Circuit Emulation Services and Voice over ATM

This chapter provides an overview of using circuit emulation services (CES) for connecting the ATM switch router and traditional time-division multiplexing (TDM) devices. This chapter also includes a description of the Simple Gateway Control Protocol (SGCP), a software feature that allows your ATM switch router to function as a call gateway in a voice over ATM environment.

Note The information in this chapter is applicable to the Catalyst 8540 MSR, Catalyst 8510 MSR, and LightStream 1010 ATM switch routers. For detailed configuration information, refer to the ATM Switch Router Software Configuration Guide and the ATM Switch Router Command Reference publication.

This chapter includes the following sections:

Circuit Emulation Services Overview

Real-time data, such as voice and video, have low tolerance for transfer delays. For voice, it is generally <50 ms; for video, it is <150 ms; delays >50 ms usually require echo cancellation. The constant bit rate (CBR) traffic category provides the low delay characteristics required for such traffic. Using CBR permanent virtual connections (PVCs) or soft PVCs, circuit emulation extends T1/E1 services carrying real-time data directly from traditional TDM devices across the ATM cloud.

Typical applications for CES include the following:

Interconnecting PBXs, TDM equipment, and video equipment over an ATM network (typically a WAN).
Merging traffic from TDM sources with other data over the ATM backbone.
Using an ATM backbone to transport real-time data instead of a leased line.

Figure 9-1 provides an example of CES applications in an ATM network. TDM devices that do not have ATM interfaces—multiplexers, PBXs, and video codecs—are directly attached to the ATM network, which transports the data over its infrastructure as CBR traffic.

Figure 9-1: T1/E1 Unstructured CES Applications in ATM Switch Router Network

The T1 and E1 CES Interfaces

There are two types of T1 and E1 port adapters available for the ATM switch router with different framing characteristics at the physical level. The T1 and E1 ATM port adapters send and receive ATM cells and must be connected to an external host that can accept native ATM. The T1 and E1 CES port adapters, however, connect TDM equipment across an ATM network using CBR ATM virtual connections.

Features and Functionality

The CES modules provide three key functions:

Circuit Emulation Services Interworking Function (CES-IWF)—enables communication between CBR and ATM User-Network Interface (UNI) interfaces.
T1/E1 CES unstructured services—also called "clear channel," allows entire T1 or E1 interfaces to be emulated across an ATM backbone.
T1/E1 CES structured services—also called "channelized T1," allows mapping of one or multiple DS0 (64 kbps) channels across an ATM cloud.

Figure 9-2 shows an example of using the CES modules in an ATM network for both unstructured and structured services.

Figure 9-2: Circuit Emulation Services Supported by CES Modules

CES-IWF

CES-IWF is based on an ATM Forum standard that allows communication between CBR and ATM UNI interfaces, that is, between non-ATM telephony devices (such as classic PBXs or TDMs) and ATM devices (such as an ATM switch or router with an ATM interface). CES-IWF works by packaging incoming native T1 or E1 frames into AAL1 cells at the ingress, and performing the opposite function at the egress.

The CES-IWF provided by the ATM switch router allows migration from interconnecting T1/E1 CBR data communications services over separate leased lines to interconnecting those services over the same ATM cloud that carries data traffic.

Figure 9-3 illustrates the use of CES-IWF between non-ATM devices (traditional PBXs) and other end devices. In the case of communication between two non-ATM end devices (traditional PBXs), CBR data in native T1 format received from an edge device on one side of the network is segmented into ATM cells and propagated through the ATM network. After traversing the network, the ATM cells are reassembled into a CBR bit stream that matches the original user data. This native T1 CBR data is then passed out of the network to the edge device at the destination endpoint. In the case of communication between the traditional PBX and the ATM data service unit (DSU), it is the DSU that performs the reassembly into the CBR bit stream for its attached PBX. Both cases illustrate the use of CES-IWF.

Figure 9-3: CES-IWF Functions in an ATM Switch Router Network

Unstructured CES

Unstructured CES in an ATM network emulates point-to-point connections over T1 or E1 leased lines. This service maps the entire bandwidth necessary for a T1 or E1 leased line connection across the ATM network. Unstructured CES operations do not decode or alter the CBR data in any way.

The CES port adapters provide the following unstructured services:

CES T1 port adapter—Supports DSX-1 physical interfaces to provide T1 unstructured (clear channel) CBR data transmission services at 1.544 Mbps.

: Using a CES T1 port adapter for unstructured CES emulates a point-to-point T1 leased line across your ATM network.

CES E1 port adapter (either 120-ohm or 75-ohm version)—Supports G.703 physical interfaces to provide E1 unstructured (clear channel) CBR data transmission services at 2.048 Mbps.

: Similarly, using a CES E1 port adapter for unstructured CES emulates a point-to-point E1 leased line across your ATM network.

Figure 9-4 shows how T1/E1 unstructured CES might be used to connect PBXs with an ATM switch router equipped with a CES T1 or E1 port adapter.

Figure 9-4: T1/E1 Unstructured CES Across Leased Lines

Structured CES

Structured CES is designed to emulate point-to-point fractional T1 (N x 64 kbps) connections. N x 64 refers to a circuit bandwidth (data transmission speed) provided by the aggregation of N x 64-kbps channels. The 64-kbps data rate, or the DS0 channel, is the basic building block of the T carrier systems (T1, T2, and T3).

With T1/E1 structured CES, networks can be simplified by eliminating TDM devices and allocating T1/E1 bandwidth to PBXs and teleconferencing equipment. In addition, the Simple Gateway Control Protocol (SGCP) can be used to control structured CES circuits for voice over ATM. See the "Simple Gateway Control Protocol" section.

