RFC 2705 - Media Gateway Control Protocol (MGCP) Version 1.0
(Formats: TXT)


Network Working Group                                          M. Arango
Request for Comments: 2705                                       RSL COM
Category: Informational                                         A. Dugan
                                                              I. Elliott
                                                   Level3 Communications
                                                              C. Huitema
                                                               Telcordia
                                                              S. Pickett
                                                       Vertical Networks
                                                            October 1999
Media Gateway Control Protocol (MGCP) Version 1.0 Status of this Memo This memo provides information for the Internet community. It does not specify an Internet standard of any kind. Distribution of this memo is unlimited. Copyright Notice Copyright (C) The Internet Society (1999). All Rights Reserved. IESG NOTE: This document is being published for the information of the community. It describes a protocol that is currently being deployed in a number of products. Implementers should be aware of developments in the IETF Megaco Working Group and ITF-T SG16 who are currently working on a potential successor to this protocol. Abstract This document describes an application programming interface and a corresponding protocol (MGCP) for controlling Voice over IP (VoIP) Gateways from external call control elements. MGCP assumes a call control architecture where the call control "intelligence" is outside the gateways and handled by external call control elements. The document is structured in 6 main sections: * The introduction presents the basic assumptions and the relation to other protocols such as H.323, RTSP, SAP or SIP. Arango, et al. Informational [Page 1] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 * The interface section presents a conceptual overview of the MGCP, presenting the naming conventions, the usage of the session description protocol SDP, and the procedures that compose MGCP: Notifications Request, Notification, Create Connection, Modify Connection, Delete Connection, AuditEndpoint, AuditConnection and RestartInProgress. * The protocol description section presents the MGCP encodings, which are based on simple text formats, and the transmission procedure over UDP. * The security section presents the security requirement of MGCP, and its usage of IP security services (IPSEC). * The event packages section provides an initial definition of packages and event names. * The description of the changes made in combining SGCP 1.1 and IPDC to create MGCP 1.0. Table of Contents 1. Introduction .............................................. 5 1.1. Relation with the H.323 standards .................... 7 1.2. Relation with the IETF standards ..................... 8 1.3. Definitions .......................................... 9 2. Media Gateway Control Interface ........................... 9 2.1. Model and naming conventions. ........................ 10 2.1.1. Types of endpoints .............................. 10 2.1.1.1. Digital channel (DS0) ...................... 11 2.1.1.2. Analog line ................................ 11 2.1.1.3. Annoucement server access point ............ 12 2.1.1.4. Interactive Voice Response access point .... 12 2.1.1.5. Conference bridge access point ............. 13 2.1.1.6. Packet relay ............................... 13 2.1.1.7. Wiretap access point ....................... 14 2.1.1.8. ATM "trunk side" interface. ................ 14 2.1.2. Endpoint identifiers ............................ 15 2.1.3. Calls and connections ........................... 17 2.1.3.1. Names of calls ............................. 20 2.1.3.2. Names of connections ....................... 20 2.1.3.3. Management of resources, attributes of ..... 20 2.1.3.4. Special case of local connections .......... 23 2.1.4. Names of Call Agents and other entities ......... 23 2.1.5. Digit maps ...................................... 24 2.1.6. Names of events ................................. 26 2.2. Usage of SDP ......................................... 29 2.3. Gateway Control Commands ............................. 30 Arango, et al. Informational [Page 2] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.3.1. EndpointConfiguration ........................... 32 2.3.2. NotificationRequest ............................. 33 2.3.3. CreateConnection ................................ 38 2.3.4. ModifyConnection ................................ 44 2.3.5. DeleteConnection (from the Call Agent) .......... 46 2.3.6. DeleteConnection (from the VoIP gateway) ........ 51 2.3.7. DeleteConnection (multiple connections, from the 51 2.3.8. Audit Endpoint .................................. 52 2.3.9. Audit Connection ................................ 55 2.3.10. Restart in progress ............................ 56 2.4. Return codes and error codes. ........................ 58 2.5. Reason Codes ......................................... 61 3. Media Gateway Control Protocol ............................ 61 3.1. General description .................................. 62 3.2. Command Header ....................................... 62 3.2.1. Command line .................................... 62 3.2.1.1. Coding of the requested verb ............... 63 3.2.1.2. Transaction Identifiers .................... 63 3.2.1.3. Coding of the endpoint identifiers and ..... 64 3.2.1.4. Coding of the protocol version ............. 65 3.2.2. Parameter lines ................................. 65 3.2.2.1. Response Acknowledgement ................... 68 3.2.2.2. Local connection options ................... 68 3.2.2.3. Capabilities ............................... 70 3.2.2.4. Connection parameters ...................... 71 3.2.2.5. Reason Codes ............................... 72 3.2.2.6. Connection mode ............................ 73 3.2.2.7. Coding of event names ...................... 73 3.2.2.8. RequestedEvents ............................ 74 3.2.2.9. SignalRequests ............................. 76 3.2.2.10. ObservedEvent ............................. 76 3.2.2.11. RequestedInfo ............................. 76 3.2.2.12. QuarantineHandling ........................ 77 3.2.2.13. DetectEvents .............................. 77 3.2.2.14. EventStates ............................... 77 3.2.2.15. RestartMethod ............................. 78 3.2.2.16. Bearer Information ........................ 78 3.3. Format of response headers ........................... 78 3.4. Formal syntax description of the protocol ............ 81 3.5. Encoding of the session description .................. 86 3.5.1. Usage of SDP for an audio service ............... 86 3.5.2. Usage of SDP in a network access service ........ 87 3.5.3. Usage of SDP for ATM connections ................ 90 3.5.4. Usage of SDP for local connections .............. 91 3.6. Transmission over UDP ................................ 91 3.6.1. Providing the At-Most-Once functionality ........ 91 3.6.2. Transaction identifiers and three ways handshake. 92 3.6.3. Computing retransmission timers ................. 93 Arango, et al. Informational [Page 3] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 3.6.4. Piggy backing ................................... 94 3.6.5. Provisional responses ........................... 94 4. States, failover and race conditions. ..................... 95 4.1. Basic Asumptions ..................................... 95 4.2. Security, Retransmission, and Detection of Lost ...... 96 4.3. Race conditions ...................................... 99 4.3.1. Quarantine list ................................. 99 4.3.2. Explicit detection ..............................103 4.3.3. Ordering of commands, and treatment of disorder .104 4.3.4. Fighting the restart avalanche ..................105 4.3.5. Disconnected Endpoints ..........................107 1. A "disconnected" timer is initialized to a random value, .107 2. The gateway then waits for either the end of this timer, .107 3. When the "disconnected" timer elapses, when a command is .107 4. If the "disconnected" procedure still left the endpoint ..107 5. Security requirements .....................................108 5.1. Protection of media connections ......................109 6. Event packages and end point types ........................109 6.1. Basic packages .......................................110 6.1.1. Generic Media Package ...........................110 6.1.2. DTMF package ....................................112 6.1.3. MF Package ......................................113 6.1.4. Trunk Package ...................................114 6.1.5. Line Package ....................................116 6.1.6. Handset emulation package .......................119 6.1.7. RTP Package .....................................120 6.1.8. Network Access Server Package ...................121 6.1.9. Announcement Server Package .....................122 6.1.10. Script Package .................................122 6.2. Basic endpoint types and profiles ....................123 7. Versions and compatibility ................................124 7.1. Differences between version 1.0 and draft 0.5 ........124 7.2. Differences between draft-04 and draft-05 ............125 7.3. Differences between draft-03 and draft-04 ............125 7.4. Differences between draft-02 and draft-03 ............125 7.5. Differences between draft-01 and draft-02 ............126 7.6. The making of MGCP from IPDC and SGCP ................126 7.7. Changes between MGCP and initial versions of SGCP ....126 8. Security Considerations ...................................128 9. Acknowledgements ..........................................128 10. References ................................................129 11. Authors' Addresses ........................................130 12. Appendix A: Proposed "MoveConnection" command .............132 12.1. Proposed syntax modification ........................133 13. Full Copyright Statement ..................................134 Arango, et al. Informational [Page 4] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 1. Introduction This document describes an abstract application programming interface and a corresponding protocol (MGCP) for controlling Telephony Gateways from external call control elements called media gateway controllers or call agents. A telephony gateway is a network element that provides conversion between the audio signals carried on telephone circuits and data packets carried over the Internet or over other packet networks. Example of gateways are: * Trunking gateways, that interface between the telephone network and a Voice over IP network. Such gateways typically manage a large number of digital circuits. * Voice over ATM gateways, which operate much the same way as voice over IP trunking gateways, except that they interface to an ATM network. * Residential gateways, that provide a traditional analog (RJ11) interface to a Voice over IP network. Examples of residential gateways include cable modem/cable set-top boxes, xDSL devices, broad-band wireless devices * Access gateways, that provide a traditional analog (RJ11) or digital PBX interface to a Voice over IP network. Examples of access gateways include small-scale voice over IP gateways. * Business gateways, that provide a traditional digital PBX interface or an integrated "soft PBX" interface to a Voice over IP network. * Network Access Servers, that can attach a "modem" to a telephone circuit and provide data access to the Internet. We expect that, in the future, the same gateways will combine Voice over IP services and Network Access services. * Circuit switches, or packet switches, which can offer a control interface to an external call control element. MGCP assumes a call control architecture where the call control "intelligence" is outside the gateways and handled by external call control elements. The MGCP assumes that these call control elements, or Call Agents, will synchronize with each other to send coherent commands to the gateways under their control. MGCP does not define a mechanism for synchronizing Call Agents. MGCP is, in essence, a master/slave protocol, where the gateways are expected to execute commands sent by the Call Agents. In consequence, this document specifies in great detail the expected behavior of the gateways, but Arango, et al. Informational [Page 5] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 only specify those parts of a call agent implementation, such as timer management, that are mandated for proper operation