Fax, Modem, and Text Support over IP Configuration Guide, Cisco IOS Release 15M&T
Fax and Modem Services over IP Overview
Fax and Modem Services over IP Overview
Last Updated: December 4, 2012
This application guide includes descriptions and configuration instructions for fax and modem transmission capabilities on Cisco Voice over IP (VoIP) networks. It is written for developers and network administrators who are installing, configuring, and maintaining fax and modem applications on Cisco voice gateways.
Finding Feature Information
Your software release may not support all the features documented in this module. For the latest caveats and feature information, see Bug Search Tool and the release notes for your platform and software release. To find information about the features documented in this module, and to see a list of the releases in which each feature is supported, see the feature information table at the end of this module.
Use Cisco Feature Navigator to find information about platform support and Cisco software image support. To access Cisco Feature Navigator, go to www.cisco.com/go/cfn. An account on Cisco.com is not required.
Information About Cisco IOS Fax Services over IP
Fax Transmission in the PSTN
Facsimile (fax) transmission is the sending of an image, drawing, or document over a distance by converting it into coded electrical signals at the originating end, passing the signals from the originator to the receiver over a transmission medium, and converting the signals into a replica of the original at the receiving end.
When sending a fax, a fax machine uses a scanner to convert the paper image into digital bits, a single-chip microprocessor called a digital signal processor (DSP) to reduce the number of bits, and a modem to convert the bits into an analog signal for transmission over an analog dial-up phone line.
When receiving a fax, the fax machine uses its modem and printer to convert the incoming bits into black and white images on paper.
The information conveyed in a fax transmission consists of both protocol (control information, capabilities, identification) and document content . The document content consists primarily of the document image plus additional metadata that accompanies the image. The means by which an image of a document is encoded within the fax content is the image data representation .
When a fax has been sent successfully, the sender receives a confirmation that indicates that the fax content was delivered. This confirmation is an internal signal and is not normally visible to the sending user, although some error messages are visible to allow a page to be resent.
The ability to send the representation of a page to a remote location developed over a number of years. The first images were sent over wires as early as 1843, but modern fax machines did not start appearing in offices until the 1960s. At that time, a single-page letter took about six minutes to send over public phone lines using the new Group 1 standard for transmission that was introduced by the International Telegraph and Telephone Consultative Committee (CCITT) in 1968. The Group 2 standard, introduced in 1976, reduced the time to send a page to three minutes, but still could not provide transmission at a dense enough resolution for the clear reproduction of small print. In 1980, the Group 3 standard was introduced. The Group 3 standard improved fax scanning resolution and introduced digital transmission techniques to enable transmission rates of 14400 bits per second (bps). Group 3 fax machines are the most common today by far. Group 4 is a standard for digital phone lines such as ISDN, and it operates at 64 kbps. Each standard specifies special tones that identify calls as fax calls and enable handshaking to define fax capabilities at both ends of the call. All of the fax standards have evolved with a goal of sending more data faster over the public switched telephone network (PSTN).
The PSTN is composed of switched time-division multiplexing (TDM) circuits, which are either single lines or trunks. A line connects a single telephony device to a switch, whereas a trunk connects a switch to a switch. The network provides exclusive and full use of a circuit between two endpoints and is full-duplex (simultaneous transmission in both directions), unless the call is data. Trunks are one of the following types:
Both circuit types have sufficient audio clarity, or dynamic range, to pass the tones required to send fax traffic across PSTN circuits.
Fax traffic consists of digital data modulated onto high-frequency carrier tones. There are various ways to modulate this information, such as Amplitude Modulation (AM), Frequency Modulation (FM) or Frequency Shift Keying (FSK), and Phase Modulation (PM) or Phase Shift Keying (PSK). In order to get higher bit rates (more information) across the same carrier circuit, these modulation techniques are often combined into forms of modulation called Quadrature Amplitude Modulation (QAM) or Trellis-Coded modulation.
Data Transmission Standards
The international standards that define data transmission techniques are used by both fax and modem transmission devices. The main difference between them is that modem payload originates as digital data, whereas fax payload is a paper image that has been encoded into a digital data stream. Another key difference is the initial handshaking that determines the facsimile or data capabilities of each party in the transmission.
There are standards that apply to both fax and modem machines and standards that apply to only fax machines, defining methods by which faxes are encoded and sent.