The CES modules provide the following structured services:

CES T1 port adapter—Supports DSX-1 physical interfaces to provide T1 channelized data transmission services at a rate of 1.544 Mbps.

: Using a CES T1 port adapter, you can map one or more digital signal level 0 (DS0) channels to an ATM virtual circuit to be connected across an ATM network. Each T1 port has up to 24 DS0 time slots per T1 port for allocation to structured CES circuits. Each time slot can transmit CBR data at a rate of 64 kbps. This represents a total CBR data transmission capacity of 1.536 Mbps (24 x 64 kbps).

CES E1 port adapter—Supports G.703 physical interfaces to provide E1 channelized data transmission services at a rate of 2.048 Mbps.

: Using either the 120- or 75-ohm version of a CES E1 port adapter, you can map a single DS0 channel (64 kbps) or multiple DS0 channels across an ATM network. Each E1 port has up to
31 DS0 available time slots for allocation to structured CES circuits. Each time slot can transmit CBR data at a rate of 64 kbps. This represents a total CBR data transmission capacity of
1.984 Mbps (31 x 64 kbps).

Note With channel-associated signaling enabled, the effective data transfer rate of the circuit is limited to 56 kbps. The "Channel-Associated Signaling and On-Hook Detection for Structured CES" section describes the CAS mechanism.

By supporting T1/E1 structured CES, the CES module can function in the same way as a classic digital access and crossconnect system (DACS) switch. Figure 9-5 illustrates the digital crossconnect and channelized mapping functions supported by an ATM switch router equipped with a CES module.

Figure 9-5: DACS Function of CES Modules

Either single or multiple DS0 time slots can be mapped across the ATM network. Each time slot (or DS0 channel) can transmit CBR data at a rate of 64 kbps. Note also that multiple N x 64 circuits can be connected to a single port, using separate time slots.

In Figure 9-6, for example, structured CES allows DS0 timeslots to be combined into circuits and transported using ATM PVCs. The PVCs can be routed to many different destination CES interfaces. Similarly, circuits from many different CES interfaces can be interconnected to a single CES interface, where the various circuit DS0 timeslots are interleaved to form an outgoing T1 bit stream. Thus, you can combine structured CBR data in a highly flexible way for transport across an ATM network.

Figure 9-6 illustrates how 24 available N x 64 DS0 time slots in a CES T1 port adapter can be combined in a number of ways to accomplish structured CBR data transport in an ATM network.

Figure 9-6: Time Slots for Structured Services in a CES T1 Port Adapter

Note that the ingress (source) DS0 channels at one end of the CES circuit can be mapped into different egress (destination) DS0 channels at the other end of the CES circuit. Mapping DS0 channels requires that the total number of time slots mapped at each end of the CES circuit match.

In Figure 9-6, for example, time slots 7, 8, and 24 are bundled to form a single 192-kbps circuit. At the other end of the connection, you can bundle any of three (available and different) DS0 time slots (such as 18, 19, and 20) to complete the CES circuit.

Note You can group DS0 channels as contiguous or noncontiguous time slots. In Figure 9-6,
time slots DS0 7, DS0 8, and DS0 24 are configured to create one structured circuit.

Figure 9-7 illustrates how 31 available N x 64 DS0 time slots can be provided for structured CES in a CES E1 port adapter. The rule for DS0 time slot allocation with a CES T1 port adapter also applies to the CES E1 port adapter: the specific DS0 time slot numbers assigned at one end of the circuit in a CES E1 port adapter do not need to map identically to the DS0 time slot numbers at the other end of the CES circuit. Only the aggregate number of DS0 time slots at each end of the circuit must agree.

Figure 9-7: Time Slots for Structured Services in CES E1 Port Adapter

Channel-Associated Signaling and On-Hook Detection for Structured CES

Since the CES T1 and E1 port adapters emulate CBR services over ATM networks, they must be able to support channel-associated signaling (CAS) information introduced into structured CES circuits by PBXs and TDMs. The optional CAS feature for the CES T1 and E1 port adapters meets this requirement. CAS, also called robbed bit signaling, uses 8k out of each 64k channel to carry signaling information, leaving 56k for the voice channel.

The alternative to CAS, Common Channel Signaling (CCS), in which one entire 64k channel is used for signaling, is not directly supported on the CES T1 and E1 port adapters. However, the Cisco VSC 2700 signaling controller, in conjunction with SGCP can provide similar functionality. See the "Simple Gateway Control Protocol" section.

A second feature, on-hook detection, allows the bandwidth of a quiet circuit to be used by other virtual connections, based upon the CAS. This feature frees unused CBR bandwidth for other preexisting ABR or UBR circuits.

These features can be configured for structured CES in the following ways:

CAS not enabled (the default state)

: In this case, the CES module does not sense the CAS information (carried as ABCD bits in the CBR bit stream) and does not provide support for CAS functions.

CAS enabled without on-hook detection

: In addition to packaging incoming CBR data into ATM adaptation layer 1 (AAL1) cells in the usual manner for transport through the network, the CES module in the ingress ATM switch router (see Figure 9-8) senses the ABCD bit patterns in the incoming data, incorporates these patterns in the ATM cell stream, and propagates the cells to the next node in the network. The ATM cells flow across the network from link to link until reaching the egress ATM switch router node.
: At the egress node, the CES module strips off the ABCD bit patterns carried by the ATM cells, reassembles the CAS ABCD bits and the user's CBR data into their original form, and passes the frames out of the ATM network on the proper DS0 time slot.