of the protocol. MGCP assumes a connection model where the basic constructs are endpoints and connections. Endpoints are sources or sinks of data and could be physical or virtual. Examples of physical endpoints are: * An interface on a gateway that terminates a trunk connected to a PSTN switch (e.g., Class 5, Class 4, etc.). A gateway that terminates trunks is called a trunk gateway. * An interface on a gateway that terminates an analog POTS connection to a phone, key system, PBX, etc. A gateway that terminates residential POTS lines (to phones) is called a residential gateway. An example of a virtual endpoint is an audio source in an audio- content server. Creation of physical endpoints requires hardware installation, while creation of virtual endpoints can be done by software. Connections may be either point to point or multipoint. A point to point connection is an association between two endpoints with the purpose of transmitting data between these endpoints. Once this association is established for both endpoints, data transfer between these endpoints can take place. A multipoint connection is established by connecting the endpoint to a multipoint session. Connections can be established over several types of bearer networks: * Transmission of audio packets using RTP and UDP over a TCP/IP network. * Transmission of audio packets using AAL2, or another adaptation layer, over an ATM network. * Transmission of packets over an internal connection, for example the TDM backplane or the interconnection bus of a gateway. This is used, in particular, for "hairpin" connections, connections that terminate in a gateway but are immediately rerouted over the telephone network. For point-to-point connections the endpoints of a connection could be in separate gateways or in the same gateway. Arango, et al. Informational [Page 6] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 1.1. Relation with the H.323 standards MGCP is designed as an internal protocol within a distributed system that appears to the outside as a single VoIP gateway. This system is composed of a Call Agent, that may or may not be distributed over several computer platforms, and of a set of gateways, including at least one "media gateway" that perform the conversion of media signals between circuits and packets, and at least one "signalling gateway" when connecting to an SS7 controlled network. In a typical configuration, this distributed gateway system will interface on one side with one or more telephony (i.e. circuit) switches, and on the other side with H.323 conformant systems, as indicated in the following table: ___________________________________________________________________ | Functional| Phone | Terminating | H.323 conformant | | Plane | switch | Entity | systems | |___________|____________|_________________|_______________________| | Signaling | Signaling | Call agent | Signaling exchanges | | Plane | exchanges | | with the call agent | | | through | | through H.225/RAS and| | | SS7/ISUP | | H.225/Q.931. | |___________|____________|_________________|_______________________| | | | | Possible negotiation | | | | | of logical channels | | | | | and transmission | | | | | parameters through | | | | | H.245 with the call | | | | | agent. | |___________|____________|_________________|_______________________| | | | Internal | | | | | synchronization| | | | | through MGCP | | |___________|____________|_________________|_______________________| | Bearer | Connection| Telephony | Transmission of VOIP | | Data | through | gateways | data using RTP | | Transport | high speed| | directly between the | | Plane | trunk | | H.323 station and the| | | groups | | gateway. | |___________|____________|_________________|_______________________| In the MGCP model, the gateways focus on the audio signal translation function, while the Call Agent handles the signaling and call processing functions. As a consequence, the Call Agent implements the "signaling" layers of the H.323 standard, and presents itself as an "H.323 Gatekeeper" or as one or more "H.323 Endpoints" to the H.323 systems. Arango, et al. Informational [Page 7] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 1.2. Relation with the IETF standards While H.323 is the recognized standard for VoIP terminals, the IETF has also produced specifications for other types of multi-media applications. These other specifications include: * the Session Description Protocol (SDP), RFC 2327, * the Session Announcement Protocol (SAP), * the Session Initiation Protocol (SIP), * the Real Time Streaming Protocol (RTSP), RFC 2326. The latter three specifications are in fact alternative signaling standards that allow for the transmission of a session description to an interested party. SAP is used by multicast session managers to distribute a multicast session description to a large group of recipients, SIP is used to invite an individual user to take part in a point-to-point or unicast session, RTSP is used to interface a server that provides real time data. In all three cases, the session description is described according to SDP; when audio is transmitted, it is transmitted through the Real-time Transport Protocol, RTP. The distributed gateway systems and MGCP will enable PSTN telephony users to access sessions set up using SAP, SIP or RTSP. The Call Agent provides for signaling conversion, according to the following table: Arango, et al. Informational [Page 8] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 _____________________________________________________________________ | Functional| Phone | Terminating | IETF conforming systems| | Plane | switch | Entity | | |___________|____________|_________________|_________________________| | Signaling | Signaling | Call agent | Signaling exchanges | | Plane | exchanges | | with the call agent | | | through | | through SAP, SIP or | | | SS7/ISUP | | RTSP. | |___________|____________|_________________|_________________________| | | | | Negotiation of session | | | | | description parameters | | | | | through SDP (telephony | | | | | gateway terminated but | | | | | passed via the call | | | | | agent to and from the | | | | | IETF conforming system)| |___________|____________|_________________|_________________________| | | | Internal | | | | | synchronization| | | | | through MGCP | | |___________|____________|_________________|_________________________| | Bearer | Connection| Telephony | Transmission of VoIP | | Data | through | gateways | data using RTP, | | Transport | high speed| | directly between the | | Plane | trunk | | remote IP end system | | | groups | | and the gateway. | |___________|____________|_________________|_________________________| The SDP standard has a pivotal status in this architecture. We will see in the following description that we also use it to carry session descriptions in MGCP. 1.3. Definitions Trunk: A communication channel between two switching systems. E.g., a DS0 on a T1 or E1 line. 2. Media Gateway Control Interface The interface functions provide for connection control and endpoint control. Both use the same system model and the same naming conventions. Arango, et al. Informational [Page 9] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1. Model and naming conventions The MGCP assumes a connection model where the basic constructs are endpoints and connections. Connections are grouped in calls. One or more connections can belong to one call. Connections and calls are set up at the initiative of one or several Call Agents. 2.1.1. Types of endpoints In the introduction, we presented several classes of gateways. Such classifications, however, can be misleading. Manufacturers can arbitrarily decide to provide several types of services in a single packaging. A single product could well, for example, provide some trunk connections to telephony switches, some primary rate connections and some analog line interfaces, thus sharing the characteristics of what we described in the introduction as "trunking", "access" and "residential" gateways. MGCP does not make assumptions about such groupings. We simply assume that media gateways support collections of endpoints. The type of the endpoint determines its functionalities. Our analysis, so far, has led us to isolate the following basic endpoint types: * Digital channel (DS0), * Analog line, * Annoucement server access point, * Interactive Voice Response access point, * Conference bridge access point, * Packet relay, * Wiretap access point, * ATM "trunk side" interface. In this section, we will develop the expected behavior of such end points. This list is not limitative. There may be other types of endpoints defined in the future, for example test endpoint that could be used to check network quality, or frame-relay endpoints that could be used to managed audio channels multiplexed over a frame-relay virtual circuit. Arango, et al. Informational [Page 10] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1.1.1. Digital channel (DS0) Digital channels provide an 8Khz*8bit service. Such channels are found in trunk and ISDN interfaces. They are typically part of digital multiplexes, such as T1, E1, T3 or E3 interfaces. Media gateways that support such channels are capable of translating the digital signals received on the channel, which may be encoded according to A or mu-law, using either the complete set of 8 bits or only 7 of these bits, into audio packets. When the media gateway also supports a NAS service, the gateway shall be capable of receiving either audio-encoded data (modem connection) or binary data (ISDN connection) and convert them into data packets. +------- +------------+| (channel) ===|DS0 endpoint| -------- Connections +------------+| +------- Media gateways should be able to establish several connections between the endpoint and the packet networks, or between the endpoint and other endpoints in the same gateway. The signals originating from these connections shall be mixed according to the connection "mode", as specified later in this document. The precise number of connections that an endpoint support is a characteristic of the gateway, and may in fact vary according with the allocation of resource within the gateway. In some cases, digital channels are used to carry signalling. This is the case for example of SS7 "F" links, or ISDN "D" channels. Media gateways that support these signalling functions shall be able to send and receive the signalling packets to and from a call agent, using the "back haul" procedures defined by the SIGTRAN working group of the IETF. Digital channels are sometimes used in conjunction with channel associated signalling, such as "MF R2". Media gateways that support these signalling functions shall be able to detect and produce the corresponding signals, such as for example "wink" or "A", according to the event signalling and reporting procedures defined in MGCP. 2.1.1.2. Analog line Analog lines can be used either as a "client" interface, providing service to a classic telephone unit, or as a "service" interface, allowing the gateway to send and receive analog calls. When the media gateway also supports a NAS service, the gateway shall be capable of receiving audio-encoded data (modem connection) and convert them into data packets. Arango, et al. Informational [Page 11] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 +------- +---------------+| (line) ===|analog endpoint| -------- Connections +---------------+| +------- Media gateways should be able to establish several connections between the endpoint and the packet networks, or between the endpoint and other endpoints in the same gateway. The audio signals originating from these connections shall be mixed according to the connection "mode", as specified later in this document. The precise number of connections that an endpoint support is a characteristic of the gateway, and may in fact vary according with the allocation of resource within the gateway. A typical gateway should however be able to support two or three connections per endpoint, in order to provide services such as "call waiting" or "three ways calling". 2.1.1.3. Annoucement server access point An announcement server endpoint provides acces to an announcement service. Under requests from the call agent, the announcement server will "play" a specified announcement. The requests from the call agent will follow the event signalling and reporting procedures defined in MGCP. +----------------------+ | Announcement endpoint| -------- Connection +----------------------+ A given announcement endpoint is not supposed to support more than one connection at a time. If several connections were established to the same endpoint, then the same announcements would be played simultaneously over all the connections. Connections to an announcement server are typically oneway, or "half duplex" -- the announcement server is not expected to listen the audio signals from the connection. 