ZTE´s MSTP Solution

ZTE´s MSTP solution includes the implementation of RPR and the MPLS protocol to fulfill the automatic end-to-end switching functions. Using brand new design and hardware and software technologies, the solution provides Ethernet, Storage Area Network (SAN) and Plesiochronous Digital Hierarchy (PDH) services by adding new function boards on the Next Generation (NG) SDH hardware system.

    ZTE´s MSTP Equipment includes:

• ZXMP S320, the MSTP for the access layer network
• ZXMP S330, the MSTP for the access and aggregation layer network
• ZXMP S380 and ZXMP S390, the MSTPs for the aggregation and core layer network
• ZXMP S385, the MSTP for the core layer network

More information at:
http://www.zte.com.cn/English/03product/list1.jsp?CateID=191

1 Multi-Service Transport  Platform (MSTP)
The MSTP is a multi-service transport network with unified protocols, which connects government agencies, enterprises, educational institutes, companies and family users. It supports multimedia applications and offers data services and integrated services such as the packet voice, images and video.

    Horizontally, the optical transport network is divided into the access layer, convergence layer and the core layer. The access layer implements access of diversified users. The convergence layer controls the convergence of distributed access points, implements data multiplexing, switches data, and provides traffic flow control and user management.The core layer completes the interaction of high-speed information over the entire network, and interconnects with the backbone network.

    Vertically, the optical transport can be divided into the service layer and the transmission layer. The service layer consists of Asynchronous Transfer Mode (ATM) and IP equipment. That may implement such services as the
narrow-band voice services, Internet services, remote calculating and transaction processing, e-commerce, videoconference, video-telephone, integrated multimedia information services, remote computer communication and control, and the line lease. The transmission layer refers to the metro optical network that carries these services, providing a unified transmission platform with high efficiency, high capacity and low cost for all services. It consists of Dense Wavelength Division Multiplexing (DWDM) and Synchronous Digital Hierarchy (SDH) equipment. Figure 1 shows the basic network architecture based on MSTP in the view of the services.

1.1 Technologies for MSTP
(1) Generic Framing Procedure (GFP)
    The GFP defined by ITU-T as Recommendation G.7041/Y.1303 is the first encapsulation mechanism capable of addressing the wide range of data transport applications by supporting a suite of client network protocols that are summarized in Table 1. The GFP provides a flexible and efficient mapping of various protocols onto a transport.
In addition to the common aspects of the GFP protocol, client-specific functions are required to handle unique differences in protocol mapping. This includes options specific to the two types of client data mapping, the
frame-mapping (GFP-F) and the transparent-mapping (GFP-T). Table 1 lists all of the GFP-F and GFP-T protocols supported in the G.7041 specification.

(2) Virtual Concatenation (VCAT) and Link Capacity Adjustment Scheme (LCAS).
    The VCAT and LCAS technologies are used for transport of synchronous storage area network and network attached storage protocols such as Enterprise Systems Connection (ESCON), Fiber Channel, High Performance Parallel Interface (HiPPI). Figure 2 demonstrates the principle of CAT and VCAT.

    Running Ethernet over the Time Division Multiplexing (TDM) is one way to leverage existing infrastructure and improve the business model. Although this scheme improves the bottom line, Ethernet over TDM has some disadvantages. The first is that Ethernet do not map efficiently into commonly supported TDM payloads. In addition, when custom-size payloads are used at the edge, other inefficiencies can arise in the core of the network.

    Furthermore, the TDM, unlike many other data protocols, requires circuits to be reprovisioned for bandwidth changes. Therefore, flexible data protocols running over TDM networks are slaved to TDM´s inefficiencies. The VCAT is defined to provide a standards-based protocol that enables Ethernet and any other traffic to be efficiently mapped to TDM payloads. It combines several disconnected paths together to transmit concatenation-level services like 622M POS interface or FE/GE interface. The LCAS, defined by ITU-T as G.7042, efficiently handles dynamic traffic patterns that change over time. It makes on-demand, hitless, bandwidth changes a reality. These additions to TDM protocols will better allow service providers to leverage their existing transport infrastructure, and thus improve their business models.

    The VCAT notation for SDH is VC-X-nv, where X is the size of the noncontiguous virtual channel fragments that will be used to transport the entire Virtual Connection Gateway (VCG). The value of n is the total number of X fragments that it takes to make up the total VCG. The "v" indicates that this is a VCAT payload.

    The VCAT only needs implementation in the path-terminating devices because resequencing and the indication of multiframing is performed via the H4 byte, a path overhead field. Path overhead is only used at the source and destination of the TDM flow.

    Changing the customer´s bandwidth profile is always an issue. You have to take something that works, change it, and make sure it works again-without anyone noticing. The aim of LCAS is to make the bandwidth change a simpler and safer task. LCAS provides the control mechanism for the "hitless" increasing or decreasing of the capacity in a VCG link to meet the bandwidth needs of the application. It also provides the capability to temporarily remove member links that have experienced a failure. The LCAS assumes that in cases of capacity initiation, increase, or decrease, the modification of the end-to-end path of each individual VCG member is the responsibility of the network and element management systems. Therefore, LCAS provides a mechanism for bandwidth reprovisioning without being controlling the time and reason of the operation.