The traditional facsimile transmission standard, also called Group 3 (G3) fax, describes implementations of ITU-T T.30 and T.4. All Cisco IOS fax applications use T.30 and T.4 standards to interface with the PSTN or fax device.
For a comprehensive list of fax and modem standards, see the Standards,
Fax Transmission Phases
The T.30 specification is over 150 pages long, but a summary of its contents is provided in the following sections to provide some familiarity with the handshaking between calling and called parties and the basic procedures involved during fax transmission. The table below lists the five phases in a fax transmission.
Phase A Establishing a Voice Call
The call originator prepares a fax and dials a destination number. The destination fax device picks up the call. The originator and the destination are now connected in a voice call, but to transition to fax transmission one party must signal that it is a fax device. Either device can send its signal first, using one of the following methods:
Once these messages have been exchanged, the transaction can move to phase B.
Phase B Identifying Facilities and Capabilities
The following sequence of events identifies facilities and capabilities for fax transmission:
Phase C Transmitting Content
Phase C is referred to as the In-message Procedure. During this phase, high-speed T.4 page data is sent one line at a time. Each burst of line data is followed by an End Of Line (EOL) message. Because the EOL information is sent as T.4 data, it would not necessarily be seen in a T.30 trace. When the sending device has finished sending pages or wishes to return back to control mode, it sends 6 EOLs in a series that constitutes a Return To Control (RTC) message. The RTC message indicates the end of phase C, and the call progresses to phase D.
Phase D Signaling End of Transmission and Confirmation
After the T.4 transmission and the subsequent return to control mode, the sending device must send one of the following signals:
Phase E Releasing the Call
Following the fax transmission and the postmessage transactions, either the calling device or the called device can send a Disconnect (DCN) message, at which point the devices tear down the call, and the telephony call control layer releases the circuit. DCN messages do not require a response from the opposite device.
Fax Transmission over IP Networks
An IP, or packet-switched, network enables data to be sent in packets to remote locations. The data is assembled by a packet assembler/disassembler (PAD) into individual packets of data, involving a process of segmentation or subdivision of larger sets of data as specified by the native protocol of the sending device. Each packet has a unique identifier that makes it independent and has its own destination address. Because the packet is unique and independent, it can traverse the network in a stream of packets and use different routes. This fact has some implications for fax transmissions that use data packets rather than using an analog signal over a circuit-switched network.
Differences from Fax Transmission in the PSTN
Individual packets that are part of the same data transmission may follow different physical paths of varying lengths. They can also experience varying levels of propagation delay (latency) and delay that is caused by being held in packet buffers awaiting the availability of a subsequent circuit. The packets can also arrive in an order different from the order in which they entered the network. The destination node of the network uses the identifiers and addresses in the packet sequencing information to reassemble the packets into the correct sequence.
Fax transmissions are designed to operate across a 64-kbps, PCM-encoded voice circuit, but in packet networks the 64-kbps stream is often compressed into a much smaller data rate by passing it through a digital signal processor (DSP). The codecs normally used to compress a voice stream in DSPs are designed to compress and decompress human speech, not fax or modem tones. For this reason, faxes and modems are rarely used in a VoIP network without some kind of relay or pass-through mechanism in place.
Fax Services over IP Networks
There are two conceptual methods of carrying virtually real-time fax-machine-to-fax-machine communication across packet networks:
In addition to the methods for real-time fax transmission, a method called store-and-forward fax breaks the fax process into distinct sending and receiving processes and allows fax messages to be stored between those processes. store-and-forward fax is based on the ITU-T T.37 standard, and it also enables fax transmissions to be received from or delivered to computers rather than fax machines.
Cisco Fax Services
Some of the methods described in this section have different characteristics depending on the call control protocol used by the network, which may be H.323, Session Initiation Protocol (SIP), or Media Gateway Control Protocol (MGCP). Where the characteristics are different, they are noted.
Finding Support Information for Platforms and Cisco IOS Software Images
Use Cisco Feature Navigator to find information about platform support and Cisco IOS software image support. Access Cisco Feature Navigator at http://www.cisco.com/go/fn . You must have an account on Cisco.com. If you do not have an account or have forgotten your username or password, click Cancel at the login dialog box and follow the instructions that appear.