2.1.1.4. Interactive Voice Response access point An Interactive Voice Response (IVR) endpoint provides acces to an IVR service. Under requests from the call agent, the IVR server will "play" announcements and tones, and will "listen" to responses from the user. The requests from the call agent will follow the event signalling and reporting procedures defined in MGCP. Arango, et al. Informational [Page 12] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 +-------------+ | IVR endpoint| -------- Connection +-------------+ A given IVR endpoint is not supposed to support more than one connection at a time. If several connections were established to the same endpoint, then the same tones and announcements would be played simultaneously over all the connections. 2.1.1.5. Conference bridge access point A conference bridge endpoint is used to provide access to a specific conference. +------- +--------------------------+| |Conference bridge endpoint| -------- Connections +--------------------------+| +------- Media gateways should be able to establish several connections between the endpoint and the packet networks, or between the endpoint and other endpoints in the same gateway. The signals originating from these connections shall be mixed according to the connection "mode", as specified later in this document. The precise number of connections that an endpoint support is a characteristic of the gateway, and may in fact vary according with the allocation of resource within the gateway. 2.1.1.6. Packet relay A packet relay endpoint is a specific form of conference bridge, that typically only supports two connections. Packets relays can be found in firewalls between a protected and an open network, or in transcoding servers used to provide interoperation between incompatible gateways, for example gateways that do not support compatible compression algorithms, or gateways that operate over different transmission networks such as IP and ATM. +------- +---------------------+ | |Packet relay endpoint| 2 connections +---------------------+ | +------- Arango, et al. Informational [Page 13] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1.1.7. Wiretap access point A wiretap access point provides access to a wiretap service, providing either a recording or a life playback of a connection. +-----------------+ | Wiretap endpoint| -------- Connection +-----------------+ A given wiretap endpoint is not supposed to support more than one connection at a time. If several connections were established to the same endpoint, then the recording or playback would mix the audio signals received on this connections. Connections to an wiretap endpoint are typically oneway, or "half duplex" -- the wiretap server is not expected to signal its presence in a call. 2.1.1.8. ATM "trunk side" interface. ATM "trunk side" endpoints are typically found when one or several ATM permanent virtual circuits are used as a replacement for the classic "TDM" trunks linking switches. When ATM/AAL2 is used, several trunks or channels are multiplexed on a single virtual circuit; each of these trunks correspond to a single endpoint. +------- +------------------+| (channel) = |ATM trunk endpoint| -------- Connections +------------------+| +------- Media gateways should be able to establish several connections between the endpoint and the packet networks, or between the endpoint and other endpoints in the same gateway. The signals originating from these connections shall be mixed according to the connection "mode", as specified later in this document. The precise number of connections that an endpoint support is a characteristic of the gateway, and may in fact vary according with the allocation of resource within the gateway. Arango, et al. Informational [Page 14] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1.2. Endpoint identifiers Endpoints identifiers have two components that both are case insensitive: * the domain name of the gateway that is managing the endpoint, * a local name within that gateway, The syntax of the local name depends on the type of endpoint being named. However, the local name for each of these types is naturally hierarchical, beginning with a term which identifies the physical gateway containing the given endpoint and ending in a term which specifies the individual endpoint concerned. With this in mind, the following rules for construction and interpretation of the Entity Name field for these entity types MUST be supported: 1) The individual terms of the naming path MUST be separated by a single slash ("/", ASCII 2F hex). 2) The individual terms are character strings composed of letters, digits or other printable characters, with the exception of characters used as delimitors ("/", "@"), characters used for wildcarding ("*", "$") and white spaces. 3) Wild-carding is represented either by an asterisk ("*") or a dollar sign ("$") for the terms of the naming path which are to be wild-carded. Thus, if the full naming path looks like term1/term2/term3 then the Entity Name field looks like this depending on which terms are wild-carded: */term2/term3 if term1 is wild-carded term1/*/term3 if term2 is wild-carded term1/term2/* if term3 is wild-carded term1/*/* if term2 and term3 are wild-carded, etc. In each of these examples a dollar sign could have appeared instead of an asterisk. Arango, et al. Informational [Page 15] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 4) A term represented by an asterisk is to be interpreted as: "use ALL values of this term known within the scope of the Media Gateway". A term represented by a dollar sign is to be interpreted as: "use ANY ONE value of this term known within the scope of the Media Gateway". The description of a specific command may add further criteria for selection within the general rules given here. If the Media Gateway controls multiple physical gateways, the first term of the naming MUST identify the physical gateway containing the desired entity. If the Media Gateway controls only a single physical gateway, the first term of the naming string MAY identify that physical gateway, depending on local practice. A local name that is composed of only a wildcard character refers to either all (*) or any ($) endpoints within the media gateway. In the case of trunking gateways, endpoints are trunk circuits linking a gateway to a telephone switch. These circuits are typically grouped into a digital multiplex, that is connected to the gateway by a physical interface. Such circuits are named in three contexts: * In the ISUP protocol, trunks are grouped into trunk groups, identified by the SS7 point codes of the switches that the group connects. Circuits within a trunk group are identified by a circuit number (CIC in ISUP). * In the gateway configuration files, physical interfaces are typically identified by the name of the interface, an arbitrary text string. When the interface multiplexes several circuits, individual circuits are typically identified by a circuit number. * In MGCP, the endpoints are identified by an endpoint identifier. The Call Agents use configuration databases to map ranges of circuit numbers within an ISUP trunk group to corresponding ranges of circuits in a multiplex connected to a gateway through a physical interface. The gateway will be identified, in MGCP, by a domain name. The local name will be structured to encode both the name of the physical interface, for example X35V3+A4, and the circuit number within the multiplex connected to the interface, for example 13. The circuit number will be separated from the name of the interface by a fraction bar, as in: X35V3+A4/13 Arango, et al. Informational [Page 16] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Other types of endpoints will use different conventions. For example, in gateways were physical interfaces by construction only control one circuit, the circuit number will be omitted. The exact syntax of such names should be specified in the corresponding server specification. 2.1.3. Calls and connections Connections are created on the call agent on each endpoint that will be involved in the "call." In the classic example of a connection between two "DS0" endpoints (EP1 and EP2), the call agents controlling the end points will establish two connections (C1 and C2): +---+ +---+ (channel1) ===|EP1|--(C1)--... ...(C2)--|EP2|===(channel2) +---+ +---+ Each connection will be designated locally by a connection identifier, and will be characterized by connection attributes. When the two endpoints are located on gateways that are managed by the same call agent, the creation is done via the three following steps: 1) The call agent asks the first gateway to "create a connection" on the first endpoint. The gateway allocates resources to that connection, and respond to the command by providing a "session description." The session description contains the information necessary for a third party to send packets towards the newly created connection, such as for example IP address, UDP port, and packetization parameters. 2) The call agent then asks the second gateway to "create a connection" on the second endpoint. The command carries the "session description" provided by the first gateway. The gateway allocates resources to that connection, and respond to the command by providing its own "session description." 3) The call agent uses a "modify connection" command to provide this second "session description" to the first endpoint. Once this is done, communication can proceed in both directions. When the two endpoints are located on gateways that are managed by the different call agents, these two call agents shall exchange information through a call-agent to call-agent signalling protocol, in order to synchronize the creation of the connection on the two endpoints. Arango, et al. Informational [Page 17] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Once established, the connection parameters can be modified at any time by a "modify connection" command. The call agent may for example instruct the gateway to change the compression algorithm used on a connection, or to modify the IP address and UDP port to which data should be sent, if a connection is "redirected." The call agent removes a connection by sending to the gateway a "delete connection" command. The gateway may also, under some circumstances, inform a gateway that a connection could not be sustained. The following diagram provides a view of the states of a connection, as seen from the gateway: Arango, et al. Informational [Page 18] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Create connection received | V +-------------------+ |resource allocation|-(failed)-+ +-------------------+ | | (connection refused) (successful) | v +----------->+ | | | +-------------------+ | | remote session | | | description |----------(yes)--------+ | | available ? | | | +-------------------+ | | | | | (no) | | | | | +-----------+ +------+ | +--->| half open |------> Delete <-------| open |<----------+ | | | (wait) | Connection |(wait)| | | | +-----------+ received +------+ | | | | | | | | | Modify Connection | Modify Connection | | | received | received | | | | | | | | | +--------------------+ | +--------------------+ | | | |assess modification | | |assess modification | | | | +--------------------+ | +--------------------+ | | | | | | | | | | |(failed) (successful) | (failed) (successful) | | | | | | | | | | +<---+ | | +-------------+-------+ | | | +<-------------------+ | | +-----------------+ | Free connection | | resources. | | Report. | +-----------------+ | V Arango, et al. Informational [Page 19] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1.3.1. Names of calls One of the attributes of each connection is the "call identifier." Calls are identified by unique identifiers, independent of the underlying platforms or agents. These identifiers are created by the Call Agent. They are treated in MGCP as unstructured octet strings. Call identifiers are expected to be unique within the system, or at a minimum, unique within the collection of Call Agents that control the same gateways. When a Call Agent builds several connections that pertain to the same call, either on the same gateway or in different gateways, these connections that belong to the same call share the same call-id. This identifier can then be used by accounting or management procedures, which are outside the scope of MGCP. 2.1.3.2. Names of connections Connection identifiers are created by the gateway when it is requested to create a connection. They identify the connection within the context of an endpoint. They are treated in MGCP as unstructured octet strings. The gateway should make sure that a proper waiting period, at least 3 minutes, elapses between the end of a connection that used this identifier and its use in a new connection for the same endpoint. (Gateways may decide to use identifiers that are unique within the context of the gateway.) 