Note All these processes occur transparently.

CAS and on-hook detection enabled

The CAS and on-hook detection features work together to allow an ingress node in an ATM network to monitor on-hook and off-hook conditions for a specified 1 x 64 structured CES circuit. As implied by 1 x 64, the on-hook detection (or bandwidth-release) feature is supported only in a structured CES circuit with a single DS0 time slot at each end of the connection, as shown in Figure 9-8.

For structured CES, you can invoke CAS with the ability to detect on-hook or off-hook conditions for any given structured CES circuit. The hook state indicates the following:

On-hook—Circuit is idle or unconnected.
Off-hook—Circuit is in use and connected.

The CAS mechanism allows dynamically allocated T1/E1 bandwidth and is released by hard PVCs or soft PVCs configured for structured CES.

When you configure CAS, the ingress CES module monitors the ABCD bits in the incoming CBR bit stream to detect on-hook and off-hook conditions in the circuit. In an off-hook condition, all the bandwidth provided for the specified CES circuit is used to transport ATM AAL1 cells across the network from the ingress node to the egress node.

Conversely, in an on-hook condition, the network periodically sends dummy ATM cells from the ingress node to the egress node to maintain the connection. However, these dummy cells consume only a fraction of the circuit's reserved bandwidth, leaving the rest of the bandwidth available for other network traffic. This bandwidth-release feature enables the network to make more efficient use of its resources by making unused bandwidth available to bursting or oversubscribed traffic. However, the released bandwidth cannot be reserved by other virtual connections.

Note The on-hook detection feature requires customer premises equipment (CPE) that supports on-hook.

Figure 9-8: CAS in a Structured CES Circuit

Enabling the CAS feature for a CES circuit limits the bandwidth of the DS0 channel to 56 kbps for user data, since CAS functions consume 8 kbps of channel bandwidth for transporting the ABCD signaling bits. These signaling bits are passed transparently from the ingress node to the egress node as part of the ATM AAL1 cell stream.

In summary, when you enable the optional CAS and on-hook detection features, the following conditions apply:

The permanent virtual channel connection (PVCC) provided for the CES circuit always exists.
The bandwidth for the CES circuit is always reserved.
During an on-hook state, most of the bandwidth reserved for the CES circuit is not in use. (Dummy cells are occasionally sent from the ingress node to the egress node to maintain the connection.) This unused bandwidth is therefore available when other network traffic, such as ABR traffic, bursts or is oversubscribed.
During an off-hook state, all the bandwidth reserved for the CES circuit is dedicated to that circuit.

Advantages

Potential advantages of using CES in your ATM network include the following:

Native T1 handoff allows you to connect directly to a PBX or other TDM device without that device knowing anything about ATM.
Interconnecting PBXs and TDM devices over an ATM network allows you to take advantage of increased bandwidth, save on leased line costs, and reduce the amount of equipment, such as DSUs and MUXs, required for separate infrastructures.

Limitations

Potential limitations of CES include the following:

Like traditional TDM, CBR virtual connections used by CES do not offer the statistical multiplexing benefits of other ATM services.

Clocking, if not properly configured, can cause problems, particularly when there are multiple clocking sources.

Network Clocking for CES and CBR Traffic

For your CES environment to function properly, clocking must be carefully set up. Clock sources and their priorities, along with a distribution mode, must be properly configured, as described in "Network Clock Synchronization."

The CES port adapters are capable of using three clocking modes to meet the timing requirements of CBR data:

Synchronous—the default clocking mode for CES. If you have a single PRS, such as a clock signal from a telephone company, you should use synchronous mode. If you are using structured CES, you must use synchronous mode.
Synchronous residual time stamp (SRTS)—allows equipment at the edges of a network to use a clocking signal that is different (and completely independent) from the clocking signal being used in the ATM network. SRTS mode also allows two CPEs to have different clocks.
Adaptive—typically used when it is not possible to implement either synchronous or SRTS mode. This is the least precise and least recommended method.

Table 9-1 summarizes, in order of preference, the characteristics of the three clocking modes you can configure on a CES module.

Table 9-1: Characteristics of CES Clocking Modes

Clocking Mode Advantages Limitations

Synchronous

Supports both unstructured (clear channel) and structured CBR traffic.

Exhibits superior control of wander and jitter.

Requires a PRS and network clock synchronization services.

Ties the CES interface to the network clock synchronization services clocking signal (PRS).

SRTS

Conveys externally generated user clocking signal through an ATM network, providing an independent clocking signal for each CES circuit.

Requires a PRS and network clock synchronization services.

Supports only unstructured (clear channel) CBR traffic.

Exhibits moderate control of wander and jitter.

Adaptive

Does not require a PRS or network clock synchronization services.

Supports only unstructured (clear channel) CBR traffic.

Exhibits poorest control of wander and jitter.

Clocking Mode	Advantages	Limitations
Synchronous	Supports both unstructured (clear channel) and structured CBR traffic. Exhibits superior control of wander and jitter.	Requires a PRS and network clock synchronization services. Ties the CES interface to the network clock synchronization services clocking signal (PRS).
SRTS	Conveys externally generated user clocking signal through an ATM network, providing an independent clocking signal for each CES circuit.	Requires a PRS and network clock synchronization services. Supports only unstructured (clear channel) CBR traffic. Exhibits moderate control of wander and jitter.
Adaptive	Does not require a PRS or network clock synchronization services.	Supports only unstructured (clear channel) CBR traffic. Exhibits poorest control of wander and jitter.