This section describes the following aspects of the fax services available on Cisco IOS gateways:
Concepts Related to Cisco Fax Services
The following concepts are useful in understanding how fax transmission methods are implemented on Cisco IP networks:
Voice Gateways and Dial Peers
A Cisco voice gateway provides an interface between the IP network and the public switched telephone network (PSTN) or telephony (fax) device. When a call comes into the IP network over a gateway, that gateway is called an originating gateway (OGW). Similarly, a gateway over which a call passes out of the IP network is called a terminating gateway (TGW).
A traditional voice call over the PSTN uses a dedicated 64-kbps circuit end to end. In contrast, a voice call over the packet network contains several discrete segments or call legs. A call leg is a logical connection between two routers or between a router and a telephony device. A voice call comprises four call legs, inbound into and outbound from both the OGW and the TGW.
Dial peers are software constructs that sort calls, route calls, and define characteristics applied to each call leg in the call connection, based on call source and destination endpoints. Dial peers are used for both inbound and outbound call legs. It is important to remember that these terms are defined from the perspective of the router. An inbound call leg is created by any call that comes in to a router, regardless of whether the router is an OGW or a TGW. An outbound call leg is created by any call that leaves a router, regardless of whether the router is an OGW or a TGW, as shown in the figure below.
Different types of dial peers handle different kinds of call legs. The following types of dial peers are used for fax over Cisco IP networks:
Tool Command Language (TCL) is used for scripts that direct interactive voice response (IVR) applications, which are used in Cisco voice networks for various purposes. IVR applications typically involve the real-time gathering of data from callers by means of digit collection and voice prompts. For example, you might have a debit card application that asks a user to enter a personal identification number (PIN) and then collects and verifies the digits that the user enters.
A gateway can have several IVR applications to accommodate different gateway services, and you can customize IVR applications to present different interfaces to various callers. IVR applications are used to implement the following fax services:
TCL scripts are provided on the Cisco Software Center website. You download them to a location that is accessible to the voice gateway that is running the fax application and then configure the gateway with the name and location of the script.
Quality of service (QoS) refers to the ability of a network--whether the network is a complex network, small corporate network, Internet service provider (ISP), or enterprise network--to provide better service to selected network traffic over various technologies, including Frame Relay, ATM, Ethernet and 802.1 networks, and SONET, as well as IP-routed networks that may use any or all of these underlying technologies.
The primary goals of QoS are to provide better and more predictable network service by providing dedicated bandwidth, controlled jitter and latency, and improved loss characteristics. QoS achieves these goals by providing tools for managing network congestion, shaping network traffic, using expensive wide-area links more efficiently, and setting traffic policies across the network.
QoS for fax transmissions means assuring that echo cancellation (EC) and voice activity detection (VAD), which are normally enabled for voice calls, are turned off as soon as a call is identified as a fax call. If EC and VAD are enabled, they can interfere with the successful reception of fax traffic.
The advantages of carrying fax over packet networks are reduced cost and saved bandwidth and are associated with QoS issues that are unique to packet networks. A major issue in the implementation of fax over IP networks is the problem of inaccurate timing of messages caused by delay through the network.
The delay of fax packets through a packet network causes the precise timing that is required for many portions of the fax protocol to be skewed and can result in the loss of the call. The fax-over-packet protocol in the interworking function must compensate for the loss of a fixed timing of messages over the packet network so that the T.30 protocol operates without error. Error Correction Mode (ECM) is enabled in the T.30 protocol.
An end-to-end fax over IP call is susceptible to the following sources of delay:
Delay issues are compounded by the need to remove jitter, which is the variable interpacket arrival time that is caused by conditions in the network that a packet traverses. An approach to removing the jitter is to collect packets and hold them long enough so that even the slowest packets arrive in time to be played in the correct sequence. This approach, however, causes additional delay. In most fax over IP methods, a time stamp is incorporated in the packet to ensure that packet data is played out at the proper instant.
The T.30 standard provides for ECM that allows a fax page to be broken into HDLC-like frames that allow transmission errors to be detected. ECM works by sending a fax page in a series of blocks. After receiving the complete page data, the receiving fax identifies any frames with errors. The sending fax then retransmits those frames. This process is repeated until all frames have been received without errors.