2.1.3.3. Management of resources, attributes of connections Many types of resources will be associated to a connection, such as specific signal processing functions or packetization functions. Generally, these resources fall in two categories: 1) Externally visible resources, that affect the format of "the bits on the network" and must be communicated to the second endpoint involved in the connection. 2) Internal resources, that determine which signal is being sent over the connection and how the received signals are processed by the endpoint. The resources allocated to a connection, and more generally the handling of the connection, are chosen by the gateway under instructions from the call agent. The call agent will provide these instructions by sending two set of parameters to the gateway: 1) The local directives instruct the gateway on the choice of resources that should be used for a connection, Arango, et al. Informational [Page 20] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2) When available, the "session description" provided by the other end of the connection. The local directives specify such parameters as the mode of the connection (e.g. send only, send-receive), preferred coding or packetization methods, usage of echo cancellation or silence suppression. (A detailed list can be found in the specification of the LocalConnectionOptions parameter of the CreateConnection command.) For each of these parameters, the call agent can either specify a value, a range of value, or no value at all. This allow various implementations to implement various level of control, from a very tight control where the call agent specifies minute details of the connection handling to a very loose control where the call agent only specifies broad guidelines, such as the maximum bandwidth, and let the gateway choose the detailed values. Based on the value of the local directives, the gateway will determine the resources allocated to the connection. When this is possible, the gateway will choose values that are in line with the remote session description - but there is no absolute requirement that the parameters be exactly the same. Once the resource have been allocated, the gateway will compose a "session description" that describes the way it intends to receive packets. Note that the session description may in some cases present a range of values. For example, if the gateway is ready to accept one of several compression algorithm, it can provide a list of these accepted algorithms. Arango, et al. Informational [Page 21] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Local Directives (from call agent 1) | V +-------------+ | resources | | allocation | | (gateway 1) | +-------------+ | | V | Local | Parameters V | Session | Description Local Directives | | (from call agent 2) | +---> Transmission----+ | | (CA to CA) | | | V V | +-------------+ | | resources | | | allocation | | | (gateway 2) | | +-------------+ | | | | | V | | Local | | Parameters | Session | Description | +---- Transmission<---+ | | (CA to CA) V V +-------------+ | modification| | (gateway 1) | +-------------+ | V Local Parameters -- Information flow: local directives & session descriptions -- Arango, et al. Informational [Page 22] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1.3.4. Special case of local connections Large gateways include a large number of endpoints which are often of different types. In some networks, we may often have to set-up connections between endpoints that are located within the same gateway. Examples of such connections may be: * Connecting a trunk line to a wiretap device, * Connecting a call to an Interactive Voice-Response unit, * Connecting a call to a Conferencing unit, * Routing a call from on endpoint to another, something often described as a "hairpin" connection. Local connections are much simpler to establish than network connections. In most cases, the connection will be established through some local interconnecting device, such as for example a TDM bus. When two endpoints are managed by the same gateway, it is possible to specify the connection in a single command that conveys the name of the two endpoints that will be connected. The command is essentially a "Create Connection" command which includes the name of the second endpoint in lieu of the "remote session description." 2.1.4. Names of Call Agents and other entities The media gateway control protocol has been designed to allow the implementation of redundant Call Agents, for enhanced network reliability. This means that there is no fixed binding between entities and hardware platforms or network interfaces. Reliability can be improved by the following precautions: * Entities such as endpoints or Call Agents are identified by their domain name, not their network addresses. Several addresses can be associated with a domain name. If a command or a response cannot be forwarded to one of the network addresses, implementations should retry the transmission using another address. * Entities may move to another platform. The association between a logical name (domain name) and the actual platform are kept in the domain name service. Call Agents and Gateways should keep track of the time-to-live of the record they read from the DNS. They should query the DNS to refresh the information if the time to live has expired. Arango, et al. Informational [Page 23] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 In addition to the indirection provided by the use of domain names and the DNS, the concept of "notified entity" is central to reliability and fail-over in MGCP. The "notified entity" for an endpoint is the Call Agent currently controlling that endpoint. At any point in time, an endpoint has one, and only one, "notified entity" associated with it, and when the endpoint needs to send a command to the Call Agent, it MUST send the command to the current "notified entity" for which endpoint(s) the command pertains. Upon startup, the "notified entity" MUST be set to a provisioned value. Most commands sent by the Call Agent include the ability to explicitly name the "notified entity" through the use of a "NotifiedEntity" parameter. The "notified entity" will stay the same until either a new "NotifiedEntity" parameter is received or the endpoint reboots. If the "notified entity" for an endpoint is empty or has not been set explicitly, the "notified entity" will then default to the source address of the last connection handling command or notification request received for the endpoint. Auditing will thus not change the "notified entity." 2.1.5. Digit maps The Call Agent can ask the gateway to collect digits dialed by the user. This facility is intended to be used with residential gateways to collect the numbers that a user dials; it may also be used with trunking gateways and access gateways alike, to collect the access codes, credit card numbers and other numbers requested by call control services. An alternative procedure is for the gateway to notify the Call Agent of the dialed digits, as soon as they are dialed. However, such a procedure generates a large number of interactions. It is preferable to accumulate the dialed numbers in a buffer, and to transmit them in a single message. The problem with this accumulation approach, however, is that it is hard for the gateway to predict how many numbers it needs to accumulate before transmission. For example, using the phone on our desk, we can dial the following numbers: Arango, et al. Informational [Page 24] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 _______________________________________________________ | 0 | Local operator | | 00 | Long distance operator | | xxxx | Local extension number | | 8xxxxxxx | Local number | | #xxxxxxx | Shortcut to local number at| | | other corporate sites | | *xx | Star services | | 91xxxxxxxxxx | Long distance number | | 9011 + up to 15 digits| International number | |________________________|_____________________________| The solution to this problem is to load the gateway with a digit map that correspond to the dial plan. This digit map is expressed using a syntax derived from the Unix system command, egrep. For example, the dial plan described above results in the following digit map: (0T| 00T|[1-7]xxx|8xxxxxxx|#xxxxxxx|*xx|91xxxxxxxxxx|9011x.T) The formal syntax of the digit map is described by the DigitMap rule in the formal syntax description of the protocol (section 3.4). A Digit-Map, according to this syntax, is defined either by a "string" or by a list of strings. Each string in the list is an alternative numbering scheme, specified either as a set of digits or timers, or as regular expression. A gateway that detects digits, letters or timers will: 1) Add the event parameter code as a token to the end of an internal state variable called the "current dial string" 2) Apply the current dial string to the digit map table, attempting a match to each regular expression in the Digit Map in lexical order 3) If the result is under-qualified (partially matches at least one entry in the digit map), do nothing further. If the result matches, or is over-qualified (i.e. no further digits could possibly produce a match), send the current digit string to the Call Agent. A match, in this specification, can be either a "perfect match," exactly matching one of the specified alternatives, or an impossible match, which occur when the dial string does not match any of the alternative. Unexpected timers, for example, can cause "impossible matches." Both perfect matches and impossible matches trigger notification of the accumulated digits. Digit maps are provided to the gateway by the Call Agent, whenever the Call Agent instructs the gateway to listen for digits. Arango, et al. Informational [Page 25] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.1.6. Names of events The concept of events and signals is central to MGCP. A Call Agent may ask to be notified about certain events occurring in an endpoint, e.g. off-hook events, and a call agent may request certain signals to be applied to an endpoint, e.g. dial-tone. Events and signals are grouped in packages within which they share the same namespace which we will refer to as event names in the following. Packages are groupings of the events and signals supported by a particular type of endpoint. For instance, one package may support a certain group of events and signals for analog access lines, and another package may support another group of events and signals for video lines. One or more packages may exist for a given endpoint-type. Event names are case insensitive and are composed of two logical parts, a package name and an event name. Both names are strings of letters, hyphens and digits, with the restriction that hyphens shall never be the first or last characters in a name. Package or event names are not case sensitive - values such as "hu", "Hu", "HU" or "hU" should be considered equal. Examples of package names are "D" (DTMF), "M" (MF), "T" (Trunk) or "L" (Line). Examples of event names can be "hu" (off hook or "hang- up" transition), "hf" (flash hook) or "0" (the digit zero). In textual representations, the package name, when present, is separated from the event name by a slash ("/"). The package name is in fact optional. Each endpoint-type has a default package associated with it, and if the package name is excluded from the event name, the default package name for that endpoint-type is assumed. For example, for an analog access line, the following two event names are equal: l/dl dial-tone in the line package for an analog access line. dl dial-tone in the line package (default) for an analog access line. This document defines a basic set of package names and event names. Additional package names and event names can be registered with the IANA. A package definition shall define the name of the package, and the definition of each event belonging to the package. The event definition shall include the precise name of the event (i.e., the code used in MGCP), a plain text definition of the event, and, when appropriate, the precise definition of the corresponding signals, for example the exact frequencies of audio signal such as dial tones or DTMF tones. Arango, et al. Informational [Page 26] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 In addition, implementers can gain experience by using experimental packages. The names of experimental packages must start with the two characters "x-"; the IANA shall not register package names that start with these characters. Digits, or letters, are supported in many packages, notably "DTMF" and "MF". Digits and letters are defined by the rules "Digit" and "Letter" in the definition of digit maps. This definition refers to the digits (0 to 9), to the asterisk or star ("*") and orthotrope, number or pound sign ("#"), and to the letters "A", "B", "C" and "D", as well as the timer indication "T". These letters can be combined in "digit string" that represent the keys that a user punched on a dial. In addition, the letter "X" can be used to represent all digits, and the sign "$" can be used in wildcard notations. The need to easily express the digit strings has a consequence on the form of event names: An event name that does not denote a digit should always contain at least one character that is neither a digit, nor one of the letters A, B, C, D, T or X. (Such names should not contain the special signs "*", "#", "/" or "$".) A Call Agent may often have to ask a gateway to detect a group of events. Two conventions can be used to denote such groups: * The wildcard convention can be used to detect any event belonging to a package, or a given event in many packages, or event any event in any package supported by the gateway. * The regular expression Range notation can be used to detect a range of digits. The star sign (*) can be used as a wildcard instead of a package name, and the keyword "all" can be used as a wildcard instead of an event name: A name such as "foo/all" denotes all events in package "foo" A name such as "*/bar" denotes the event "bar" in any package supported by the gateway The names "*" or "*/all" denote all events supported by the gate way. The call agent can ask a gateway to detect a set of digits or letters either by individually describing those letters, or by using the "range" notation defined in the syntax of digit strings. For example, the call agent can: Arango, et al. Informational [Page 27] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Use the letter "x" to denote "any letter or digit." Use the notation "[0-9#]" to denote the digits 0 to 9 and the pound sign. In some cases, Call Agents will request the gateway to generate or detect events on connections rather than on the end point itself. For example, gateways may be asked to provide a ringback tone on a connection. When an event shall be applied on a connection, the name of the connection is added to the name of the event, using an "at" sign (@) as a delimiter, as in: G/rt@0A3F58 The wildcard character "*" (star) can be used to denote "all connections". When this convention is used, the gateway will generate or detect the event on all the connections that are connected to the endpoint. An example of this convention could be: R/qa@* The wildcard character "$" can be used to denote "the current connection." It should only be used by the call agent, when the event notification request is "encapsulated" within a command creation or modification command. When this convention is used, the gateway will generate or detect the event on the connection that is currently being created or modified. An example of this convention is: G/rt@$ The connection id, or a wildcard replacement, can be used in conjunction with the "all packages" and "all events" conventions. For example, the notation: */all@* can be used to designate all events on all connections. Events and signals are described in packages. The package description must provide, for each events, the following informations: * The description of the event and its purpose, which should mean the actual signal that is generated by the client (i.e., xx ms FSK tone) as well as the resulting user observed result (i.e., MW light on/off). * The detailed characteristics of the event, such as for example frequencies and amplitude of audio signals, modulations and repetitions, Arango, et al. Informational [Page 28] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 * The typical and maximum duration of the event. Signals are divided into different types depending on their behavior: * On/off (OO) Once applied, these signals last forever until they are turned off. This may happen either as the result of an event or a new SignalRequests (see later). * Time-out (TO) Once applied, these signals last until they are either turned off (by an event or SignalRequests) or a signal specific period of time has elapsed. Depending on package specifications, a signal that times out may generate an "operation complete" event. * Brief (BR) The duration of these signals is so short, that they stop on their own. If an event occurs the signal will not stop, however if a new SignalRequests is applied, the signal will stop. (Note: this point should be debated. One could make a case that events such as strings of DTMF digits should in fact be allowed to complete.) TO signals are normally used to alert the endpoints' users, to signal them that they are expected to perform a specific action, such as hang down the phone (ringing). Transmission of these signals should typically be interrupted as soon as the first of the requested events has been produced. Package descriptions should describe, for all signals, their type (OO, TO, BR). They should also describe the maximum duration of the TO signals. 2.2. Usage of SDP The Call Agent uses the MGCP to provision the gateways with the description of connection parameters such as IP addresses, UDP port and RTP profiles. These descriptions will follow the conventions delineated in the Session Description Protocol which is now an IETF proposed standard, documented in RFC 2327. SDP allows for description of multimedia conferences. This version limits SDP usage to the setting of audio circuits and data access circuits. The initial session descriptions contain the description of exactly one media, of type "audio" for audio connections, "nas" for data access. Arango, et al. Informational [Page 29] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.3. Gateway Control Commands This section describes the commands of the MGCP. The service consists of connection handling and endpoint handling commands. There are nine commands in the protocol: * The Call Agent can issue an EndpointConfiguration command to a gateway, instructing the gateway about the coding characteristics expected by the "line-side" of the endpoint. * The Call Agent can issue a NotificationRequest command to a gateway, instructing the gateway to watch for specific events such as hook actions or DTMF tones on a specified endpoint . * The gateway will then use the Notify command to inform the Call Agent when the requested events occur. * The Call Agent can use the CreateConnection command to create a connection that terminates in an "endpoint" inside the gateway. * The Call Agent can use the ModifyConnection command to change the parameters associated to a previously established connection. * The Call Agent can use the DeleteConnection command to delete an existing connection. The DeleteConnection command may also be used by a gateway to indicate that a connection can no longer be sustained. * The Call Agent can use the AuditEndpoint and AuditConnection commands to audit the status of an "endpoint" and any connections associated with it. Network management beyond the capabilities provided by these commands are generally desirable, e.g. information about the status of the gateway. Such capabilities are expected to be supported by the use of the Simple Network Management Protocol (SNMP) and definition of a MIB which is outside the scope of this specification. * The Gateway can use the RestartInProgress command to notify the Call Agent that the gateway, or a group of endpoints managed by the gateway, is being taken out of service or is being placed back in service. These services allow a controller (normally, the Call Agent) to instruct a gateway on the creation of connections that terminate in an "endpoint" attached to the gateway, and to be informed about events occurring at the endpoint. An endpoint may be for example: Arango, et al. Informational [Page 30] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 * A specific trunk circuit, within a trunk group terminating in a gateway, * A specific announcement handled by an announcement server. Connections are grouped into "calls". Several connections, that may or may not belong to the same call, can terminate in the same endpoint . Each connection is qualified by a "mode" parameter, which can be set to "send only" (sendonly), "receive only" (recvonly), "send/receive" (sendrecv), "conference" (confrnce), "data", "inactive" (inactive), "loopback", "continuity test" (conttest), "network loop back" (netwloop) or "network continuity test" (netwtest). The handling of the audio signals received on these connections is determined by the mode parameters: * Audio signals received in data packets through connections in "receive", "conference" or "send/receive" mode are mixed and sent to the endpoint. * Audio signals originating from the endpoint are transmitted over all the connections whose mode is "send", "conference" or "send/receive." * In addition to being sent to the endpoint, audio signals received in data packets through connections in "conference" mode are replicated to all the other connections whose mode is "conference." The "loopback" and "continuity test" modes are used during maintenance and continuity test operations. There are two flavors of continuity test, one specified by ITU and one used in the US. In the first case, the test is a loopback test. The originating switch will send a tone (the go tone) on the bearer circuit and expect the terminating switch to loopback the circuit. If the originating switch sees the same tone returned (the return tone), the COT has passed. If not, the COT has failed. In the second case, the go and return tones are different. The originating switch sends a certain go tone. The terminating switch detects the go tone, it asserts a different return tone in the backwards direction. When the originating switch detects the return tone, the COT is passed. If the originating switch never detects the return tone, the COT has failed. If the mode is set to "loopback", the gateway is expected to return the incoming signal from the endpoint back into that same endpoint. This procedure will be used, typically, for testing the continuity of trunk circuits according to the ITU specifications. Arango, et al. Informational [Page 31] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 If the mode is set to "continuity test", the gateway is informed that the other end of the circuit has initiated a continuity test procedure according to the GR specification. The gateway will place the circuit in the transponder mode required for dual-tone continuity tests. If the mode is set to "network loopback", the audio signals received from the connection will be echoed back on the same connection. If the mode is set to "network continuity test", the gateway will process the packets received from the connection according to the transponder mode required for dual-tone continuity test, and send the processed signal back on the connection. 