Although the wander and jitter characteristics of these clocking modes differ, all clocking modes preserve the integrity of the your CBR data, ensuring error-free data transport from source to destination. The differences among the three modes are further described in the following sections.

Synchronous Clocking

Synchronous clocking mode is the only one that supports full CES functionality. SRTS and adaptive clocking do not support structured CES. In addition, synchronous clocking is typically used in public telephony systems, making a precision reference signal readily and widely available for synchronizing CBR data transport. With synchronous clocking mode, every device must get its clocking from a single PRS, such as a PBX connecting to a public telephone network.

Figure 9-9, for example, shows how a PRS for synchronous clocking can be provided to an edge node of an ATM network and propagated through the network to synchronize the flow of CBR data between the communicating ATM end nodes.

In this network scenario, a PRS is available to the network by the PBX at the edge of the network. The PRS is present at the port of a CES module in edge node A (the ingress node). From there, the PRS is propagated into the first ATM network through an ATM port and conveyed across an OC-3 trunk to an adjacent ATM network. This same clocking signal is then used to synchronize the handling of CBR data in edge node B (the egress node).

Figure 9-9: Synchronous Clocking in an ATM Switch Router Network

Configuration Overview

Since synchronous clocking is the default for CES, you do not need to perform any per-interface configuration when using this mode. This assumes that you have one PRS and have properly configured your network clocking, as described in "Network Clock Synchronization."

SRTS Clocking

Synchronous residual time stamp (SRTS) clocking mode is typically used when there are multiple different clock sources; for example, a when you already have your ATM network synchronized to a reference source, then add a PBX that receives its clock from another service provider source. If you want to avoid resynchronizing your network to their source, SRTS provides a solution by reconciling the different timing signals in the course of CBR data transport through the ATM network.

A common scenario for using SRTS clocking is when your edge equipment is driven by a different clocking signal than that being used in the ATM network. For example, user equipment at the edges of the network can be driven by clock A, while the devices within the ATM network are being driven by clock B. Figure 9-10 shows such an operating scenario, in which a timing signal is provided to edge nodes independently from the ATM network.

Figure 9-10: SRTS Clocking in an ATM Switch Router Network

Using Figure 9-10, assume that the user of edge node 1 wants to send CBR data to a user at edge
node 3. In this scenario, SRTS clocking works as follows:

1. Clock A is driving the devices within the ATM network.

2. At edge node 1, the user introduces CBR traffic into the ATM network according to clock B.

3. As edge node 1 segments the CBR bit stream into ATM cells, it measures the difference between user clock B, which drives it, and network clock A.

4. As edge node 1 generates the ATM cell stream, it incorporates this delta value into every eighth cell.

5. The cells then propagate through the network in the usual manner.

6. As destination edge node 3 receives the cells, this node not only reassembles the ATM cells into the original CBR bit stream, but also reconciles, or reconstructs, the user clock B timing signal from the delta value carried within every eighth ATM cell.

Thus, during SRTS clocking, CBR traffic is synchronized between the ingress (segmentation) side of the CES circuit and the egress (reassembly) side of the circuit according to user clock signal B, while the ATM network continues to function according to clock A.

Configuration Overview

Configuring SRTS clocking for CES requires the following steps:

Step 1 Configure the global clocking as described in "Network Clock Synchronization."

Step 2 From interface configuration mode, enable SRTS clocking mode on the CES interfaces.

Adaptive Clocking

The name adaptive clocking mode reflects the fact that the rate at which CBR data is propagated through an ATM network is driven essentially by the rate at which CBR data is introduced into the network by the user's edge equipment. The actual rate of CBR data flow through the network might vary from time to time during adaptive clocking, depending on how rapidly (or how slowly) CBR data is being introduced into the network. Nevertheless, CBR data transport through the network occurs in a "pseudo synchronous" manner that ensures the integrity of the data.

Adaptive clocking requires neither the network clock synchronization service nor a global PRS for effective handling of CBR traffic. Rather than using a clocking signal to convey CBR traffic through an ATM network, adaptive clocking in a CES module infers appropriate timing for data transport by calculating an "average" data rate upon arrival and conveying that data to the output port of the module at an equivalent rate.

For example, if CBR data is arriving at a CES module at a rate of so many bits per second, then that rate is used, in effect, to govern the flow of CBR data through the network. What happens behind the scenes, however, is that the CES module automatically calculates the average data rate using microcode (firmware) built into the board. This calculation occurs dynamically as user data traverses the network.

When the CES module senses that its segmentation and reassembly (SAR) buffer is filling up, it increases the rate of the transmit (TX) clock for its output port, thereby draining the buffer at a rate that is consistent with the rate of data arrival.

Similarly, the CES module slows down the transmit clock of its output port if it senses that the buffer is being drained faster than CBR data is being received. Adaptive clocking attempts to minimize wide excursions in SAR buffer loading, while at the same time providing an effective means of propagating CBR traffic through the network.

Relative to the other clocking modes, implementing adaptive clocking is simple and straightforward. It does not require network clock synchronization services, a PRS, or the advance planning typically associated with developing a logical network timing map. However, adaptive clocking does not support structured CES, and it exhibits relatively high wander characteristics.

Configuration Overview

Unlike synchronous or SRTS modes, configuring adaptive clocking mode for CES does not require selection of clocking sources, priorities, and distribution mode. You must only enable adaptive clocking from interface configuration mode on each of the CES interfaces.

CES Configurations

This section provides some general guidelines and considerations when configuring CES connections. This section also includes examples of various types of unstructured and structured service connections you can configure for CES.