If a receiving fax machine is not able to receive an error-free page, the fax transmission may fail, and one of the fax machines may disconnect. If a network has packet-loss levels greater than 3 to 5 percent, fax transmissions consistently fail when ECM is enabled. Fax relay packet loss concealment disables ECM so that fax calls with up to 9 percent packet loss succeed and calls with packet loss of 5 to 7 percent succeed with acceptable quality.
For more information, see the following documents:
Fax Pass-Through and Fax Pass-Through with Upspeed
Fax pass-through is the simplest technique for sending fax over IP networks, but it is not the default, nor is it the most desirable method of supporting fax over IP. T.38 fax relay provides a more reliable and error-free method of sending faxes over an IP network, but some third-party H.323 and SIP implementations do not support T.38 fax relay. These same implementations often support fax pass-through.
Fax pass-through is the state of the channel after the fax upspeed process has occurred. In fax pass-through mode, gateways do not distinguish a fax call from a voice call. Fax communication between the two fax machines is carried in its entirety in-band over a voice call. When using fax pass-through with upspeed, the gateways are to some extent aware of the fax call. Although relay mechanisms are not employed, with upspeed the gateways do recognize a CED fax tone and automatically change the voice codec to G.711 if necessary (thus the designation upspeed ) and turn off echo cancellation (EC) and voice activity detection (VAD) for the duration of the call.
Fax pass-through is also known as Voice Band Data (VBD) by the International Telecommunication Union (ITU). VBD refers to the transport of fax or modem signals over a voice channel through a packet network with an encoding appropriate for fax or modem signals. The minimum set of coders for VBD mode is G.711 u-law and a-law with VAD disabled.
Once a terminating gateway (TGW) detects a CED tone from a called fax machine, the TGW exchanges the voice codec that was negotiated during the voice call setup for a G.711 codec and turns off EC and VAD. This switchover is communicated to the originating gateway (OGW), which allows the fax machines to transfer modem signals as though they were traversing the PSTN. If the voice codec that was configured and negotiated for the VoIP call is G.711 when the CED tone is detected, there is no need to make any changes to the session other than turning off EC and VAD.
Before pass-through features were introduced (in Cisco IOS Release 12.1(3)T for the Cisco AS5300, and later for other Cisco IOS gateway platforms), fax pass-through was achieved by manually configuring a dial peer that only matched fax calls to set the codec parameters to G.711 with no EC and no VAD (or to clear-channel codec). Control of fax pass-through is achieved through named signaling events (NSEs) that are sent in the RTP stream.
NSEs are a Cisco-proprietary version of IETF-standard named telephony events (NTEs), which are specially marked data packets used to digitally convey telephony signaling tones and events. NSEs use different event values than NTEs and are generally sent with RTP payload type 100, whereas NTEs use payload type 101. NSEs and NTEs provide a more reliable way to communicate tones and events by using a single packet rather than a series of in-band packets that can be corrupted or partially lost.
Fax pass-through and fax pass-through with upspeed use peer-to-peer NSEs within the Real-Time Transport Protocol (RTP) stream or bearer stream to coordinate codec switchover and the disabling of EC and VAD. Redundant packets can be sent to improve reliability when the probability of packet loss is high.
When a DSP is put into voice mode at the beginning of a VoIP call, the DSP is informed by the call control stack whether the control protocol can support pass-through or not. If pass-through is supported, the following events occur:
For configuration instructions, see Chapter 1, "Configuring Fax Pass-Through."
Fax pass-through call flow is shown in the figure below.
Cisco Fax Relay
Cisco fax relay is the oldest method of supporting fax on Cisco IOS gateways and has been supported since Cisco IOS Release 11.3. Cisco fax relay uses Real-Time Transport Protocol (RTP) as the method of transport. In Cisco fax relay mode, gateways terminate T.30 fax signaling by spoofing a virtual fax machine to the locally attached fax machine. The gateways use a Cisco-proprietary fax-relay RTP-based protocol to communicate between them.
Unlike fax pass-through, fax relay demodulates the fax modem bits at the local gateway, sends the information across the voice network using the fax relay protocol, and then remodulates the bits back into tones at the far gateway. The fax machines on either end are sending and receiving tones and are not aware that a demodulation/modulation fax relay process is occurring.
