SDH multi-business platform leads the development of the metro network
Http://www.chinatelecom.com.cn June 29, 2004 Wei Leping China Telecom Group Corporation
Abstract First introduces the deficiencies of traditional SDH and the need to develop the next-generation SDH standard, then briefly introduce several major new standards and applications, and finally focus on the evolution and latest development of SDH multi-business platform traditional SDH. Key words SDH Multi-Service Platform 1 Traditional SDH defects and roles have changed capacity of greater and lower costs of WDM systems, so that SDH is a historical change in the long-distance network. In addition to the region where the capacity demand in the long-distance network is not too large, the role of the SDH starts to transfer to the network edge. Given the complex client signal characteristics of the network edge, SDH must transform from the pure transmission network to the multi-service platform integrated with the traffic network and the traffic network, called the fused SDH multi-service platform. Second, in order to effectively support dynamic IP services, multi-service delivery platform needs to eventually realize fast and automatic service assignment. The requirements of these two aspects result in the next generation of SDH devices. Starting, the SDH multi-service platform is simply implemented to implement the fixed package and transparency of the data service on the existing SDH device, thereby providing more efficient Layer 2 exchange and local aggregation. However, the bandwidth finger of the traditional SDH network is achieved by a centralized network management system, which cannot accommodate a large-capacity IP business dynamic and unpredictable characteristics. Second, in the original SDH multiplexing structure, 4 times the bandwidth increment is too large for the optimal configuration of the data service. Finally, in a typical metropolitan network, all layers, including the SDH layer, the ATM layer, and IP layers are separate networks, which can only be configured and utilized by manual way. These defects have prompted the next generation SDH device to implement a series of new features in addition to a standardized VC virtual level function. To this end, ITU has developed a series of next-generation SDH standards, ie integrated universal group frame programs (GFPs), link capacity adjustment schemes (LCAS), and auto-swap optical network (ASON) standards. 2 Integrated Universal Frame Program CFP is a general standard signal adaptation mapping technology that can be transparently encapsulated into an open general standard signal of existing networks into existing networks. CFP can support frame mapping and transparent transmission within the same conveyor path. Frame Mapping Mode (GFP-F) can map customer signal frames into a variable length GFP frame, support the rate adaptation and multiplexing of the package particle level to achieve the flow of the virtual container level, the most suitable A frame mapping without a fixed length of the Ethernet / MAC or I swollen PP. This mapping mode is re-processed after receiving a complete data frame, so there is a need to have cache and Mac (media access control), which is currently only for Ethernet, less suitable for video and storage services. Transparent transmission (GFP-T) mode is completely different, and can be handled in time in a transparent mapping without waiting for the arrival of the entire frame. This mapping method is suitable for processing real-time services such as video signals and a storage domain (SAN) service with fixed frame length block-shaped signal format with Fiber Pass, ESCON, FICON, and the like. A general mapping standard GFP can replace many different and dedicated mapping methods, which can be applied to any customer signal. In addition, GFP is not limited to mapping to traditional VCs, and can be used for larger capacity bandwidth pipes, from 2.5 Gbit / s until 40Gbit / s. GFP uses a self-synchronous framework technology similar to ATM, simple and flexible, high cost, high standardization, and high standardization, facilitating multi-vendor equipment interconnection, can support various network topologies, can support user data implement statistical reuse, The error frame can be more effectively prevented, and the GFP also has a QoS mechanism. The corresponding transport layer is not limited to SDH, which can be a physical channel for OTN or other byte synchronization. In addition, using the simplified arbitrary character block each time the process, GFP reduces the processing requirements for data link mapping and de-mapping processes.