Before You Begin

Before you begin configuring physical interfaces and virtual circuits for CES operation, you should be aware of the information you need and the associated tasks:

Network clocking

: You must have clocking properly configured for CES operations to be successful. Configuring synchronous clocking is described in "Network Clock Synchronization." Special clocking cases for CES are described in the "Network Clocking for CES and CBR Traffic" section in this chapter.

Service type

: You must decide whether to use unstructured or structured services. See the "Unstructured CES" section and the "Structured CES" section for descriptions.

Physical level characteristics:
- Framing type
- Line buildout
- Line code type
- Loopback test method

: Refer to your ATM switch router software documentation for details on configuring these and other physical level parameters.

Cell Delay Variation (CDV)

: CDV can be critical due to the delay-sensitive nature of CBR data; refer to the "About Cell Delay Variation" section.

PVC (hard, soft)

: You must decide whether to use hard or soft PVCs. Manual configuration is required for hard PVCs, while soft PVCs are set up through signaling. Also, soft PVCs can reroute in the event of failure, while hard PVCs cannot. For more information, see the "General Procedure for Creating Soft PVCCs for CES" section.

About Cell Delay Variation

Cell delay variation (CDV) refers to the distortion caused by change in interarrival times between cells, also known as jitter, measured in microseconds.

Each end-to-end CES circuit exhibits delay characteristics, based on the following factors:

The delay characteristics of the individual devices participating in the CES circuit.

: Each network device contributes some increment of delay, reflecting the unique electrical characteristics of that device.

The number of intermediate hops through which the CBR data must pass in traversing the network from source to destination.
The type and speed of the trunk lines interconnecting the ATM networks.
The volume of traffic being handled by the trunk lines at any given time; that is, the degree to which the network is experiencing congestion conditions.

The network designer or administrator calculates a CDV value for each hop in the data path as a means of establishing a maximum allowable CDV value for the network as a whole. To some degree, the network's maximum allowable CDV value is a measure of the network's expected performance. By establishing this CDV threshold for the network, appropriate buffer sizing can be derived for the network devices involved in any given CES circuit, ensuring that the network operates as expected.

In a CES module, for example, the maximum allowable CDV value for the network is used to determine an appropriate size (depth) for the SAR buffer built into the board. This sizing of the SAR buffer is done to prevent buffer overflow or underflow conditions. An overflow condition can cause a loss of frames, while an underflow condition can cause frames to be repeated.

The actual CDV value for a circuit varies according to the particular data path used for the circuit. Consequently, the depth of the SAR buffer increases or decreases in proportion to the CDV value for the CES circuit being set up.

Tips You can issue the CLI show ces circuit interface command in an unstructured circuit to measure the current CDV value. In many cases the configured default value is satisfactory.

For an unstructured hard PVC, the CDV value for the circuit (including all hops) must not exceed a maximum allowable CDV value.

For an unstructured soft PVC, the network automatically determines the best data path through the network and handles the routing of CBR traffic. The network accomplishes this task dynamically through the ATM connection admission control (CAC) mechanism. The CAC mechanism determines the best path through the network by executing a routing algorithm that consults local routing tables in network devices.

If the requested data path is equal to or less than the maximum allowable CDV value established by the network administrator, the connection request is granted. If the requested CES circuit exceeds the maximum allowable CDV value, the connection request is denied.

For example, when a user requests a connection from source node A at one edge of the network to destination node B at the opposite edge of the network, the CAC mechanism accounts for the CDV value for each hop in the requested connection to determine a suitable path through the network that does not exceed the network's maximum allowable CDV value.

General Procedure for Creating Soft PVCCs for CES

You can create either hard PVCCs or soft PVCCs, depending on your particular CES application requirements, for use in your CES operations. This section provides a general procedure for configuring soft PVCCs for CES.

The following steps must be performed in the prescribed order when you configure soft PVCCs for either unstructured or structured CES:

Note The steps in these guidelines assume that you have already configured circuits on the CES interfaces. If you have not, you will not see addresses in some displays, or the output will be null.

Step 1 Determine which CES interfaces are currently configured in your ATM switch router chassis. The show ces status command displays this information for you.

Step 2 Determine which two ports you want to define as participants in the soft PVCC.

Step 3 Decide which of the two ports you want to designate as the destination (or passive) side of the soft PVCC.

Note This is an arbitrary decision—you can choose either port as the destination end of the circuit. However, you must decide which port is to function in this capacity and proceed accordingly.

Step 4 Configure the destination (passive) side of the soft PVCC.

You must configure the destination end of the soft PVCC first, as this end defines an ATM Forum-compliant CES-IWF ATM address for that port.

Step 5 Retrieve the CES-IWF ATM address of the soft PVCC's destination end. You can use the show ces address command to display the CES-IWF ATM addresses.

You must determine this address, as well as the VPI/VCI values for the circuit (see Step 6), and use these elements as part of the command string when you configure the source (active) end of the soft PVCC (see Step 8).

Step 6 Retrieve the VPI/VCI values for the circuit. You can use the show ces circuit command to display the VPI/VCI values.

Step 7 Shut down the interface.

Step 8 Configure the source (active) end of the soft PVCC last, using the information derived from Step 5 and Step 6.

You must configure the source end of the soft PVCC last, because that end not only defines the configuration information for the source port, but also requires you to enter the CES-IWF ATM address and VPI/VCI values for the destination port.

Step 9 Reenable the interface.

Note The soft PVC route optimization feature is not supported for CBR data.

T1/E1 Unstructured CES

This section provides an overview of the procedures for configuring CES modules for unstructured CES.