The default method for fax transmission on Cisco IOS gateways is Cisco fax relay. This is an RTP-based transmission method that uses proprietary signaling and encoding mechanisms. Cisco fax relay capability is widely available and has been in the Cisco IOS gateway software since Cisco IOS Release 11.3, which introduced DSPs to enable voice applications. The mechanism for Cisco fax relay is the same for calls that are controlled by SIP, MGCP, or H.323 call control protocols.
Before T.38 standards-based fax relay was introduced, no command-line interface (CLI) was required to enable Cisco fax relay. Today Cisco fax relay is still the default, but explicit CLI enables a choice between the fax relay methods.
Cisco fax relay is the default operation and, in the absence of any explicit CLI on the dial peer, is used when a fax transmission is detected. If voice calls are being completed successfully between two routers, fax calls should also work. Events that occur during a Cisco fax relay call fall into the following call phases:
For configuration information, see Chapter 1, "Configuring Cisco Fax Relay."
Cisco Fax Relay Fax Setup Phase
When a DSP is put into voice mode at the beginning of a VoIP call, the DSP is informed by the call control stack whether fax relay is supported and if it is supported, whether it is Cisco fax relay or T.38 fax relay. If Cisco fax relay is supported, the following events occur:
Cisco fax relay fax setup is shown in the figure below.
Cisco Fax Relay Data Transfer Phase
During fax relay operation, the T.30 analog fax signals are received from the PSTN or from a directly attached fax machine. The T.30 fax signals are demodulated by a DSP on the gateway and then packetized and sent across the VoIP network as data. The TGW decodes the data stream and remodulates the T.30 analog fax signals to be sent to the PSTN or to a destination fax machine.
The messages that are demodulated and remodulated are predominantly the phase B, phase D, and phase E messages of a T.30 transaction. Most of the messages are passed across without any interference, but certain messages are modified according to the constraints of the VoIP network.
During phase B, fax machines interrogate each other's capabilities. They expect to communicate with each other across a 64-kbps PSTN circuit, and they attempt to make best use of the available bandwidth and circuit quality of a 64-kbps voice path. However, in a VoIP network, the fax machines do not have a 64-kbps PSTN circuit available. The bandwidth per call is probably less than 64 kbps, and the circuit is not considered a clear circuit.
Because transmission paths in VoIP networks are more limited than in the PSTN, Cisco IOS CLI is used to adjust fax settings on the VoIP dial peer. The adjusted fax settings restrict the facilities that are available to fax machines across the VoIP call leg and are also used to modify values in DIS and NSF messages that are received from fax machines.
The call flow of the Cisco fax relay data transfer phase is shown in the figure below.
T.38 Fax Relay
The T.38 fax relay feature provides an ITU-T standards-based method and protocols for fax relay. Data is packetized and encapsulated according to the T.38 standard. The encoding of the packet headers and the mechanism to switch from VoIP mode to fax relay mode are clearly defined in the specification. Annexes to the basic specification include details for operation under Session Initiation Protocol (SIP) and H.323 call control protocols.
T.38 fax relay provides an ITU-standard mechanism for a voice gateway to inform another voice gateway of the desire to change the media stream from a voice stream to a data stream. The desire to change the media stream is indicated by the call control protocol, and not through a change in the RTP payload or bearer information. Annexes to the T.38 specification define the switchover mechanism for the following call control protocols:
T.38 fax relay uses data redundancy to accommodate packet loss. During T.38 call establishment, voice gateways indicate the level of packet redundancy that they incorporate in their transmission of Facsimile User Datagram Packet Transport Layer packets (UDPTLs). The level of redundancy (the number of times that the packet is repeated) can be configured on Cisco IOS gateways.
There is work under way to implement T.38 fax switchover independently of the call control mechanisms. This is referred to as "bearer level signaling" and makes use of named signaling events (NSEs). The following sections address call-control-initiated switchover mechanisms:
For configuration information, see Chapter 1, "Configuring T.38 Fax Relay."
H.323 T.38 Fax Relay
The T.38 Annex B standard defines the mechanism that is used to switch over from voice mode to T.38 fax mode during a call. The ability to support T.38 must be indicated during the initial VoIP call setup. If the DSP on the gateway is capable of supporting T.38 mode, this information is indicated during the H.245 negotiation procedures as part of the regular H.323 VoIP call setup.
