hash
stringlengths
32
32
doc_id
stringlengths
7
13
section
stringlengths
3
121
content
stringlengths
0
3.82M
dde9e2552d661348f67d747426d8c5c3
183 002
6.4 Comfort noise insertion and silence suppression
If a codec has built-in support for silence suppression and comfort noise insertion, the activation or deactivation of these features shall be indicated using the a= line according to RFC 3551 [35] and RFC 4855 [40]. If the selected codec does not have built in support for silence suppression and comfort noise (CN) insertion, the CN payload code (see RFC 3389 [51]) may be included in the media description. EXAMPLE: (for ITU-T Recommendation G.711 [27]): v= 0 c= IN <address type> <connection address> m= audio <port number> RTP/AVP 0 13 a= ptime: 10 If the CN payload is included in the Local Descriptor, the MG shall be prepared to receive CN packets during silence periods. If the CN payload is included in the Remote Descriptor, the MG shall send CN packets during silence periods. Comfort noise analysis, voice activity detection and discontinuous transmission algorithms are outside the scope of the present document.
dde9e2552d661348f67d747426d8c5c3
183 002
6.5 DTMF transmission
When a G.711 codec is used (see ITU-T Recommendation G.711 [27]), Media GateWays (MGWs) shall be able to generate, detect and forward DTMF tones inband. When other codecs are used, the MGC should request the use of the procedures defined in RFC 4733 [37] to send and receive DTMF tones: • If the Local Descriptor sent by the MGC includes the support for RFC 4733 [37], Media GateWays (MGWs) shall be prepared to receive DTMF tones in the form of named events and relay the appropriate audio signal to the physical terminations. • If the Remote Descriptor indicates that RFC 4733 [37] is supported, Media GateWays (MGWs) shall be prepared to relay in the form of named events, any DTMF tone received from the physical terminations. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 72 A Dynamic Payload type shall be used to indicate support of RFC 4733 [37] for DTMF relay. EXAMPLE: v= 0 c= IN <address type> <connection address> m= audio <port number> RTP/AVP 18 110 a= ptime: 10 a= rtpmap: 110 telephone-event/8000 a= fmtp: 110 0-15
dde9e2552d661348f67d747426d8c5c3
183 002
6.6 Call progress tones
Call progress tones shall be sent in-band using a voice codec.
dde9e2552d661348f67d747426d8c5c3
183 002
6.7 Support of G.711 variants
dde9e2552d661348f67d747426d8c5c3
183 002
6.7.1 G.711 encoding law
Media GateWays (MGWs) conforming to the present document are required to support ITU-T Recommendation G.711 [27] A-Law and may also support μ-Law in order to avoid call failure or transcoding in case the remote entity supports μ-Law only. How and where to perform transcoding in IP networks in case both terminals/gateways do not support the same variant is outside the scope of this profile.
dde9e2552d661348f67d747426d8c5c3
183 002
6.7.2 G.711 silence suppression mode
ITU-T Recommendation G.711 [27] -over-IP may be operated with or without silence suppression. In case of silence suppression, comfort noise generation shall be based on ITU-T Recommendation G.711 Appendix II [29]. These features may be enabled/disabled on a per session basis, using the procedure described in clause 6.4.
dde9e2552d661348f67d747426d8c5c3
183 002
6.7.3 G.711 packet loss concealment
ITU-T Recommendation G.711 [27] -over-IP may be operated with or without error loss concealment. Typically is that decision dependent on the IP packet loss rate conditions. ITU-T Recommendation G.711 [27] error loss concealment is based on RTP packet granularity, therefore called as packet loss concealment (PLC). ITU-T Recommendation G.711, Appendix I [28] provides a framework for ITU-T Recommendation G.711 [27] PLC mode.