The circuit you set up on a CBR port for unstructured service is always identified as circuit 0, since you can establish only one unstructured circuit on any given CBR port. An unstructured circuit uses the entire bandwidth of a T1 or E1 port, as follows:

A hard PVCC on a T1 or E1 port—At least the entire bandwidth of a T1 port (1.544 Mbps) or an E1 port (2.048 Mbps) is allocated to the soft PVCC along its path as long as the soft PVCC is up. A network failure along the path of the soft PVCC relinquishes this bandwidth until the connection can be set up again.
A soft PVCC on a T1 or E1 port—The entire bandwidth of a T1 port (1.544 Mbps) or an E1 port (2.048 Mbps) is also allocated manually when you set up the circuit, but this bandwidth is only used by the unstructured circuit when it is active.

Hard PVCCs for Unstructured Services

A CES module converts CBR traffic into ATM cells for propagation through an ATM network. CBR traffic arriving on a given CES module port must first be segmented into ATM cells. This cell stream is then directed to an outgoing ATM port or CBR port using a PVCC.

Note As a general rule when configuring a hard PVCC, you must cross-connect a CBR port and an ATM port in the same ATM switch router chassis.

Figure 9-11 shows an example of using a hard PVCC to connect ATM and CES interface modules for unstructured CES on the ATM switch router.

Figure 9-11: Hard PVCC Configured for Unstructured CES

Configuration Overview

Configuring a hard PVCC, such as the one shown in Figure 9-11, requires the following steps:

Step 1 Display the current CES and ATM interface information. Use this information to identify the source CBR and destination ATM interfaces.

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 clock mode.

For descriptions of the clocking modes, see the "Network Clocking for CES and CBR Traffic" section.

If you are using synchronous or SRTS clocking mode, you must first configure the global clocking distribution mode for the chassis and the clock source on the interface; you do not need to perform these steps if you are using adaptive clocking. For more information on configuring clock sources, see "Network Clock Synchronization."

Step 4 Configure the CES AAL1 service as unstructured (the default).

Step 5 If needed, configure the line coding, framing, and line buildout.

Step 6 Configure the CES interface circuit identifier and specify a circuit name.

Step 7 Configure the hard PVCC cross-connect to the ATM interface with VPI/VCI values.

Soft PVCCs for Unstructured Services

In a soft PVCC, as well as a hard PVCC, you configure both ends of the CES circuit. However, a soft PVCC typically involves CES modules at opposite edges of an ATM network, so a soft PVCC can be set up between any two CES modules anywhere in your network.

For guidelines when configuring soft PVCCs, see the "General Procedure for Creating Soft PVCCs for CES" section.

The destination address of a soft PVCC can point to either of the following:

A port in the same ATM switch router chassis
A port in any other CES module in the network

Note For simplicity, the procedure in this section assumes that you are creating a soft PVCC between interfaces in the same ATM switch router chassis. In a typical scenario, you would configure a soft PVCC to connect CES interfaces on ATM switch routers at opposite ends of a network, where the CES interfaces are attached to PBXs or other TDM devices.

Figure 9-12 shows a logical representation of the soft PVCC used in the following example procedure.

Figure 9-12: Soft PVCC Configured for Unstructured CES

Configuration Overview

Configuring a soft PVCC for unstructured CES is a two-phase process:

Phase 1—Configure the destination (passive) side of the soft PVCC.
Phase 2—Configure the source (active) side of the soft PVCC.

Configuring the destination side of the soft PVCC, as shown in Figure 9-12, requires the following steps:

Step 1 Display the current CES interface information. Use this information to identify the destination CBR interface.

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 clock mode.

For descriptions of the clocking modes, see the "Network Clocking for CES and CBR Traffic" section.

Step 4 Configure the CES AAL1 service as unstructured (the default).

Step 5 If needed, configure the line coding, framing, and line buildout.

Step 6 Configure the CES interface circuit identifier and specify a circuit name.

Configuring the source side of the soft PVCC requires the following steps:

Step 1 Display the CES interface information for the destination side. Use this information to identify the ATM address and VPI/VCI values to use when configuring the source side of the soft PVCC.

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 service as unstructured (the default).

Step 4 If needed, configure the line coding, framing, and line buildout.

Step 5 Configure the CES interface circuit identifier and specify a circuit name.

Step 6 Configure the soft PVCC to the destination CES-IWF ATM address and VPI/VCI of the circuit.

T1/E1 Structured CES

This section provides an overview of the procedures you use when configuring CES modules for structured (N x 64 kbps) CES.

An important distinction between structured and unstructured CES is that structured CES allows you to allocate the entire T1/E1 bandwidth. Structured CES only uses the T1/E1 bandwidth actually required to support the active structured circuit(s) you configure. For example, configuring a CES module for structured services allows you to define multiple hard PVCCs or soft PVCCs for any CES T1 or E1 port.

In both module types, any bits not available for structured CES are used for framing and out-of-band control.

Note Structured CES requires synchronous clocking mode. See the "Network Clocking for CES and CBR Traffic" section.

Hard PVCCs for Structured Services without CAS

This section describes configuring a hard PVCC for structured CES without channel-associated signaling (CAS).

Figure 9-13 illustrates a hard PVCC for a structured CES connection configured with the following parameters:

Four time slots (DS0 channels 1 to 3, and 7) are configured for a circuit named CBR-PVC-A.
CAS is not used.
ATM port 0/0/0 in the ATM switch router chassis is designated as the destination port of the hard PVCC.