2.3.1. EndpointConfiguration The EndpointConfiguration commands are used to specify the encoding of the signals that will be received by the endpoint. For example, in certain international telephony configurations, some calls will carry mu-law encoded audio signals, while other will use A-law. The Call Agent will use the EndpointConfiguration command to pass this information to the gateway. The configuration may vary on a call by call basis, but can also be used in the absence of any connection. ReturnCode <-- EndpointConfiguration( EndpointId, BearerInformation) EndpointId is the name for the endpoint in the gateway where EndpointConfiguration executes, as defined in section 2.1.1. The "any of" wildcard convention shall not be used. If the "all of" wildcard convention is used, the command applies to all the endpoint whose name matches the wildcard. BearerInformation is a parameter defining the coding of the data received from the line side. These information is encoded as a list of sub-parameters. The only sub-parameter defined in this version of the specification is the encoding method, whose values can be set to "A-law" and "mu-law". ReturnCode is a parameter returned by the gateway. It indicates the outcome of the command and consists of an integer number optionally followed by commentary. Arango, et al. Informational [Page 32] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 2.3.2. NotificationRequest The NotificationRequest commands are used to request the gateway to send notifications upon the occurrence of specified events in an endpoint. For example, a notification may be requested for when a gateway detects that an endpoint is receiving tones associated with fax communication. The entity receiving this notification may decide to use a different type of encoding method in the connections bound to this endpoint. ReturnCode <-- NotificationRequest( EndpointId, [NotifiedEntity,] [RequestedEvents,] RequestIdentifier, [DigitMap,] [SignalRequests,] [QuarantineHandling,] [DetectEvents,] [encapsulated EndpointConfiguration]) EndpointId is the name for the endpoint in the gateway where NotificationRequest executes, as defined in section 2.1.1. NotifiedEntity is an optional parameter that specifies where the notifications should be sent. When this parameter is absent, the notifications should be sent to the originator of the NotificationRequest. RequestIdentifier is used to correlate this request with the notifications that it triggers. RequestedEvents is a list of events that the gateway is requested to detect and report. Such events include, for example, fax tones, continuity tones, or on-hook transition. To each event is associated an action, which can be: * Notify the event immediately, together with the accumulated list of observed events, * Swap audio, * Accumulate the event in an event buffer, but don't notify yet, * Accumulate according to Digit Map, * Keep Signal(s) active, Arango, et al. Informational [Page 33] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 * process the Embedded Notification Request, * Ignore the event. Some actions can be combined. In particular: * The "swap audio" action can be combined with "Notify", "Accumulate" and "Ignore." * The "keep signal active" action can be combined with "Notify", "Accumulate", "Accumulate according to Digit Map", "Ignore" and "Embedded Notification Request." * The "Embedded Notification Request" can be combined with "Accumulate" and with "Keep signals active." It can also be combined with Notify, if the gateway is allowed to issue several Notify commands in response to a single Notification request. In addition to the requestedEvents parameter specified in the command, some profiles of MGCP have introduced the concept of "persistent events." According to such profiles, the persistent event list is configured in the endpoint, by means outside the scope of MGCP. The basic MGCP specification does not specify any persistent event. If a persistent event is not included in the list of RequestedEvents, and the event occurs, the event will be detected anyway, and processed like all other events, as if the persistent event had been requested with a Notify action. Thus, informally, persistent events can be viewed as always being implicitly included in the list of RequestedEvents with an action to Notify, although no glare detection, etc., will be performed. Non-persistent events are those events explicitly included in the RequestedEvents list. The (possibly empty) list of requested events completely replaces the previous list of requested events. In addition to the persistent events, only the events specified in the requested events list will be detected by the endpoint. If a persistent event is included in the RequestedEvents list, the action specified will then replace the default action associated with the event for the life of the RequestedEvents list, after which the default action is restored. For example, if "Ignore off-hook" was specified, and a new request without any off-hook instructions were received, the default "Notify off-hook" operation then would be restored. A given event MUST NOT appear more than once in a RequestedEvents. Arango, et al. Informational [Page 34] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 The gateway will detect the union of the persistent events and the requested events. If an event is not specified in either list, it will be ignored. The Swap Audio action can be used when a gateway handles more than one active connection on an endpoint. This will be the case for three-way calling, call waiting, and possibly other feature scenarios. In order to avoid the round-trip to the Call Agent when just changing which connection is attached to the audio functions of the endpoint, the NotificationRequest can map an event (usually hook flash, but could be some other event) to a local function swap audio, which selects the "next" connection in a round robin fashion. If there is only one connection, this action is effectively a no-op. If signal(s) are desired to start when an event being looked for occurs, the "Embedded NotificationRequest" action can be used. The embedded NotificationRequest may include a new list of RequestedEvents, SignalRequests and a new digit map as well. The semantics of the embedded NotificationRequest is as if a new NotificationRequest was just received with the same NotifiedEntity, and RequestIdentifier. When the "Embedded NotificationRequest" is activated, the "current dial string" will be cleared; the list of observed events and the quarantine buffer will be unaffected. MGCP implementations shall be able to support at least one level of embedding. An embedded NotificationRequest that respects this limitation shall not contain another Embedded NotificationRequest. DigitMap is an optional parameter that allows the Call Agent to provision the gateways with a digit map according to which digits will be accumulated. If this optional parameter is absent, the previously defined value is retained. This parameter must be defined, either explicitly or through a previous command, if the RequestedEvent parameters contain an request to "accumulate according to the digit map." The collection of these digits will result in a digit string. The digit string is initialized to a null string upon reception of the NotificationRequest, so that a subsequent notification only returns the digits that were collected after this request. Digits that were accumulated according to the digit map are reported as any other accumulated event, in the order in which they occur. It is therefore possible that other events be accumulated may be found in between the list of digits. SignalRequests is a parameter that contains the set of signals that the gateway is asked to apply to the endpoint, such as, for example ringing, or continuity tones. Signals are identified by their name, which is an event name, and may be qualified by parameters. Arango, et al. Informational [Page 35] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 The action triggered by the SignalRequests is synchronized with the collection of events specified in the RequestedEvents parameter. For example, if the NotificationRequest mandates "ringing" and the event request ask to look for an "off-hook" event, the ringing shall stop as soon as the gateway detect an off hook event. The formal definition is that the generation of all "Time Out" signals shall stop as soon as one of the requested events is detected, unless the "Keep signals active" action is associated to the specified event. The specific definition of actions that are requested via these SignalRequests, such as the duration of and frequency of a DTMF digit, is out side the scope of MGCP. This definition may vary from location to location and hence from gateway to gateway. The RequestedEvents and SignalRequests refer to the same event definitions. In one case, the gateway is asked to detect the occurrence of the event, and in the other case it is asked to generate it. The specific events and signals that a given endpoint can detect or perform are determined by the list of event packages that are supported by that end point. Each package specifies a list of events and actions that can be detected or performed. A gateway that is requested to detect or perform an event belonging to a package that is not supported by the specified endpoint shall return an error. When the event name is not qualified by a package name, the default package name for the end point is assumed. If the event name is not registered in this default package, the gateway shall return an error. The Call Agent can send a NotificationRequest whose requested signal list is empty. It will do so for example when tone generation should stop. The optional QuarantineHandling parameter specifies the handling of "quarantine" events, i.e. events that have been detected by the gateway before the arrival of this NotificationRequest command, but have not yet been notified to the Call Agent. The parameter provides a set of handling options: * whether the quarantined events should be processed or discarded (the default is to process them.) * whether the gateway is expected to generate at most one notification (step by step), or multiple notifications (loop), in response to this request (the default is exactly one.) When the parameter is absent, the default value is assumed. Arango, et al. Informational [Page 36] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 We should note that the quarantine-handling parameter also governs the handling of events that were detected but not yet notified when the command is received. DetectEvents is an optional parameter that specifies a list of events that the gateway is requested to detect during the quarantine period. When this parameter is absent, the events that should be detected in the quarantine period are those listed in the last received DetectEvents list. In addition, the gateway should also detect the events specified in the request list, including those for which the "ignore" action is specified. Some events and signals, such as the in-line ringback or the quality alert, are performed or detected on connections terminating in the end point rather than on the endpoint itself. The structure of the event names allow the Call Agent to specify the connection (or connections) on which the events should be performed or detected. The command may carry an encapsulated EndpointConfiguration command, that will apply to the same endpoint. When this command is present, the parameters of the EndpointConfiguration command are inserted after the normal parameters of the NotificationRequest, with the exception of the EndpointId, which is not replicated. The encapsulated EndpointConfiguration command shares the fate of the NotificationRequest command. If the NotificationRequest is rejected, the EndpointConfiguration is not executed. ReturnCode is a parameter returned by the gateway. It indicates the outcome of the command and consists of an integer number optionally followed by commentary. .NH 3 Notifications Notifications are sent via the Notify command and are sent by the gateway when the observed events occur. ReturnCode <-- Notify( EndpointId, [NotifiedEntity,] RequestIdentifier, ObservedEvents) EndpointId is the name for the endpoint in the gateway which is issuing the Notify command, as defined in section 2.1.1. The identifier should be a fully qualified endpoint identifier, including the domain name of the gateway. The local part of the name shall not use the wildcard convention. Arango, et al. Informational [Page 37] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 NotifiedEntity is an optional parameter that identifies the entity to which the notifications is sent. This parameter is equal to the last received value of the NotifiedEntity parameter. The parameter is absent if there was no such parameter in the triggering request. The notification is sent to the "current notified entity" or, if no such entity was ever specified, to the address from which the request was received. RequestIdentifier is parameter that repeats the RequestIdentifier parameter of the NotificationRequest that triggered this notification. It is used to correlate this notification with the request that triggered it. ObservedEvents is a list of events that the gateway detected. A single notification may report a list of events that will be reported in the order in which they were detected. The list may only contain the identification of events that were requested in the RequestedEvents parameter of the triggering NotificationRequest. It will contain the events that were either accumulated (but not notified) or treated according to digit map (but no match yet), and the final event that triggered the detection or provided a final match in the digit map. ReturnCode is a parameter returned by the call agent. It indicates the outcome of the command and consists of an integer number optionally followed by commentary. 