dde9e2552d661348f67d747426d8c5c3
183 002
6.8 MG-Internal redirection of RTP traffic
There might be situations where both RTP session endpoints will be located in the same media gateway (e.g. for a local call). This is related to the case where the two corresponding H.248 RTP terminations, of a single RTP session, belong to two different H.248 Contexts. If RTP traffic turnaround is not supported by the edge routers, it is recommended that the MG try to redirect internally the corresponding RTP/RTCP bearer traffic. This function is related to routing and forwarding of IP packet traffic. The function is therefore also known as MG-embedded IP Router function (figure 2). ETSI ETSI TS 183 002 V3.3.1 (2009-08) 73 Control Subsystem AGW/RGW AGW/RGW Context1 Phy1 RTP1 Context2 Phy2 RTP2 IP Router Data Path IP Forwarding Engine IP Transport Media Gateway Controller Internal IP Router => must be capable of IP Packet Forwarding for IP Packets with identical source & destination IP@s Analogue/ISDN Analogue/ISDN Figure 2: Single RTP session - internal bearer redirection via embedded IP router The MGC may resolve such a two-Context configuration by appropriate H.248 Commands. This may be done in a very early stage, i.e. already during call/bearer establishment phase, or in a later stage during active call phase. 6.9 Residential Media Gateway (RMG) - Keep pinhole open mechanisms for signalling and media traffic in IP domain due to NAT Traversal support
dde9e2552d661348f67d747426d8c5c3
183 002
6.9.1 Problem statement in general
This profile supports residential and access gateways (clause 4.1). The physical MG would be located as Residential Media Gateway (RMG) on customer premises, or as an Access Media Gateway (AMG) on public network side (see figure 1). The media traffic (e.g. RTP-over-IP) and signalling traffic (H.248-over-IP) is transported by the IP network and may face network address translators ("NAT devices"). Any NAT device is separating an IP network in two or more IP address realms (briefly "realm"). There are following network configurations (from H.248 perspective): 1) MG and MGC located in the same IP realm; and - MG and peer IP bearer node located in the same IP realm; or - MG and peer IP bearer node located in the different IP realms; or 2) MG and MGC located in different IP realms; and - MG and peer IP bearer node located in the same IP realm; or - MG and peer IP bearer node located in the different IP realms. Traversal of a NAT device, - so called NAT Traversal support function -, may be thus needed in the IP-based bearer plane and/or IP-based control plane. This clause focuses on the different realms scenarios. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 74
dde9e2552d661348f67d747426d8c5c3
183 002
6.9.2 Problem statement for multiple IP realm network configurations
The different realm configuration is the general case for RGW deployments. Figure 3 illustrates a more detailed network model with a NAT/FW device located at the border between customer network and public network. The RMG is on customer side located, the controlling RMGC is located in the public network. The NAT function is leading to two IP realms (internal and external; or private and public). H.248 signalling is consequently going through the same NAT/FW device as the media traffic. The peering node of the RMG in the transport stratum might be likely a Border Gateway Function (BGF) or possibly other H.248 MG types (like e.g. a RMG, an AMG, TMGW, MRFP), a SIP UA, H.323 terminal, etc. H.248 Residential Media Gateway (RMG) PSTN/ISDN Phy Internal IP network External IP network NAT/FW Residential Media Gateway Controller (RMGC) H.248 Media Gateway (MG) H.248 RAGW Profile H.248 Profile IP Bearer connection (e.g. RTP) Gateway control (H.248) „Media Pinhole“ „Signalling Pinhole“ a) H.248 gateway control traffic b) PSTN call control signalling via H.248 c) ISDN call control signalling via IUA/SCTP/IP Figure 3: Network model - H.248 RMG and RMGC located in different IP realms (The peer IP bearer node might be an H.248 MG as indicated in this example.) There will be two pinholes, which may require keep pinhole open mechanisms (see clause 6.9.3) as NAT Traversal support function. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 75 6.9.3 Possible NAT Traversal support functions when MG and MGC located in different IP realms Figure 4 indicates possible solutions for each pinhole. The signalling pinhole related keep open mechanism is particularly required during the active call phase (i.e. between context creation and release phases). The media pinhole related keep open mechanism may be required in the call/bearer establishment phase, or e.g. in case of suppressed media traffic (e.g. muted microphone). H.248 Residential Media Gateway (RMG) PSTN/ISDN Phy Internal IP network External IP network NAT/FW Residential Media Gateway Controller (RMGC) H.248 Media Gateway (MG) H.248 RAGW Profile (this specification) H.248 Profile IP Bearer connection (e.g. RTP) Gateway control (H.248) „Media Pinhole“ „Signalling Pinhole“ Keep pinhole open mechanism: e.g. H.248 Inactivity Timer Package (it; H.248.14) Keep pinhole open mechanism: e.g. H.248 Keepalive Request Package (kar; H.248.50) Figure 4: Network model - Possible NAT Traversal support functions 6.9.4 Customer access via a Network Address Translation (NAT) using H.248.50 Clause 4.2 provides a functional requirement that where a RGW also provides customer access via a NAT device that "it does not interfere with, and explicitly takes account of, the operation of the H.248 gateway function in the RGW". However in some network architectures there may be an intermediate NAT/s placed between the RGW and the BGF. The address latching function supported by BGF for hosted NAT traversal requires that a BGF receives a RTP packet from the RGW to set the NAT binding. A RGW cannot receive media RTP packets as long as it has not sent at least one RTP packet to create a pinhole through any intermediate NATs. In order to create a pinhole and maintain a NAT binding, the MGC should request the RGW to transmit an RTP keep-alive packet. This is achieved via the H.248.50 "kar" package. If requested to send a keep alive the RGW shall follow the procedures of H.248.50 [i.4], section 9.2.6. The kar package is required for RGW but not necessary for AGW. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 76
dde9e2552d661348f67d747426d8c5c3
183 002
7 Procedures for Physical H.248 terminations
dde9e2552d661348f67d747426d8c5c3
183 002
7.1 General procedures
dde9e2552d661348f67d747426d8c5c3
183 002
7.1.1 Initial configuration
A default digit map shall be provisioned in the MG, so that it can be referred to by name rather than by value.
dde9e2552d661348f67d747426d8c5c3
183 002
7.1.2 DTMF detection
When a series of digit maps is used during called party's number dialling, it is recommended that the values of the timers defined in ITU-T Recommendation H.248.1 [15] be set in such a way that the T timer of the subsequent digit maps be set the same value than the L timer.
dde9e2552d661348f67d747426d8c5c3
183 002
7.1.3 Sending of tones
Signals shall be sent to the exterior of the gateway, according to the following principles: • When sending a tone to the calling party from the originating MG, the signal shall be applied to the physical termination. • When sending a tone to the calling party from the terminating MG, the signal shall be applied to the ephemeral termination. • When sending a tone to the called party from the terminating MG, the signal shall be applied to the physical termination. • When sending a tone to the called party from the originating MG, the signal shall be applied to the ephemeral termination. Table 93 summarizes where usual tones are generated. Table 93: Sending of tones Tone Generation side Perceived by Signal Dial tone Local Calling Party cg/dt Ring tone Local or Remote Calling Party cg/rt Busy tone Local Calling Party cg/bt Release tone Local Calling Party cg/ct Special Dial Tone Local Calling Party xcg/spec Special Information Tone Local or Remote Calling Party cg/sit Call Waiting tone Local Called Party cg/cw or alert/cw (see note) Congestion tone Local or Remote Calling Party cg/ct Caller Waiting Tone Remote Calling Party cg/cr Message Waiting Tone Local Calling Party srvtn/mwt Confirmation Tone Local Calling Party srvtn/conf Negative Acknowledgment Local Calling Party xcg/nack Off-Hook warning tone Local Both xcg/roh Vacant Tone (Number Unobtainable) Local Calling Party xcg/vac NOTE: The call waiting tone may also be embedded in the andisp/dwa signal if associated with display information. In an originating MG, the context topology and the termination modes shall be configured in such a way that in-band information can be received from the remote side before the called party's answer. This setting shall occur not later than the receipt of the ACO message (or an equivalent message from another signalling protocol that ISUP). ETSI ETSI TS 183 002 V3.3.1 (2009-08) 77
dde9e2552d661348f67d747426d8c5c3
183 002
7.1.4 Sending of announcements
Residential Gateways are not required to store recorded announcements nor to support the generic announcement package. Access Gateways may be able to store recorded announcements and shall support the generic announcement package. However, Access Gateways are not required to support variable announcements. When the announcement to be delivered is not available in the gateway and cannot be autonomously retrieved by the gateway from a remote repository, the MGC shall initiate a connection to an external announcement machine, by temporarily adding a termination into the context. Instructions to play announcements are sent directly from the MGC to the server. The announcement server may itself be implemented as an MG, controlled using the H.248 protocol. However, this interface is outside the scope of the present document. In an originating MG, the context topology and the termination modes shall be configured in such a way that in-band information can be received from the remote side before the called party's answer. This setting shall occur not later than the receipt of the ACO message (or an equivalent message from another signalling protocol that ISUP).