Figure 9-13: Hard PVCC Configured for Structured CES

Configuration Overview

Configuring a hard PVCC for structured CES without CAS requires the following tasks:

Step 1 Display the current CES and ATM interface information. Use this information to identify the source CBR and destination ATM interfaces.

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 service as structured.

When the AAL1 service is configured as structured, clocking mode is synchronous by default. No other clocking configuration for CES is required, assuming that you have properly configured the global clocking distribution mode for the chassis and the clock source on the interface (network-derived by default). For more information on configuring clock sources, see "Network Clock Synchronization."

Step 4 If needed, configure the line coding, framing, and line buildout.

Step 5 Configure the CES interface circuit identifier and list of T1 time slots that comprise the circuit, and specify a name for the circuit.

Step 6 Configure the hard PVCC to the ATM interface with VPI/VCI values.

Hard PVCCs for Structured Services through a VP Tunnel

A common application of CES is tunneling through a public network. Using VP tunnels to connect end systems, you can send CBR data using structured or unstructured CES over long distances without the cost of leased lines or long distance telephone charges. Figure 9-14 shows an example of this application for CES.

Figure 9-14: CES over VP Tunnel

The VP tunnel type for this application is commonly shaped or hierarchical. For a full description of the types of VP tunnels and their restrictions, see the "VP Tunnels" section of the chapter "Virtual Connections."

Configuration Overview

Configuring a hard PVCC for structured CES through a VP tunnel requires the following steps:

Step 1 Configure the VP tunnel on the ATM interface, as described in the "VP Tunnels" section in the chapter "Virtual Connections."

Step 2 Configure the hard PVCC that connects the CBR and ATM interfaces, as described in the "Hard PVCCs for Structured Services without CAS" section. You could also connect the two ends through the VP tunnel using soft PVCCs, as described in the following section, "Soft PVCCs for Structured Services without CAS."

Soft PVCCs for Structured Services without CAS

This section describes the procedures used to configure a soft PVCC for structured service based on the following assumptions, as illustrated in Figure 9-15:

Four time slots (DS0 channels) are configured for the soft PVCC, as follows:
- For circuit CBR-PVC-A: DS0 channels 1 to 3 and 7 are used on port CBR0/1/0
- For circuit CBR-PVC-B: DS0 channels 10 to 13 are used on port CBR0/1/1
Channel-associated signaling (CAS) is not used.
The source (active) side of the soft PVCC is named CBR-PVC-A.
The destination (passive) side of the soft PVCC is named CBR-PVC-B.

Figure 9-15: Soft PVCC Configured for Structured CES (without CAS)

Configuration Overview

Configuring a soft PVCC for structured CES without CAS is a two-phase process:

Phase 1—Configure the destination (passive) side of a soft PVCC.
Phase 2—Configure the source (active) side of a soft PVCC.

Configuring the destination side of the soft PVCC requires the following steps:

Step 1 Display the current CES interface information. Use this information to identify the destination CBR interface.

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 service as structured.

Step 4 If needed, configure the line coding, framing, and line buildout.

Step 5 Configure the CES interface circuit identifier and list of T1 time slots that comprise the circuit, and specify a name for the circuit.

Configuring the source side of a soft PVCC requires the following steps:

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 service as structured.

Step 4 If needed, configure the line coding, framing, and line buildout.

Step 5 Configure the CES interface circuit identifier and list of T1 time slots that comprise the circuit, and specify a name for the circuit.

Step 6 Configure the soft PVCC to the destination CES-IWF ATM address with VPI/VCI values.

Soft PVCCs for Structured Services with CAS

The procedures in this section build on the configuration information in the "Soft PVCCs for Structured Services without CAS" section. However, this procedure enables channel associated signaling (CAS) for the soft PVCC.

The following procedure is based on the following assumptions, as illustrated in Figure 9-16:

One time slot is configured for the soft PVCC:
- For circuit CBR-PVC-A: DS0 channel 1 on port CBR0/1/0
- For circuit CBR-PVC-B: DS0 channel 10 on port CBR0/1/1
CAS is enabled for the circuit.
The signaling mode for the CBR ports is set to robbedbit.

Figure 9-16: Soft PVCC Configured for Structured CES with CAS

Configuration Overview

Configuring the destination side of the soft PVCC requires the following steps:

Step 1 Display the current CES interface information. Use this information to identify the destination CBR interface.

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 service as structured.

Step 4 If needed, configure the line coding, framing, and line buildout.

Step 5 Configure the DSX1 signal mode to robbedbit.

Step 6 Configure the CES interface circuit identifier and list of T1 time slots that comprise the circuit, and specify a name for the circuit. Specify CAS when configuring the circuit.

Configuring the source soft PVCC requires the following steps:

Step 2 From global configuration mode, select the CBR interface to configure and enter interface configuration mode.

Step 3 Configure the CES AAL1 service as structured.

Step 4 If needed, configure the line coding, framing, and line buildout.

Step 5 Configure the CES interface circuit identifier and list of T1 time slots that comprise the circuit, and specify a name for the circuit. Specify CAS when configuring the circuit.

Step 6 Configure the soft PVCC to the destination CES-IWF ATM address with VPI/VCI values.

Soft PVCCs for Structured Services with CAS and On-Hook Detection Enabled

This section outlines the additional steps that you must take to activate the on-hook detection (bandwidth-release) feature in a 1 x 64 structured soft PVCC CES circuit. For a description of on-hook detection, see the "Channel-Associated Signaling and On-Hook Detection for Structured CES" section.

Configuration Overview

Configuring a soft PVCC for structured services with CAS and on-hook detection requires the following steps:

Step 1 Configure the soft PVCC with CAS enabled, as described in the previous section, "Soft PVCCs for Structured Services with CAS."