2.3.3. CreateConnection This command is used to create a connection between two endpoints. ReturnCode, ConnectionId, [SpecificEndPointId,] [LocalConnectionDescriptor,] [SecondEndPointId,] [SecondConnectionId] <--- CreateConnection(CallId, EndpointId, [NotifiedEntity,] [LocalConnectionOptions,] Mode, [{RemoteConnectionDescriptor | SecondEndpointId}, ] [Encapsulated NotificationRequest,] [Encapsulated EndpointConfiguration]) Arango, et al. Informational [Page 38] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 A connection is defined by its endpoints. The input parameters in CreateConnection provide the data necessary to build a gateway's "view" of a connection. CallId is a globally unique parameter that identifies the call (or session) to which this connection belongs. Connections that belong to the same call share the same call-id. The call-id can be used to identify calls for reporting and accounting purposes. It does not affect the handling of connections by the gateway. EndpointId is the identifier for the connection endpoint in the gateway where CreateConnection executes. The EndpointId can be fully-specified by assigning a value to the parameter EndpointId in the function call or it may be under-specified by using the "anyone" wildcard convention. If the endpoint is underspecified, the endpoint identifier will be assigned by the gateway and its complete value returned in the SpecificEndPointId parameter of the response. The NotifiedEntity is an optional parameter that specifies where the Notify or DeleteConnection commands should be sent. If the parameter is absent, the Notify or DeleteConnection commands should be sent to the last received Notified Entity, or to originator of the CreateConnection command if no Notified Entity was ever received for the end point. LocalConnectionOptions is a parameter used by the Call Agent to direct the handling of the connection by the gateway. The fields contained in LocalConnectionOptions are the following: * Encoding Method, * Packetization period, * Bandwidth, * Type of Service, * Usage of echo cancellation, * Usage of silence suppression or voice activity detection, * Usage of signal level adaptation and noise level reduction, or "gain control." * Usage of reservation service, * Usage of RTP security, Arango, et al. Informational [Page 39] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 * Type of network used to carry the connection. This set of field can be completed by vendor specific optional or mandatory extensions. The encoding of the first three fields, when they are present, will be compatible with the SDP and RTP profiles: * The encoding method shall be specified by using one or several valid encoding names, as defined in the RTP AV Profile or registered with the IANA. * The packetization period is encoded as either the length of time in milliseconds represented by the media in a packet, as specified in the "ptime" parameter of SDP, or as a range value, specifying both the minimum and maximum acceptable packetization periods. * The bandwidth is encoded as either a single value or a range, expressed as an integer number of kilobit per seconds. For each of the first three fields, the Call Agent has three options: * It may state exactly one value, which the gateway will then use for the connection, * It may provide a loose specification, such as a list of allowed encoding methods or a range of packetization periods, * It may simply provide a bandwidth indication, leaving the choice of encoding method and packetization period to the gateway. The bandwidth specification shall not contradict the specification of encoding methods and packetization period. If an encoding method is specified, then the gateway is authorized to use it, even if it results in the usage of a larger bandwidth than specified. The LocalConnectionOptions parameter may be absent in the case of a data call. The Type of Service specifies the class of service that will be used for the connection. When the connection is transmitted over an IP network, the parameters encodes the 8-bit type of service value parameter of the IP header. When the Type of Service is not specified, the gateway shall use a default or configured value. The gateways can be instructed to perform a reservation, for example using RSVP, on a given connection. When a reservation is needed, the call agent will specify the reservation profile that should be used, which is either "controlled load" or "guaranteed service." The Arango, et al. Informational [Page 40] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 absence of reservation can be indicated by asking for the "best effort" service, which is the default value of this parameter. When reservation has been asked on a connection, the gateway will: * start emitting RSVP "PATH" messages if the connection is in "send-only", "send-receive", "conference", "network loop back" or "network continuity test" mode (if a remote connection descriptor has been received,) * start emitting RSVP "RESV" messages as soon as it receives "PATH" messages if the connection is in "receive-only", "send-receive", "conference", "network loop back" or "network continuity test" mode. The RSVP filters will be deduced from the characteristics of the connection. The RSVP resource profiles will be deduced from the connection's bandwidth and packetization period. By default, the telephony gateways always perform echo cancellation. However, it is necessary, for some calls, to turn off these operations. The echo cancellation parameter can have two values, "on" (when the echo cancellation is requested) and "off" (when it is turned off.) The telephony gateways may perform gain control, in order to adapt the level of the signal. However, it is necessary, for example for modem calls, to turn off this function. The gain control parameter may either be specified as "automatic", or as an explicit number of decibels of gain. The default is to not perform gain control, which is equivalent to specifying a gain of 0 decibels. The telephony gateways may perform voice activity detection, and avoid sending packets during periods of silence. However, it is necessary, for example for modem calls, to turn off this detection. The silence suppression parameter can have two values, "on" (when the detection is requested) and "off" (when it is turned off.) The default is "off." The Call agent can request the gateway to enable encryption of the audio Packets. It does so by providing an key specification, as specified in RFC 2327. By default, encryption is not used. The Call Agent may instruct the gateway to prepare the connection on a specified type of network. The type of network is encoded as in the "connection-field" parameter of the SDP standard. Possible values are IN (Internet), ATM and LOCAL. The parameter is optional; if absent, the network is determined by the type of gateway. Arango, et al. Informational [Page 41] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 RemoteConnectionDescriptor is the connection descriptor for the remote side of a connection, on the other side of the IP network. It includes the same fields as in the LocalConnectionDescriptor, i.e. the fields that describe a session according to the SDP standard. This parameter may have a null value when the information for the remote end is not known yet. This occurs because the entity that builds a connection starts by sending a CreateConnection to one of the two gateways involved in it. For the first CreateConnection issued, there is no information available about the other side of the connection. This information may be provided later via a ModifyConnection call. In the case of data connections (mode=data), this parameter describes the characteristics of the data connection. The SecondEndpointId can be used instead of the RemoteConnectionDescriptor to establish a connection between two endpoints located on the same gateway. The connection is by definition a local connection. The SecondEndpointId can be fully- specified by assigning a value to the parameter SecondEndpointId in the function call or it may be under-specified by using the "anyone" wildcard convention. If the secondendpoint is underspecified, the second endpoint identifier will be assigned by the gateway and its complete value returned in the SecondEndPointId parameter of the response. Mode indicates the mode of operation for this side of the connection. The mode are "send", "receive", "send/receive", "conference", "data", "inactive", "loopback", "continuity test", "network loop back" or "network continuity test." The expected handling of these modes is specified in the introduction of the "Gateway Handling Function" section. Some end points may not be capable of supporting all modes. If the command specifies a mode that the endpoint cannot support, and error shall be returned. The gateway returns a ConnectionId, that uniquely identifies the connection within one endpoint, and a LocalConnectionDescriptor, which is a session description that contains information about addresses and RTP ports, as defined in SDP. The LocalConnectionDescriptor is not returned in the case of data connections. The SpecificEndPointId is an optional parameter that identifies the responding endpoint. It can be used when the EndpointId argument referred to a "any of" wildcard name. When a SpecificEndPointId is returned, the Call Agent should use it as the EndpointId value is successive commands referring to this call. Arango, et al. Informational [Page 42] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 When a SecondEndpointId is specified, the command really creates two connections that can be manipulated separately through ModifyConnection and DeleteConnection commands. The response to the creation provides a SecondConnectionId parameter that identifies the second connection. After receiving a "CreateConnection" request that did not include a RemoteConnectionDescriptor parameter, a gateway is in an ambiguous situation. Because it has exported a LocalConnectionDescriptor parameter, it can potentially receive packets. Because it has not yet received the RemoteConnectionDescriptor parameter of the other gateway, it does not know whether the packets that it receives have been authorized by the Call Agent. It must thus navigate between two risks, i.e. clipping some important announcements or listening to insane data. The behavior of the gateway is determined by the value of the Mode parameter: * If the mode was set to ReceiveOnly, the gateway should accept the voice signals and transmit them through the endpoint. * If the mode was set to Inactive, Loopback, Continuity Test, the gateway should refuse the voice signals. * If the mode was set to Network Loopback or Network Continuity Test, the gateway should perform the expected echo or Response. Note that the mode values SendReceive, Conference, Data and SendOnly don't make sense in this situation. They should be treated as errors, and the command should be rejected (Error code 517). The command may optionally contain an encapsulated Notification Request command, in which case a RequestIdentifier parameter will be present, as well as, optionally, the RequestedEvents DigitMap, SignalRequests, QuarantineHandling and DetectEvents parameters. The encapsulated NotificationRequest is executed simultaneously with the creation of the connection. For example, when the Call Agent wants to initiate a call to an residential gateway, it should: * ask the residential gateway to prepare a connection, in order to be sure that the user can start speaking as soon as the phone goes off hook, * ask the residential gateway to start ringing, * ask the residential gateway to notify the Call Agent when the phone goes off-hook. Arango, et al. Informational [Page 43] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 This can be accomplished in a single CreateConnection command, by also transmitting the RequestedEvent parameters for the off hook event, and the SignalRequest parameter for the ringing signal. When these parameters are present, the creation and the NotificationRequests should be synchronized, which means that bothshould be accepted, or both refused. In our example, the CreateConnection may be refused if the gateway does not have sufficient resources, or cannot get adequate resources from the local network access, and the off-hook Notification-Request can be refused in the glare condition, if the user is already off-hook. In this example, the phone should not ring if the connection cannot be established, and the connection should not be established if the user is already off hook. The NotifiedEntity parameter, if present, applies to both the CreateConnection and the NotificationRequest command. It defines the new "notified entity" for the endpoint. The command may carry an encapsulated EndpointConfiguration command, that will apply to the same endpoint. When this command is present, the parameters of the EndpointConfiguration command are inserted after the normal parameters of the CreateConnection with the exception of the EndpointId, which is not replicated. The EndpointConfiguration command may be encapsulated together with an encapsulated NotificationRequest command. The encapsulated EndpointConfiguration command shares the fate of the CreateConnection command. If the CreateConnection is rejected, the EndpointConfiguration is not executed. ReturnCode is a parameter returned by the gateway. It indicates the outcome of the command and consists of an integer number optionally followed by commentary. 