|
|
The following paragraphs describe the Cisco WAN switches (nodes) used to build cell-based Wide Area Networks: the ATM broadband BPX Service Node, the multiband IGX, and the narrowband IPX. For further information, refer to the BPX, IGX, and IPX reference manuals. Also, for further information on associated products, refer to the Release 4.1 Cisco MGX 8220 Installation and Configuration and the Release 4.1 Cisco MGX 8220 Command Reference publications, and to the FastPAD Reference publication. For Network Management Information, refer to the Cisco WAN Manager Operations publication.
This chapter includes the following:
BXM card sets provide either service or trunk interfaces. The BXM card sets are compliant with ATM UNI 3.1 and Traffic Management 4.0 standards including ABR VS/VD and provide high scalability with up 144 T3/E3 ports, up to 96 OC3c/STM-1 ports, or up to 24 OC-12c/STM-4 ATM ports.
The BPX provides the following features:
With the BCC-4, the BPX employs a redundant 19.2 Gbps non-blocking crosspoint switch matrix for cell switching. The switch matrix can establish up to 20 million point-to-point connections per second between ports. A single BPX provides twelve card slots, with each card capable of operating at 800 Mbps for ASI and BNI cards. The BXM cards support egress at up to 1600 Mbps and ingress at up to 800 Mbps. Access to and from the crosspoint switch is through multi-port network and user access cards.
An arbiter is utilized to set up and release each cell path based on those ports with traffic data to switch. The arbiter can set up 20 million point-to-point connections per second between the BPX switch ports. For maximum reliability, the switch matrix, data paths, and all control and arbiter functions are provisioned in a fail-safe 1:1 redundant configuration.
Each BPX node is configured with twelve slots (switch ports) for network or user interfaces and three slots for common equipment cards.
The BXM cards may be configured to operate in either port (UNI) mode for connection to CPE or in trunk mode for connection to other BPX nodes or networks. The BXM-T3/E3, BXM-155, and BXM-622 provide T3, E3, OC3/STM-1, and OC12/STM-4 interfaces, respectively.
The Broadband Network Interface (BNI) card provides high-speed DS3, E3, or OC-3/STM-1 ATM network interfaces for trunking to other BPX nodes, to IPX nodes equipped with an ATM trunking cards or to other networks. The BNI high-speed trunk cards can be configured for 1:1 redundancy.
An ATM Service Interface (ASI) card provides two DS3, E3, or OC-3/STM-1 ATM ports for connection to ATM customer premise equipment.
The BPX is a virtual circuit switch. It can switch on VCI and/or VPI fields in the ATM cell header. Individual connections, as well as groups of connections within a path, may be switched. This greatly enhances the versatility of the BPX as a network switch.
Three techniques are used for transmitting user information over ATM cells.
The BPX may be used to provide a smooth migration path for network operators currently using IPX and IGX nodes in their networks who need additional capacity and would like to build on their existing equipment. Since the BPX can be integrated into an existing IPX/IGX network by adding only a single card to each of the IPX or IGX nodes that need to connect to the BPX, the hardware costs are kept to a minimum.
The ATM cell payload is ideally suited to transporting IPX FastPackets since both are fixed length and the size of the ATM cell is larger than the FastPacket cell length (no segmentation of the cell is necessarily required). Packets of data that are created by IPX and IGX nodes in a mixed environment are encapsulated in the ATM cells for transmission throughout the network.
For more details on the BPX hardware, refer to the BPX Reference. For information on the BPX commands, refer to the Command Reference.
The Extended Services Processor (ESP) is an adjunct shelf that provides ATM and frame relay SVC signaling to the BPX Service Node. The ESP also provides PNNI routing management for the ATM and frame relay SVCs. For further information about the ESP, refer to the BPX Service Node Extended Service Processor document.