With the low-error characteristics of modern optical communication, GFP further reduces the complexity, device size and cost of the receiver, so that GFP is particularly suitable for high-speed transmission link applications. Finally, the marker space in the linear extension of the GFP can be integrated with the MPLS integration, using MPLS to enhance GFP, provide end-to-end operation and maintenance, protection, and business reuse capabilities. Since there is sufficient space to provide end-to-end OAM & P information and business characteristics, some people want to develop it into an independent networked layer similar to ATM, compensate for the client's layer signal lack of electricity-to-end OAM & P, protection and business The weakness of the reuse capabilities uses GFP to unify these features. However, most people still tend to adhere to the flat principles of the network architecture, do not want to have a new network layer, do not want GFP to take the ATM all-around the old road, will only use GFP only as a general adaptive mapping technology, but will Business features enhanced roles left to MPLS to complete. The data stream and GFP are simultaneously managed using MPLS, and the convergence of the transport layer and the data layer is completed. The evolution step of the GFP is first is a new generation of MSTP interface mapping scheme, and then expands to the access layer, enter the urban core switching layer, and finally enter the package switching layer. From the long way, it will replace Packet over SDH. 3 Link Capacity Regulation LCAS defines a method of flexible, dynamic, lossless change transmission network virtual level linkage bandwidth to automatically adapt to effective business bandwidth, particularly data service bandwidth requirements such as bandwidth dynamic changes like Ethernet. The specific implementation method is to define a virtual level group (VCG), using the overhead byte of SDH to transfer control information, and adjust the VC number through the VCG in this VCG, so that the Upper service bandwidth is quickly adapted. In addition, when some paragraphs have failed or repaired, some failed resource can be deleted or resumed via the LCAS protocol. The maximum advantage of using LCAS relative to traditional NMS is used, and the effective net load can be automatically mapped to the available VC, which not only avoids complex manual circuit cross-connect configuration, improves bandwidth finger speed, and does not Damage, and when the system part of the virtual container has failed, you can use the LCAS standard command to reconstruct the VCG, with a soft protection mode or fault tolerance mechanism. This mechanism can help enable free protection capacity in the traditional SDH network to improve bandwidth resource utilization. Further, it is also possible to decompose traffic into a few parts of the traffic to transmit, and each part of the traffic has different availability levels, accounting for less protection resources, thereby achieving more economical and effective sharing protection mechanisms. Briefly, LCAS does not need overlapping two-layer mechanism to provide bandwidth control and direct implementation of on-demand assignment bandwidth, providing end-to-end dynamic bandwidth adjustment mechanism to adapt to business bandwidth changes, can guarantee service quality The premise will significantly improve network utilization. In theory, LCAS can provide dynamic operation with GMPLS. But in fact, the current LCAS is still a modification of the virtual level, although the network management is simplified, but does not provide real dynamic operation. People worry that LCAS is like ATM SVC, can't develop due to dynamic operation, I would rather be as a long-term road sign, not fundamental requirements. 4 Automatic Switching Optical Network Automatic Exchange Optical Network (Ason) is an automatic connection management that automatically completes the network connection, using a separate Ason control plane through various transfer networks (including SDH or OTN). This optical network with independent control surface is commonly known as intelligent optical transmission network, referred to as an intelligent optical network.
Ason learned the experience of IP network intelligence, and improved the effective circuitry and signaling protocol used in IP networks to adapt to the application needs of optical networks, effectively solved the integration problem of IP layers and optical network layers, representing The direction of the next generation of optical network. The main benefits of introducing ASON in the network are: allowing network resources to dynamically assign network resources to routing; with scalable signaling capabilities; fast business provisioning and expansion; reduce maintenance management operation costs; rapid business recovery capabilities of optical layer; Reduce the needs of the running support system software for new technology configuration management, just maintain a dynamic database; you can also introduce new business types, such as on-demand bandwidth services, wavelength wholesale, wavelength rental, grading bandwidth business, dynamic wavelength Allocate leased business, bandwidth transactions, dynamic routing, optical layer virtual private network (VPN), etc., so that traditional transfer networks evolved in the direction of the service network. If the next generation of SDH multi-service platforms can integrate the above VC cascades, GFP, LCAS, and then cooperate with the core Ason's automatic circuit and assignment, not only greatly enhance its own flexible and effective support data. The capacity of the business, and the intelligence of the core smart optical network can be extended to the network edge to enhance the intelligence range of the entire network. 5 SDH multi-service platforms, SDH is in principle, and the SDH is the transfer layer technology, which does not know the net load carrying, so traditional SDH technology is difficult to effectively process data transfer requirements. However, facing the accelerated dataization of telecommunications services and IP and diverse business environments, SDH is continuously improved to effectively handle various important data services. Theoretically, SDH can know the type of net load content according to the channel overhead by this, thereby developing a new method of maping different data services into the SDH virtual container, resulting in the concept of SDH multi-service platform. The SDH multi-service platform is much largely increased in port density and integration, but also supports transmission and end of multiple data services on the SDH facility, implementing data service transparent or Layer 2 exchange and local aggregation. The point-to-point transmission method directly encapsulates the data map to the pre-allocated SDH virtual container transmission, simple; reduces POS port, low cost; has good user bandwidth guarantee and security isolation function, suitable for high QoS requirements Data Rental Service and Core Application. However, the bandwidth utilization is low, and the network hardware resource consumption is large, and the lack of flexibility. The two-layer exchange and aggregation mode users enter the network in the port, with bandwidth sharing, port aggregation capability, and user isolation and rate limits can be achieved by virtual local area network (VLAN), using SDH ring or fast Generate tree protection Implementation Layer 2 Ethernet business protection and ribbon sharing, saving network resources and ports. However, the two-layer exchange of competitive bandwidth makes it difficult to ensure that the user's actual bandwidth in the network congestion, especially the Etheri-ring network application, and the security is slightly poor. SDH multi-service platform with Layer 2 exchange and agglomeration can significantly reduce the service port of the node, reduce network cost, reduce the three-layer switch / router burden, the networking of the network, suitable for aggregation and access layer applications, and security Requires lower users and their browsing and video on demarcation services. Of course, the data service can also be further processed by the integrated router function. At this time, more and more flexible data networking functions can be enjoyed, the key is routing aggregation, which can reduce the complexity of the core router, but add three-layer transfer. Complexity. Therefore, the practice of supporting the simplified three-layer function is also an optional compromise. The starting point of the SDH multi-service platform is to make full use of people tried to SDH technology, especially their protection recovery, ensuring delayed performance and powerful network management capabilities to adapt to multi-service applications, support two-story or third floors. intelligent.