2.3.4. ModifyConnection This command is used to modify the characteristics of a gateway's "view" of a connection. This "view" of the call includes both the local connection descriptors as well as the remote connection descriptor. Arango, et al. Informational [Page 44] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 ReturnCode, [LocalConnectionDescriptor] <--- ModifyConnection(CallId, EndpointId, ConnectionId, [NotifiedEntity,] [LocalConnectionOptions,] [Mode,] [RemoteConnectionDescriptor,] [Encapsulated NotificationRequest,] [Encapsulated EndpointConfiguration]) The parameters used are the same as in the CreateConnection command, with the addition of a ConnectionId that identifies the connection within the endpoint. This parameter is returned by the CreateConnection function, as part of the local connection descriptor. It uniquely identifies the connection within the context of the endpoint. The EndpointId should be a fully qualified endpoint identifier. The local name shall not use the wildcard convention. The ModifyConnection command can be used to affect parameters of a connection in the following ways: * Provide information about the other end of the connection, through the RemoteConnectionDescriptor. * Activate or deactivate the connection, by changing the value of the Mode parameter. This can occur at any time during the connection, with arbitrary parameter values. * Change the sending parameters of the connection, for example by switching to a different coding scheme, changing the packetization period, or modifying the handling of echo cancellation. Connections can only be activated if the RemoteConnectionDescriptor has been provided to the gateway. The receive only mode, however, can be activated without the provision of this descriptor. The command will only return a LocalConnectionDescriptor if the local connection parameters, such as RTP ports, were modified. (Usage of this feature is actually for further study.) The command may optionally contain an encapsulated Notification Request command, in which case a RequestIdentifier parameter will be present, as well as, optionnally, the RequestedEvents DigitMap, SignalRequests, QuarantineHandling and DetectEvents parameters. The Arango, et al. Informational [Page 45] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 encapsulated NotificationRequest is executed simultaneously with the modification of the connection. For example, when a connection is accepted, the calling gateway should be instructed to place the circuit in send-receive mode and to stop providing ringing tones. This can be accomplished in a single ModifyConnection command, by also transmitting the RequestedEvent parameters, for the on hook event, and an empty SignalRequest parameter, to stop the provision of ringing tones. When these parameters are present, the modification and the NotificationRequests should be synchronized, which means that both should be accepted, or both refused. The NotifiedEntity parameter, if present, applies to both the ModifyConnection and the NotificationRequest command. The command may carry an encapsulated EndpointConfiguration command, that will apply to the same endpoint. When this command is present, the parameters of the EndpointConfiguration command are inserted after the normal parameters of the ModifyConnection with the exception of the EndpointId, which is not replicated. The EndpointConfiguration command may be encapsulated together with an encapsulated NotificationRequest command. The encapsulated EndpointConfiguration command shares the fate of the ModifyConnection command. If the ModifyConnection is rejected, the EndpointConfiguration is not executed. ReturnCode is a parameter returned by the gateway. It indicates the outcome of the command and consists of an integer number optionally followed by commentary. 2.3.5. DeleteConnection (from the Call Agent) This command is used to terminate a connection. As a side effect, it collects statistics on the execution of the connection. ReturnCode, Connection-parameters <-- DeleteConnection(CallId, EndpointId, ConnectionId, [Encapsulated NotificationRequest,] [Encapsulated EndpointConfiguration]) The endpoint identifier, in this form of the DeleteConnection command, shall be fully qualified. Wildcard conventions shall not be used. Arango, et al. Informational [Page 46] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 In the general case where a connection has two ends, this command has to be sent to both gateways involved in the connection. Some connections, however, may use IP multicast. In this case, they can be deleted individually. After the connection has been deleted, any loopback that has been requested for the connection should be cancelled. When all connections to an endpoint have been deleted, that endpoint should be placed in inactive mode. In response to the DeleteConnection command, the gateway returns a list of parameters that describe the status of the connection. These parameters are: Number of packets sent: The total number of RTP data packets transmitted by the sender since starting transmission on this connection. The count is not reset if the sender changes its synchronization source identifier (SSRC, as defined in RTP), for example as a result of a Modify command. The value is zero if the connection was set in "receive only" mode. Number of octets sent: The total number of payload octets (i.e., not including header or padding) transmitted in RTP data packets by the sender since starting transmission on this connection. The count is not reset if the sender changes its SSRC identifier, for example as a result of a ModifyConnection command. The value is zero if the connection was set in "receive only" mode. Number of packets received: The total number of RTP data packets received by the sender since starting reception on this connection. The count includes packets received from different SSRC, if the sender used several values. The value is zero if the connection was set in "send only" mode. Number of octets received: The total number of payload octets (i.e., not including header or padding) transmitted in RTP data packets by the sender since starting transmission on this connection. The count includes packets received from different SSRC, if the sender used several values. The value is zero if the connection was set in "send only" mode. Arango, et al. Informational [Page 47] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Number of packets lost: The total number of RTP data packets that have been lost since the beginning of reception. This number is defined to be the number of packets expected less the number of packets actually received, where the number of packets received includes any which are late or duplicates. The count includes packets received from different SSRC, if the sender used several values. Thus packets that arrive late are not counted as lost, and the loss may be negative if there are duplicates. The count includes packets received from different SSRC, if the sender used several values. The number of packets expected is defined to be the extended last sequence number received, as defined next, less the initial sequence number received. The count includes packets received from different SSRC, if the sender used several values. The value is zero if the connection was set in "send only" mode. This parameter is omitted if the connection was set in "data" mode. Interarrival jitter: An estimate of the statistical variance of the RTP data packet interarrival time measured in milliseconds and expressed as an unsigned integer. The interarrival jitter J is defined to be the mean deviation (smoothed absolute value) of the difference D in packet spacing at the receiver compared to the sender for a pair of packets. Detailed computation algorithms are found in RFC 1889. The count includes packets received from different SSRC, if the sender used several values. The value is zero if the connection was set in "send only" mode. This parameter is omitted if the connection was set in "data" mode. Average transmission delay: An estimate of the network latency, expressed in milliseconds. This is the average value of the difference between the NTP timestamp indicated by the senders of the RTCP messages and the NTP timestamp of the receivers, measured when this messages are received. The average is obtained by summing all the estimates, then dividing by the number of RTCP messages that have been received. This parameter is omitted if the connection was set in "data" mode. When the gateway's clock is not synchronized by NTP, the latency value can be computed as one half of the round trip delay, as measured through RTCP. When the gateway cannot compute the one way delay or the round trip delay, the parameter conveys a null value. For a detailed definition of these variables, refer to RFC 1889. Arango, et al. Informational [Page 48] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 When the connection was set up over an ATM network, the meaning of these parameters may change: Number of packets sent: The total number of ATM cells transmitted since starting transmission on this connection. Number of octets sent: The total number of payload octets transmitted in ATM cells. Number of packets received: The total number of ATM cells received since starting reception on this connection. Number of octets received: The total number of payload octets received in ATM cells. Number of packets lost: Should be determined as the number of cell losts, or set to zero if the adaptation layer does not enable the gateway to assess losses. Interarrival jitter: Should be understood as the interarrival jitter between ATM cells. Average transmission delay: The gateway may not be able to assess this parameter over an ATM network. It could simply report a null value. When the connection was set up over an LOCAL interconnect, the meaning of these parameters is defined as follows: Number of packets sent: Not significant. Number of octets sent: The total number of payload octets transmitted over the local connection. Number of packets received: Not significant. Number of octets received: The total number of payload octets received over the connection. Number of packets lost: Not significant. A value of zero is assumed. Arango, et al. Informational [Page 49] RFC 2705 Media Gateway Control Protocol (MGCP) October 1999 Interarrival jitter: Not significant. A value of zero is assumed. Average transmission delay: Not significant. A value of zero is assumed. The standard set of connection parameters can be extended by the creation of extension parameters. The command may optionally contain an encapsulated Notification Request command, in which case a RequestIdentifier parameter will be present, as well as, optionnally, the RequestedEvents DigitMap, SignalRequests, QuarantineHandling and DetectEvents parameters. The encapsulated NotificationRequest is executed simultaneously with the deletion of the connection. For example, when a user hang-up is notified, the gateway should be instructed to delete the connection and to start looking for an off hook event. This can be accomplished in a single DeleteConnection command, by also transmitting the RequestedEvent parameters, for the off hook event, and an empty SignalRequest parameter. When these parameters are present, the DeleteConnection and the NotificationRequests should be synchronized, which means that both should be accepted, or both refused. The command may carry an encapsulated EndpointConfiguration command, that will apply to the same endpoint. When this command is present, the parameters of the EndpointConfiguration command are inserted after the normal parameters of the DeleteConnection with the exception of the EndpointId, which is not replicated. The EndpointConfiguration command may be encapsulated together with an encapsulated NotificationRequest command. The encapsulated EndpointConfiguration command shares the fate of the DeleteConnection command. If the Delet