The Cisco MGX 8220 is a standards-based ATM interface shelf for the BPX that a provides a service interface to multi-service networks and is usually co-located with a BPX. The Cisco MGX 8220 allows users to economically concentrate large numbers of PVC connections over high-speed ATM trunks. Release 4 of the Cisco MGX 8220 provides T1/E1 and subrate frame relay, T1/E1 ATM, and FUNI (frame based UNI over ATM), circuit emulation, inverse ATM multiplexing, and frame relay service interfaces for routing traffic over the ATM network via the BPX. For additional information on the Cisco MGX 8220, refer to the Cisco MGX 8220 Installation and Configuration document.
The BPX consists of the following elements:
ATM connections can be made between user devices equipped with DS3, E3, OC3, and OC12 ATM interfaces. These user devices connect directly to a BPX switch port over an ATM User-Network Interface on a BXM card (service access mode) or on the ATM Service Interface (ASI) card and through Cisco MGX 8220 or IPX shelves. The BPX supports constant bit rate (CBR), variable bit rate (VBR), available bit rate (ABR), and unspecified bit rate (UBR) ATM connections.
Functions performed by the BXM (trunk mode) and BNI modules include queuing and servicing cells, implementing congestion control mechanisms, address translation and routing of cells to the switching matrix. Physical connection to the network is accomplished through simple line interface cards.
Segmentation and reassembly, cell switching, interfacing to the switch matrix, and system timing and status monitoring are performed by the BPX controller card. A non-blocking, crosspoint switching matrix in the BCC card is the heart of the BPX. With the BCC-3, which employs a 16 x 16 crosspoint switch, the BPX operates at 9.6 Gbps. With the BCC-4, which employs a 16 x 32 crosspoint switch, the BPX can operate at 19.6 Gbps when it is also equipped with BXM cards. A switch arbiter is used to determine which matrix input port needs to connect to which matrix output port to properly route the cells. This individual crosspoint switching technique provides superior performance at broadband speeds when compared with a system bus architecture.
Each BPX switch port has a capacity of operating at 800 Mbps and is capable of supporting OC-12 data rates in each of the twelve switch ports on the BPX
All data flow through the switch is monitored at every process to detect any corruption of the data or addressing. Flow-through parity recomputation and checking is performed within the individual cards and data paths on the backplane to detect data path and memory faults that would otherwise be difficult to detect with periodic checks.
The controller, switch matrix, backplane data paths, and all control and timing signals are redundant throughout the BPX switch to provide maximum reliability. Both Broadband Controller Cards (BCCs) are synchronized to each other and data paths are buffered so any switchover will be transparent.
Cisco's IGX product family, consists of the IGX 8, IGX 16, and IGX 32 multiservice ATM switches. As multi-band members of the Cisco WAN Switching ATM portfolio, IGXs seamlessly integrate with Cisco's IPX, BPX, Cisco MGX 8220, INS, and FastPAD platforms under StrataSphere management, to provide multiband ATM solutions from the access to the core layer with integrated network management and call processing.
The IGX 16 and IGX 32 are network backbone node systems for large sites with multiple trunks and considerable local traffic requirements, where a large number of physical ports and gigabit-scale throughput are required. Both IGX systems are available as stand-alone units or can be rack mounted with other equipment.
All IGX services are supported by standard ATM narrowband and broadband trunk resources. IGX systems support trunk speeds from 128 Kbps to T3/E3 and FastPAD access trunk connectivity from 9.6 Kbps to 2 Mbps. IGX systems can be fully interconnected in a logical mesh through public ATM services, or provide edge switching in and out of such services, and can network to FastPAD systems over public frame relay services at up to 2 Mbps. IGX systems also support the IPX FastPacket trunk protocol for seamless connectivity with IPXs at smaller sites.
The IGX uses a 1.2 Gbps cell switching redundant bus to pass ATM cells between optionally redundant adaptation, trunking and gateway modules within the system. This architecture allows any amount of bandwidth to be assigned to any slot, and makes the IGX the only system in its class with more than 16 slots, for greater scalability.