The basic idea is to map different SDH time slots directly or after processes, and the SDH device and the second layer and the three-layer grouping device are generated into an entity in physical integration. The next-generation SDH node integrated with the transfer layer is called the fused SDH multi-service node, which is positioned on the network edge. The emergence of the SDH multi-service platform not only reduces a large number of independent business nodes and transmitting node devices, simplifies the node structure, reduces equipment cost, reduces the number of racks, equipment room, power consumption and frame interconnection, simplified circuit Assignment, speed up the service providing speed, improve network scalability, save operation and maintenance and training costs, can also support various data services, especially Ethernet, Frame Relay, ATM, Storage Domain Network (SAN) After the IP routine, the bandwidth utilization rate of the TDM pathway can be improved by statistical multiplexing and excess ordering, and the number of ports of the local device is optimized. In addition, the SDH multi-service platform can provide any of the ports, two or three-layer business, regardless of the physical interface type. With the increase in data traffic in the network, the SDH multi-service platform is evolving and developing new generation systems that support data services to more flexibly and effectively support data services. The latest development is to support GFP, LCAS, RPR, and ASON standards. The main disadvantage of this solution is based on synchronous operation, the jitter is strict, and the equipment cost is high. Second, this solution is difficult to flexibly generate business. Further, the use of SDH fixed frame length and time slot to support the bandwidth efficiency of burst data services, and the current data service function is still not flexible enough. Finally, it is difficult to manage multiple network-oriented and unconnected networks, and management cost is still high. In the long run, this solution is not a most effective way after the data service is the absolute leading business type of the network. 6 SDH multi-service platform in embedded RPR functionality In order to extend the Ethernet to the core network of the electric telecommunications, the IEEE 802.17 elastic packet ring (RPR) is one of the solutions. This is a two-layer protocol, which is independent of the following layer technology and the three-layer technology, which belongs to intermediate layer enhancement technology. It uses a new MAC layer and shared access method to deliver the IP package through the new MAC layer. Entering a layer of data intra or bare fiber, there is no need to make a split restructuring, and the exchange processing power is improved, and performance and flexibility are improved. RPR is not a separate structure that can work both from SDH or Gbit / S Ethernet, or directly working on the bare fiber as a router's line interface board. Early independent RPR equipment main architecture on Ethernet, current trend is architectural in SDH, becoming a new generation of MSTP's embedded smart layer function, so that it can make full use of the advantages of both. If the bottom layer uses the SDH frame format, the TDM service and data traffic can be mapped into different SDH time slots. The TDM service high quality requirements continues to be used along the SDH environmental protection, and the data service that is less high for quality requirements may not participate in SDH. Environmental protection, improve bandwidth utilization. RPR simplifies the packet processing process. Do not allow the business stream to make the business stream in each node in the network, implement queuing, plasticity, and processing, and can directly turn the non-landing IP package directly, obvious Improve the exchange of exchange processing, the best group business. On the other hand, RPR can ensure that the service quality of the circuit exchange service and the line service can implement 50 ms protection switching time on the second floor, so it is allowed to run according to different services, and there is no need to protect the reservation bandwidth. RPR has automatic topology discovery capabilities, using a trigger protocol similar to OSPF but working on the second layer, can automatically identify changes in the two-layer topology, enhance self-healing, support plug, to avoid manual configuration Time-consuming powerful erroneous problems.
