and bandwidth control capabilities. Therefore, how to ensure service quality when EPON superimposes voice signals is an urgent problem to be solved.
1. TDM business
At present, the most questionable EPON multi-service capability is its ability to transmit traditional TDM services.
The TDM services mentioned here include two types of voice services (POTS, Popular Old Telephone Service) and circuit services (T1 / El, N´64kbit / s leased lines).
When EPON systems carry data dedicated line services (2048kbit / s or 13´64kbit / s data services), TDM over Ethernet is recommended. EPON system can adopt circuit switching or VolP when carrying voice services.
In the next few years, because the market demand for circuit services is still very large, the EPON system is required to carry both packet-switched services and circuit-switched services. How does EFM carry TDM on EPON and how to ensure the quality of TDM services. There are no specific provisions in technology, but they must be compatible with the Ethernet frame format. Multi-service EPON (MS-EPON) adopts E1 Over Ethernet technology, which efficiently solves the problem of adaptation of TDM services on Ethernet frames, enabling EPON to realize multi-service transmission and access. At the same time, MS-EPON overcomes the gap between OLT and ONU. The shared bandwidth contention phenomenon provides Ethernet users with a guaranteed bandwidth guarantee.
The encapsulation method of Ethernet makes EPON technology very suitable for carrying IP services, but it also faces a major problem-it is difficult to carry TDM services such as voice or circuit data. EPON is an Ethernet-based asynchronous transmission network. It does not have a high-precision clock synchronized across the network, and it is difficult to meet the timing and synchronization requirements of TDM services. To solve the problem of timing synchronization of TDM services while ensuring technical difficulties such as QoS of TDM services, we must not only improve the design of the EPON system itself, but also need to adopt some specific technologies.
The performance index of the circuit switched voice service indicates that when the EPON system uses the circuit switched method to carry voice services, it should meet the requirements of YDN 065-1997 “General Technical Specification for Telephone Switching Equipment of the Ministry of Posts and Telecommunications” and YD / T 1128-2001 “General Telephone Switching Equipment” Technical Specifications (Supplement 1) “requirements for pure circuit switched voice quality. Therefore, EPON currently has the following problems with TDM services.
① TDM service QoS guarantee: Although the bandwidth occupied by the TDM service is small, it has high requirements on indicators such as delay, jitter, drift, and bit error rate. This requires not only to consider how to reduce the transmission delay and jitter of the TDM service during uplink dynamic bandwidth allocation, but also to ensure that the TDM service strictly controls the delay and jitter in the downlink bandwidth control strategy.
② Timing and synchronization of TDM services: TDM services have particularly strict requirements on timing and synchronization. EPON is essentially an asynchronous transmission network based on Ethernet technology. There is no high-precision telecommunication clock synchronized throughout the network. The clock accuracy defined by Ethernet is ± 100´10 and the clock accuracy required by traditional TDM services is ± 50´10. In addition, while providing the telecommunications clock synchronized throughout the network, TDM data must be transmitted as periodically as possible to meet its jitter and error requirements.
③ EPON survivability: The TDM service also requires that the bearer network must have good survivability. When a major failure occurs, the service can be reliably switched in the shortest possible time. Because EPON is mainly used for access network construction, it is relatively close to users, and various applications and use environments are complex. It is easily affected by unknown factors such as urban construction, causing accidents such as link interruptions. Therefore, the EPON system is urgently required to provide a cost-effective system protection solution.
2. IP services
EPON transmits IP data packets without protocol conversion and has high efficiency, which is very suitable for data services.
VolP technology, as a hot technology in development, has achieved a certain scale of application in recent years, and is an effective means for carrying voice services over IP networks. In the EPON system, it is also possible to implement access to traditional telephone services by adding certain VoIP equipment or functions. Utilizing VoIP technology, as long as the delay and jitter characteristics of the EPON voice service are guaranteed, other functions are left to the user-side integrated access device (IAD, Integrated Access Device) and the central access gateway device to process the voice service Transmission. This method is relatively simple to implement and can directly port existing technologies, but requires expensive central office access gateway equipment, higher network construction costs, and is limited by the shortcomings of the VoIP technology itself. In addition, E1 and N´64kbit / s data services cannot be provided.
When the EPON system uses VoIP to carry voice services, it should meet the following performance indicators for VoIP voice services.
① The dynamic switching time of voice coding is less than 60ms.
② It should have 80ms buffer storage capacity to ensure that no speech discontinuities and jitters occur.
③ Objective evaluation of voice: When the network conditions are good, the average value of PSQM is less than 1.5; when the network conditions are poor (packet loss rate = 1%, jitter = 20ms, delay = 100ms), the average value of PSQM is <1.8; When the conditions are bad (packet loss rate = 5%, jitter = 60ms, delay = 400ms), the average PSQM is less than 2.0.
④ Subjective assessment of speech: When the network conditions are good, the average value of MOS is> 4.0; when the network conditions are poor (packet loss rate = 1%, jitter = 20ms, delay = 100ms), the average value of MOS is <3.5; network When the conditions are bad (packet loss rate = 5%, jitter = 60ms, delay = 400ms), the average value of MOS <3.0.
⑤ Encoding rate: G.711, encoding rate = 64kbit / s. For G.729a, the required coding rate is <18kbit / s. For G.723.1, the G.723.1 (5.3) coding rate is <18kbit / s, and the G.723.1 (6.3) coding rate is <15kbit / s.
⑥ Delay index (loopback delay): VoIP delay includes codec delay, input buffer delay at the receiving end, and internal queue delay. When G.729a encoding is used, the loopback delay is <150ms. When G.723.1 encoding is used, the loopback delay is <200ms.
3． CATV business
For analog CATV services, EPON can also be carried in the same way as GPON: add a wavelength (actually this is a WDM technology and has nothing to do with EPON and GPON itself).
PON technology is the best way to achieve FTTx broadband access. EPON is a new optical access network technology created by combining Ethernet technology and PON technology. It can be used to transmit voice, data and video services and is compatible. For some new services in the future, EPON will become the dominant technology for full-service broadband optical access with its absolute advantages such as high bandwidth, high efficiency, and easy expansion.
Protection scheme of PON system
In order to improve network reliability and survivability, a fiber protection switching mechanism can be used in the PON system. The optical fiber protection switching mechanism can be performed in two ways: ① automatic switching, triggered by fault detection; ② forced switching, triggered by management events.
There are three main types of fiber protection: backbone fiber redundancy protection, OLT PON port redundancy protection, and full protection, as shown in Figure 1.16.
Backbone fiber redundancy protection (Figure 1.16 (a)): using a single PON port with a built-in 1´2 optical switch at the OLT PON port; using a 2: N optical splitter; the OLT detects the line status; There are no special requirements for the ONU.
OLT PON port redundancy protection (Figure 1.16 (b)): The standby PON port is in a cold standby state, using a 2: N optical splitter; the OLT detects the line status, and the switching is done by the OLT, with no special requirements for the ONU.
Full protection (Figure 1.16 (c)): both the main and backup PON ports are in working state; two 2: N optical splitters are used; an optical switch is built in front of the ONU PON port, and the ONU detects the line status and determines the main use Lines and switching are done by the ONU.
The protection switching mechanism of the PON system can support automatic return or manual return of the protected services. For the automatic return mode, after eliminating the switching failure, after a certain return waiting time, the protected service should automatically return to the original working route. The return waiting time can be set.