Once the VoIP call setup is completed, the DSP continues to listen for a fax tone. When a fax tone is heard, the DSP signals the receipt of fax tone to the call control layer, which then initiates fax changeover as specified in the T.38 Annex B procedures. The H.245 message flow shown in the figure below contains the following events:
SIP T.38 Fax Relay
When the call control protocol is SIP, T.38 Annex D procedures are used for the changeover from VoIP to fax mode during a call. Initially, a normal VoIP call is established using SIP INVITEs. The DSP needs to be informed that it can support T.38 mode while it is put into voice mode. Then, during the call, when the DSP detects fax HDLC flags, it signals the detection of the flags to the call control layer, and the call control layer initiates a SIP INVITE mid-call to signal the desire to change the media stream.
The SIP T.38 fax relay call flow shown in the figure below contains the following events:
MGCP T.38 Fax Relay
The MGCP T.38 fax relay feature conforms to ITU-T T.38, Procedures for Real-Time Group 3 Facsimile Communication over IP Networks, which determines procedures for real-time facsimile communication in various gateway control protocol (XGCP) applications.
MGCP T.38 fax relay provides two modes of implementation:
MGCP-based T.38 fax relay enables interworking between the T.38 application that already exists on Cisco gateways and the MGCP applications on call agents.
MGCP-based T.38 fax relay has the following call flow:
A fax relay MGCP event allows the gateway to notify the call agent of the status (start, stop, or failure) of T.38 processing for the connection. This event is sent in both call-agent-controlled and gateway-controlled mode.
Gateway-Controlled MGCP T.38 Fax Relay
In gateway-controlled mode, a call agent uses the fx: extension of the local connection option (LCO) to instruct a gateway about how to process a call. Gateways do not need instruction from the call agent to switch to T.38 mode. This mode is used if the call agent has not been upgraded to support T.38 and MGCP interworking, or if the call agent does not want to manage fax calls. Gateway-controlled mode can also be used to bypass the message delay overhead caused by call agent handling; for example, to meet time requirements for switchover to T.38 mode. If the call agent does not specify the mode to the gateway, the gateway defaults to gateway-controlled mode.
In gateway-controlled mode, the gateways exchange NSEs that do the following:
CA-Controlled MGCP T.38 Fax Relay
In call-agent (CA)-controlled mode, the call agent can instruct the gateway to switch to T.38 for a call. In Cisco IOS Release 12.3(1) and later releases, CA-controlled mode enables T.38 fax relay interworking between H.323 gateways and MGCP gateways and between two MGCP gateways under the control of a call agent. This feature supersedes previous methods for CA-controlled fax relay and introduces the following gateway capabilities to enable this functionality:
T.37 Store-and-Forward Fax
The T.37 store-and-forward feature provides an ITU-T standards-based method for store-and-forward fax. The fax transmission process is divided into distinct sending and receiving phases with the potential to store the fax between sending and receiving, if necessary.
A store-and-forward fax gateway takes calls from G3 fax machines, converts them into e-mail messages, and sends them over an IP network. Another store-and-forward fax gateway at the terminating end of the network receives the e-mail message, converts it back into a fax message, and delivers it to a far-end G3 fax machine. The transmitting gateway is referred to as an on-ramp gateway, and the terminating gateway is referred to as an off-ramp gateway. With store-and-forward fax, you can do the following:
Cisco fax gateways support the T.37 standard as independent on-ramp gateways, independent off-ramp gateways, or combined on-ramp and off-ramp gateways. The two phases, on-ramp fax and off-ramp fax, are often combined to provide fax throughput over an IP network. Advantages of T.37 store-and-forward fax include delivery at off-peak hours, sophisticated retry-on-busy algorithms, and the ability to broadcast a single fax to multiple receiving fax machines.
With store-and-forward fax, the on-ramp gateway receives a fax from a traditional PSTN-based Group 3 fax device and converts it into a Tagged Image File Format (TIFF) file attachment. The gateway creates a standard Multipurpose Internet Mail Extension (MIME) e-mail message and attaches the TIFF file to the e-mail. The gateway forwards the e-mail, now called a fax mail, and its attachment to the messaging infrastructure of a designated Simple Mail Transport Protocol (SMTP) server. The messaging infrastructure performs message routing, message storage, and transport, and can be custom store-and-forward SMTP software or a standard Internet mail transfer agent (MTA) such as UNIX sendmail or Netscape MailServer. The IETF standards for fax transmission are covered by RFC 2301 through 2306. TIFF-F describes the data format for compressed fax images.