1.2 Main Features of MSTP
MSTP has capabilities to transmit multiple services. It has a series of features that are different from other networks, such as:

• The MSTP is a uniform network in a multi-user oriented environment, which requires a certain level of Quality of Service assurance.
• It supports various signals from all client layers, and rapidly provides the bandwidth required by these signals.
• It is a multi-service transport network that supports data, voice and image services. It covers SDH, PDH, SAN, Ethernet, fiber level and ATM services access capabilities. Therefore, it supports ATM/SDH/PDH/SAN/ETH/Lamda level service granularity.
• It is the development trend of the transmission network.
  

    The MSTP integrates more functions in one system than the classical SDH plus data service system does. In additon, it is smaller and low in power consumption. Contrasting with the classical SDH plus data equipment, the MSTP means smaller equipment room, lower power consumption, more convenient Network Management System (NMS), less expenses and better QoS or SLA capability.

1.3 Multi Protocol Label Switch (MPLS)
The MPLS helps implement the automatic end-to-end service creation. Therefore, the manual configuration is not necessary any longer.

    The MPLS gives network operators a great deal of flexibility to divert and route traffic around link failures, congestion and bottlenecks. From the QoS standpoint, Internet Service Providers (ISPs) will be able to better manage different data streams based on priority and service plan. For instance, subscribers of a premium service plan, or those who receive lots of streaming media or high-bandwidth content can see minimal latency and packet loss.

    When the packets enter an MPLS-based network, Label Edge Routers (LERs) assign each of them a label (identifier). These labels not only contain the information based on the routing table entry (i.e., the destination, bandwidth, delay, and other metrics), but also refer to the IP header field (the source IP address), to the Layer 4 socket number information, and to the differentiated services. Once this classification is complete and mapped, different packets are assigned to the corresponding Labeled Switch Paths (LSPs). At these LSPs, the Label Switch Routers (LSRs) place outgoing labels on the packets.

    With these LSPs, network operators are able to divert and route traffic based on data-stream type and
Internet-access customer.

    The following are the principles for the Ethernet over the MPLS:

• The Ethernet or Virtual Local Area Network (VLAN) and the VC label　are together formed as the Pseudo-Wire (PW or Virtual Circuit).
• The Multiple PWs and the external layer tunnel label multiplexes it as one LSP port.
  

    These principles are demonstrated in Figure 3. The Ethernet services  accord to the customer domain, check Media Access Control (MAC) of the destination, and then decide the forwarding Tunnel LSP port. When the Ethernet services go through the intermediate node (P), the Tunnel Label is switched, but the VC label keeps unchanged.

    The embedded MPLS data plane maps the Ethernet over MPLS traffic into STM-N traffic in four steps, as shown in Figure 4.

(1) The logical port and the LAN port are regarded as the Ethernet switch port. The Ethernet port is divided into the Vrtual Bridges (VBs). The VBs are segregated from each other, and the traffic inspection implements the bit rate limit of users´ access.

(2) According to the VPN ID, a VC label is assigned to the Ethernet service and then the PWs/ are multiplexed according to the destination node, the QoS and the Tunnel Label.

(3) The switch accords the furthest external Tunnel Label. It supports the Replace, Push and POP action of the label, and realizes QoS of the congestion avoidance scheme.

(4) Different VCG realizes circuit segregation and strict QoS. All LSPs share the bandwidth of the VCG, are logically segregated from each other, and provide E2E QoS assurance.

1.4 Resilient Packet Ring (RPR)  Technology
The RPR technology provides MSTP more choices. It is an emerging standard that enables the efficient transmission of data traffic over the SDH ring infrastructure, while continuing to leverage the network for TDM services. The RPR technology combines the low cost and simplicity of packet-based connectionless networking with the reliability, bandwidth, and scalability of optical networks. The result is the best of both worlds—a resilient,
packet-oriented, ring-based solution that provides virtual mesh networking connectivity. Figure 5 shows the architecture of the RPR over SDH.

    The RPR technology is developed for carrier networks with rapidly growing data traffic. Currently, data networks in the metro areas must be transported through TDM circuits, like SDH. These circuits are based on point-to-point, fixed bandwidth connections most appropriate for voice and constant bit rate services. An addressable, connectionless network for bursty traffic, best handles data communications and RPR if it is optimized for this traffic type running over optical networks in the metro and wide area.

    The RPR is optimized as a high-availability protocol for transporting data, packet video and voice over the ring topologies, while it provides sub-50 millisecond protection switching. It is intended for carrier applications in the Metropolitan Area Network (MAN) and WAN, but can also provide the reliable and high-speed connectivity to the campus and data center.