Step 2 Specify on-hook detection in addition to CAS when configuring the circuit.

Multiple Soft PVCCs on the Same CES Port

The procedures in this section describe creating more than one structured service PVCC on the same T1 or E1 port. Figure 9-17 illustrates how you can configure multiple CES circuits on a single T1 or E1 port.

The following assumptions apply to creating multiple soft PVCCs on the same T1 or E1 port, as illustrated in Figure 9-17:

The source (active) side of a soft PVCC named CBR-PVC-A is already created on port CBR0/1/0.
The destination (passive) side of a soft PVCC named CBR-PVC-B is already created on port CBR0/1/1.
A new source (active) side of a soft PVCC named CBR-PVC-AC is created on port CBR0/1/0 of the CES module, thereby creating a multiple CES circuit on this particular port.
A new destination (passive) side of a soft PVCC named CBR-PVC-CA is created on port CBR0/1/2 of the CES module.

Figure 9-17: Configuring Multiple Soft PVCCs on the Same T1 or E1 Port

Configuration Overview

Configuring the additional soft PVCCs, as shown in Figure 9-17, requires the following steps:

Step 1 On the destination side, configure the parameters for the additional CBR interface and configure the circuit and timeslots. The configuration steps in the "Soft PVCCs for Structured Services without CAS" section or in the "Soft PVCCs for Structured Services with CAS" section show how to do this configuration.

Step 2 On the source side, perform the following steps:

a. Configure the additional circuit and time slots on the existing CBR interface.

b. Configure the soft PVCC to cross-connect the two circuits.

Simple Gateway Control Protocol

The Simple Gateway Control Protocol (SGCP) running on the ATM switch router, in combination with the Cisco VSC 2700, a virtual switch controller, provides switched voice on demand within a campus or metropolitan-area network. Designed to support common channel signaling protocols, this solution can eliminate the need for expensive tandem switches when interconnecting multiple PBXs.

Networks that include multiple PBXs require a tandem switch, which inter-PBX calls must traverse. To accommodate multiple PBXs, they might all be connected to the tandem switch. Or, as in Figure 9-18, the PBXs might be connected through the ATM network using CES interfaces and clear-channel T1 or E1 links. All inter-PBX calls are required to traverse the tandem switch. This scenario works well where the number of PBXs to be connected is low. As the size of the campus or MAN grow, however, the size of the required tandem PBX also grows, as does the number of primary rate interfaces (PRIs) required.

Figure 9-18: PBXs Connected using Common Channel Signaling and CES

In Figure 9-19 the virtual switch controller (VSC) is connected to an ATM switch router. SGCP, running on both the VSC and the ATM switch router, carries VSC instructions to set up and tear down connections based on the signaling between the VSC and the PBXs. When a call must be established between any two PBXs, the VSC instructs the ATM switch router to provision a soft PVC (64 kbps CBR) between the appropriate two endpoints, providing bandwidth on demand. This reduces the total number of interfaces required and eliminates the need for a tandem PBX.

Figure 9-19: PBXs Connected using the VSC

Additional Advantages

In addition to potentially eliminating the need for the tandem PBX, the VSC and SGCP solution provides the following advantages:

Reduces the total interfaces required for PBXs.
Provides bandwidth on demand for devices capable of using common channel signaling over a PRI interface.
Allows toll bypass where a MAN crosses a local access and transport area (LATA) boundary.

How It Works

SGCP controls Voice over IP gateways by an external call control element (called a call-agent). This feature has been adapted to provide control of ATM switch router CES circuits (called endpoints in SGCP). The resulting system (call-agents and gateways) allows for the call-agent to engage in common channel signaling (CCS) over a 64-kbps CES circuit, governing the interconnection of bearer channels on the CES interface. In this system, the ATM switch router acts as a Voice over ATM gateway.

Figure 9-20 illustrates how 64-kbps CCS channels on the CES T1 and E1 ports are backhauled or carried to the VSC 2700 units.

Figure 9-20: Common Channeling Signaling over a CES Circuit

A single trunk circuit sigA (on ATM switch router A's CES port) carries or backhauls the CCS control call setup for the port. Trunk circuit sigB also controls a similar port on ATM switch router B. SigA is backhauled over the ATM network to ATM switch router C by a CES soft PVC to a circuit on another CES card port directly attached to a call-agent (VSC 2700 (1)). Similarly, sigB is backhauled to a CES circuit on ATM switch router D.

Both call-agents are configured to handle backhauled signaling circuits to the CES trunk circuits. When a call-setup request is received on sigA by VSC 2700 (1), the VSC 2700 (1) cooperates with VSC 2700 (2) to establish the connection.

To dynamically connect the CES circuits located on ATM switch routers A and B, the call-agents use SGCP to allocate the CES circuits on each switch and then establishes a soft PVC between the switches. The resulting connection is callAB. Call-agents use SGCP to cause the ATM switch routers to set up and delete end-to-end connections between circuits.

Configuration Overview

Configuring SGCP requires the following steps:

Step 1 Enable SGCP on the switch.

SGCP is disabled by default.

Step 2 Configure CES circuits for SGCP.

Enable structured AAL1 service on the CES interface and allocate the time slot to a circuit identifier.

Step 3 Configure SGCP request handling.

Modify the default timeout and retry intervals for SGCP request handling, as needed.

Step 4 Configure call-agent address.

Specify the address of the call agent SGCP is to use.

Table of Contents

Circuit Emulation Services and Voice over ATM