Hardware, firmware and software are designed for maximum availability, non-stop networking, even during maintenance windows. Availability design features, common to all Cisco WAN switching systems, include:
All switches use a mid-plane design with front cards performing processing functions and back cards providing interfacing and physical connectivity. This allows most system maintenance to be performed at the front cards, without disconnecting interface cables.
The following are some of the features provided by the IPX narrowband switch:
The IPX family consists of different configurations designed to meet the needs of small, medium, and large sites. There are three basic IPX packaging options: the IPX 8, IPX 16, and IPX 32. All three systems use virtually identical hardware and software; they differ only in the number of card slots provided. The IPX 8 extends the benefits of FastPacket technology to the edge of networks without losing bandwidth efficiency.
 |
Note The term IPX is used to refer to the generic system. Where necessary to differentiate between the various systems, the terms IPX 8, IPX 16, or IPX 32 are used. |
The IPX also uses voice compression to provide increased capacity. Voice interfaces, such as PABX's and channel banks, transmit Pulse Code Modulation (PCM) voice streams to the IPX for processing. The IPX processes voice streams using Voice Activity Detection VAD). With VAD, voice packets are transmitted only when speech is present (in typical phone conversations, speech is present only 40% of the time). This results in greater than 2-to-1 compression, with no degradation in voice quality. This compression ratio allows many more voice channels to share a single trunk.
In addition to Pulse Code Modulation (PCM), the IPX provides various levels of Adaptive Differential Pulse Code Modulation (ADPCM) to give another 2, 3, or 4-to-1 compression to voice. In this way, the IPX provides voice compression for as many as 256 (E1) or 192 (T1) voice channels on a single trunk.
The IPX uses Data Frame Multiplexing (DFM) to provide low-speed data compression. DFM is implemented using Repetitive Pattern Suppression (RPS). If data packets contain a repetitive data pattern, the IPX will suppress sending the packets across the network and merely regenerate the pattern at the far end. The net result of this suppression is a significant savings in bandwidth.
Non-disruptive diagnostics continuously monitor the performance of each component in the node. In the event of a failure, a backup component is automatically switched into service with no effect on the user. All cards and power supplies are hot-replaceable; they can be added or replaced on-line with no interruption in service.
IPX systems can be configured to operate in network applications at any world-wide network location. Options include interfaces to both North American and International standard trunks and ports. Power supply options include 208 VAC, 240 VAC, and 48 VDC modules. All cards are programmable, with downloadable firmware, to operate with various network parameters and are easily upgraded to take advantage of new features as they are developed.
For more details on the IPX hardware, refer to the IPX Reference Manual. For information on the IPX commands, refer to the Command Reference Manual.
Information enters an IPX from attached terminal equipment in the form of digital streams, through a circuit mode service interface, or in the form of LAP-D frames through a packet mode service interface. These interfaces segments the streams or frames into cells, addresses the cells, and transmits them onto a switching bus.
Each trunk interface monitors the system bus and receives any cells to be transmitted through it. Queuing of cells for transmission is done by the trunk interface. Cells that are being switched through an intermediate node are transmitted onto the system bus by the receiving trunk interface and picked up by another trunk for transmission out. When cells reach the destination node, they are put on the bus by the receiving trunk, picked up by the appropriate service interface, converted back to data streams or frames and transmitted to the attached user equipment.
All information entering an IPX, whether circuit or packet, streaming or framed, constant or variable bit rate, is converted to a single common multimedia format, a fixed-sized cell. These cells are switched and transmitted to their destination, where the original information is reconstructed and delivered. The benefits of IPX FastPacket networks arise from the power and flexibility of the universal switching architecture.
Posted: Tue Jan 16 11:03:00 PST 2001
All contents are Copyright © 1992--2001 Cisco Systems, Inc. All rights reserved.
Important Notices and Privacy Statement.