2 ZTE´s MSTP Solution
Numerous vendors and operators show strong interests on MSTP since it is viewed as a solution heading to Next Generation Network (NGN). ZTE has also put great efforts on the research of MSTP. ZTE´s MSTP solution includes the implementation of RPR and the MPLS protocol to fulfill the automatic end-to-end switching functions. Using brand new design and hardware and software technologies, the solution provides Ethernet, Storage Area Network (SAN) and Plesiochronous Digital Hierarchy (PDH) services by adding new function boards on the Next Generation SDH hardware system. Figure 6 shows the MSTP target applications with ZTE´s MSTP equipment, ZXMP.

    The MSTP solutions include the single-wavelength solution and the multi-wavelength solution.

    The former is mainly applied to the convergence layer and the access layer of the metro area optical network. It also supports data service transmission and processing, and the traditional SDH functions, such as the Ethernet processing, ATM processing, MPLS services and RPR services.

    Based on the Wavelength Division Multiplexing (WDM) technologies, the latter is mainly applied to the backbone layer and the convergence layer of the optical network. Generally, the topology structure of the backbone layer is a ring type since the services are scattered and there are numerous add/drop services on the way. In this case, Optical
Add-Drop Multiplexer (OADM) equipment is adopted in most of the Optical Transport Networks (OTNs). The protection can be directly implemented in the optical layer due to its ring topology. The wavelength cross-connection and the add/drop function can be configured.

2.1 ZTE´s MSTP Equipment
(1) ZXMP S320, the MSTP for the access layer network, as shown in Figure 7.

• Compact structure:
  It has a compact structure with the height of 4U, convenient and compatible with the rate of 155/622 Mb/s.
• Wide temperature adaptability:
  Its working temperature is from
  -20 °C to 60 °C, therefore it is suited for various environments.
• More measures are used to ensure the reliability:
  High reliability is guaranteed by the power supply, clock, cross connect 1+1 and tributary 1:N protection.
• Full service interfaces:
  It supports the 5 STM-1 equivalent access capacity and 16×16 VC4 full cross connection, the 1.5 M/2 M/34 M/45 M/155 M interface with added AI/DI (auxiliary), the10/100 M Ethernet interface, and the 155 M ATM UNI.
• More choices for working voltage and easy installation:
  The working voltage is -48/24 V. For installation, wall, desk and rack mounting can be selected.
  

(2) ZXMP S330, the MSTP for the access and aggregation layer network, as shown  in Figure 8.

• Compact structure:
  It has a compact structure with the height of 10U, and the rates of 155 M/622 M/2.5 G.
• More measures are used to ensure the reliability:
  High reliability is guaranteed by the clock, cross-connect 1+1 backup, distributed power supply and tributary
  1:N  protection for four different interfaces.
• Full service interfaces:
  ZXMP S330 supports the 13.5 G access capacity and 120×120/104×104 VC4 full cross connection. It also has the 1.5 M/2 M/34 M/45 M/155 M/622 M/2.5 G interfaces, the 10/100 M Ethernet interfaces, the 155 M ATM UNI, as well as the RPR function with the 10 M/FE/GE interfaces.
• Easy installation:
  For installation, desk and rack mounting can be selected.
  

(3) ZXMP S380 and ZXMP S390, the MSTPs for the aggregation and core layer network, as shown in Figure 9.

• High reliability:
  High reliability of ZXMP S380 and S390 is implemented by the dual-bus, intelligent fan with infinitely variable speeds, and by the safe and reliable double-system of power distribution.
• Full service interfaces:
  ZXMP S380 supports the 30 G access capacity and 256×256 VC4 cross connection. ZXMP S390 supports and the 140 G access capacity and 1 024×1 024 VC4 cross connection (ZXMP S390). S380 and S390 also have the 1.5 M/2 M/34 M/45 M/155 M/622 M/2.5 G/10 G interfaces, the 10 M/FE/GE Ethernet interfaces, the 155 M ATM UNI, and the RPR function with the 10 M/FE/GE interfaces.
• The initiated logic subnet protection:
  The initiated logic sub-equipment and subnet protection technique provides the four-fiber bi-direction multiplex section sharing protection ring, and the dual-node interconnection (DNI) protection.
  

(4) ZXMP S385, the MSTP for the core layer network, as shown in Figure 10.

• Full service interfaces:
  It supports the 35 G access capacity and 256×256 VC4 cross connection (ZXMP S385), and the 140 G access capacity and 1 024×1 024 VC4 cross connection (ZXMP S395). It also has the 1.5 M/2 M/34 M/45 M/155 M/622 M/2.5 G/10 G interface, the 10 M/FE/GE Ethernet interface, and the RPR function with the 10 M/FE/GE interface.
• Smooth upgrade:
  It can be upgraded to 10 G smoothly.
• High reliability:
  High reliability is implemented by the dual-bus, intelligent fan with infinitely variable speeds, the safe and reliable double-system of power distribution, and 1:N  tributary protection.
• The initiated logic subnet protection:
  The initiated logic sub-equipment and subnet protection technique provides the four-fiber bi-direction multiplex section sharing protection ring, and the DNI protection.

Manuscript received: 2005-02-09