Many MTAs on the market work without modification with both the on-ramp and off-ramp features of store-and-forward fax. We recommend that you dedicate a mail server to fax mail and avoid the conflicting configuration requirements of traditional e-mail and fax-mail servers. Optimize each mail server for its individual functions--for example, fax messages should usually retry transmissions every 5 minutes whereas normal e-mail should retry every 30 minutes, and fax messages should give up after 3 to 4 hours whereas normal e-mail should not give up for 4 to 5 days.
After the fax mail is stored on the SMTP server, it can be delivered in two ways: either as an e-mail message with attachment when the recipient downloads e-mail messages or as a fax to a standard PSTN-based G3 fax device. In the latter case, the SMTP server mail delivery infrastructure delivers the fax mail to the off-ramp gateway, which converts the attached TIFF file back into standard fax format and then sends the information to a standard PSTN-based G3 fax device. The off-ramp gateway is also responsible for generating delivery status notifications (DSNs) and message disposition notifications (MDNs), as appropriate.
A topology for T.37 store-and-forward fax is shown in the figure below.
T.37 store-and-forward fax is implemented on Cisco gateways using TCL IVR applications. For configuration information, see Chapter 1, "Configuring T.37 Store-and-Forward Fax."
IVR Applications for Fax
The following IVR applications have been developed for fax:
Fax Detection IVR Application
Fax detection supports the use of a single E.164 number for both voice mail and fax mail by providing the capability to detect through an interactive voice response interface whether an incoming call is voice or fax. Fax detection can be configured to use either the distinctive fax calling tones (CNG) or a manually dialed digit or both to distinguish fax calls from voice calls. Fax detection supports the following modes of operation:
For configuration information, see Chapter 1, "Configuring Fax Detection."
Fax Rollover IVR Application
The fax rollover IVR application provides a configured fallback to T.37 store-and-forward fax if a call attempts to use fax relay and fails. An OGW must be configured with fax relay, store-and-forward fax, and also with the fax rollover application. Then, if a fax relay attempt fails, the call is forwarded to an SMTP server by a mail transfer agent (MTA) using T.37-standard protocols for store-and-forward fax.
For configuration information, see Chapter 1, "Configuring Fax Rollover."
Information About Cisco IOS Modem Services over IP
Modem Passthrough over VoIP
When service providers and aggregators are implementing VoIP, they sometimes cannot separate data traffic from voice traffic. These carriers that aggregate voice traffic over VoIP infrastructures require service offerings to carry data as easily as voice.
Modem passthrough over VoIP provides for the transport of modem signals through a packet network by using pulse code modulation (PCM)-encoded packets.
Modem passthrough performs the following functions:
On detection of the modem answer tone, the gateways switch into modem passthrough mode. With modem passthrough, the modem traffic is carried between the two gateways in real-time transport protocol (RTP) packets, using an uncompressed or lightly compressed voice codec--G.711 u-law, G.711 a-law, or Voice Band Data (VBD). Packet redundancy may be used to mitigate the effects of packet loss in the IP network. Even so, modem passthrough remains susceptible to packet loss, jitter, and latency in the IP network.
The figure below illustrates the connection from the client modem to a modem ISDN channel aggregation (MICA) technologies modem network access server (NAS).
Voice Band Data
The modem passthrough feature is also known as Voice Band Data (VBD) by the International Telecommunication Union (ITU). VBD refers to the transport of modem signals over a voice channel through a packet network with an encoding appropriate for modem signals. The minimum set of coders for VBD mode is G.711 ulaw and alaw.
For VBD mode of operation, the path between the originating and answering gateway remains in a voice configuration. The modem signals are encoded using an appropriate speech codec suitable for the task, and samples are transported across a packet network. Currently G.711 is supported.
Some system requirements for the use of VBD follow:
When the gateways detect a data modem, both the originating gateway and the terminating gateway switch to modem passthrough mode. This switchover includes the following:
At the end of the modem or fax call, the voice ports revert to the previous configuration and the DSPs switch back to the original voice codec.
Packet loss is a persistent issue in voice applications. The disruption of speech, which is characteristic of packet loss, can be somewhat resolved with controlled redundancy and the RTP (RFC 2198). Controlled redundancy reconstructs missing information at the receiver end from the redundant data that arrives in the transmitted packets.
Some of the requirements for a controlled redundancy are as follows:
You can enable redundancy so that the modem and fax passthrough switchover causes the gateway to transmit redundant packets and redundancy can be enabled in one or both of the gateways. When only one gateway is configured, the other gateway receives the packets correctly, but does not produce redundant packets. When redundancy is enabled, 10-ms sample-sized packets are sent. When redundancy is disabled, 20-ms sample-sized packets are sent.
Clock Slip Buffer Management
When the gateways detect a data modem, both the originating gateway and the terminating gateway switch from dynamic and adaptive buffers to static de-jitter buffers. The use of a static de-jitter buffer is required for modem passthrough because the adaptation process in a dynamic de-jitter buffer causes a retrain on the modem connection. When the modem call is concluded, the voice ports revert to dynamic jitter buffers.
In addition, the modem passthrough data management algorithm is designed to handle and compensate for clocking differences in the PSTN between the originating and terminating gateways. This additional clock-slip monitoring prevents issues that show up in long duration modem calls.
Modem Relay over VoIP
The Modem Relay feature provides support for modem connections across traditional time-division multiplexing (TDM) networks. Modem relay demodulates a modem signal at one voice gateway and passes it as packet data to another voice gateway where the signal is remodulated and sent to a receiving modem. On detection of the modem answer tone, the gateways switch into modem passthrough mode and then, if the call menu (CM) signal is detected, the two gateways switch into modem relay mode.
Differences Between Modem Passthrough and Modem Relay
There are two ways to transport modem traffic over VoIP networks:
In this implementation, the call starts out as a voice call, then switches into modem passthrough mode, and then into modem relay mode.
Modem Tone Detection and Signaling
This implementation of modem relay supports V.34 modulation and the V.42 error correction and link layer protocol with maximum transfer rates of up to 33.6 kbps. It forces higher-rate modems to train down to the supported rates. Signaling support includes the Session Initiation Protocol (SIP), MGCP/SGCP, and H.323:
Benifits of Modem Relay
Modem relay on VoIP offers the following benefits:
You can enable payload redundancy so that the modem relay VoIP switchover causes the gateway to send redundant packets. Redundancy can be enabled in one or both of the gateways. When only a single gateway is configured for redundancy, the other gateway receives the packets correctly, but does not produce redundant packets. When redundancy is enabled, 10-ms sample-sized packets are sent. When redundancy is disabled, 20-ms sample-sized packets are sent.
Clock Slip Buffer Management
When the gateways detect a data modem, both the originating and the terminating gateways switch from dynamic jitter buffers to static jitter buffers of 200-ms depth. The switch from dynamic to static is designed to compensate for Public Switched Telephone Network (PSTN) clocking differences at the originating and terminating gateways. When the modem call is concluded, the voice ports revert to dynamic jitter buffers.
Developers using this guide may be interested in joining the Cisco Developer Support Program. This program was created to provide you with a consistent level of support that you can depend on while leveraging Cisco interfaces in your development projects.
The Developer Support Program provides formalized support for Cisco Systems interfaces to enable developers, customers, and partners in the Cisco Technology Developer program to accelerate their delivery of compatible solutions.
The Developer Support Engineers are an extension of the product technology engineering teams. They have direct access to the resources necessary to provide expert support in a timely manner.
For additional information on this General Support and Program FAQ's refer to the Developer Support Program Web Site at www.cisco.com/go/developersupport/ or contact firstname.lastname@example.org.
1 Not all supported standards are listed.
2 Not all supported MIBs are listed.
3 Not all supported RFCs are listed.
Cisco and the Cisco logo are trademarks or registered trademarks of Cisco and/or its affiliates in the U.S. and other countries. To view a list of Cisco trademarks, go to this URL: www.cisco.com/go/trademarks. Third-party trademarks mentioned are the property of their respective owners. The use of the word partner does not imply a partnership relationship between Cisco and any other company. (1110R)
Any Internet Protocol (IP) addresses and phone numbers used in this document are not intended to be actual addresses and phone numbers. Any examples, command display output, network topology diagrams, and other figures included in the document are shown for illustrative purposes only. Any use of actual IP addresses or phone numbers in illustrative content is unintentional and coincidental.
© 2012 Cisco Systems, Inc. All rights reserved.