dde9e2552d661348f67d747426d8c5c3
183 002
7.1.5 Support of emergency calls
The MGC is responsible for detecting emergency calls and setting the Emergency Call context property when creating the associated context. Prior to the context being created the MG makes no assumption on the priority of the events that take place. The priority context property shall not be used for emergency calls.
dde9e2552d661348f67d747426d8c5c3
183 002
7.1.6 Echo control
Physical terminations are required to support echo cancellation. Echo cancellation is automatically activated on physical terminations by the MG and may be deactivated using the TDM Circuit package (see note 1). NOTE 1: H.248 TDM Circuit Package Version 1 (see ITU-T Recommendation H.248.1 [15], annex E.13) is providing the basic control capabilities for ECDs in H.248 MGs. The SPNE Control Package (see ITU-T Recommendation Q.115.0 [44], clause 7.2) is extending the tdmc Version 1 package by further ECD control possibilities. The SPNE Control Package is not required and is beyond the scope of this version of the profile. Deactivation by the MGC occurs on ISDN physical terminations in case of Unrestricted 64 kbit/s calls and on ANALOG terminations in case the PSTN subscriber line is marked as supporting data calls only. Echo cancellation may also be deactivated by the MG when entering the VBD mode. An Echo Control Device/Function (ECD; see ITU-T Recommendation Q.115.1 [45], clause 3.1) is therefore always associated with a physical H.248 termination (see notes 2 and 3). NOTE 2: A VoIP Media GateWay (MGW) defined by this H.248 Profile is a so-called "type 1 exchange/node" from ECD point of view (see ITU-T Recommendation Q.115.1 [45], clause A.2.4.3.1). NOTE 3: The configuration of the HECD (or ECD) in a media gateway is the "reverse associated" mode (see ITU-T Recommendation Q.115.1 [45], clause A.1.1, note 1 and figures A.2a, A.2b and A.3). An ECD is responsible for a single echo path, therefore also known as half-way ECD (HECD) (see notes 4 and 5). The ITU-T Recommendation G.168 [41] Digital Network ECD is required for the echo generated at "legacy terminal" side of the MG. This is the echo path on which the HECD is intended to operate, called as G.168 Cancelled End (or formerly as Near End). The required ECD tail length capacity is given by the echo path at the cancelled end. NOTE 4: A local call, resulting in a single H.248 Context with two physical H.248 terminations (Phy-to-Phy bearer interworking) may result in the allocation of two HECDs, one per direction. Such two complementary HECDs representing a full-way ECD (called full ECD (FECD), see ITU-T Recommendation Q.115.1 [45], clause 3.7). A FECD configuration for Phy-to-Phy H.248 Context types is not required and supported in this Profile version. It is rather anticipated from MG side, that the MGC is disabling the ECD resources for such a Context type, due to the small end-to-end propagation delay here. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 78 NOTE 5: More detailed ECD notation: In case of an outgoing call the ECD in the originating media gateway, responsible for the hybrid echo generated by the calling party, is playing the role of the outgoing ECD (OECD, see ITU-T Recommendation Q.115.1 [45], clause 3.12). In case of an incoming call the ECD in the terminating media gateway, responsible for the hybrid echo generated by the called party, is playing the role of the incoming ECD (IECD, see ITU-T Recommendation Q.115.1 [45], clause 3.11).
dde9e2552d661348f67d747426d8c5c3
183 002
7.2 Specific procedures for analog lines
Example signalling flows for call-dependent procedures for analog line access are provided by TR 183 040 [i.2].
dde9e2552d661348f67d747426d8c5c3
183 002
7.2.1 Autonomous actions
Normally the MG detects and applies analogue signals to the analogue line under the instructions of the MGC. However, a MG may be provisioned to perform the following time critical autonomous actions: • Apply normal power feed when the analogue line state changes from On-Hook to Off-Hook. • Apply idle voltage/current feed when the analogue line state changes from Off-Hook to On-Hook. • Apply reduced power feed when the analogue line continues to remain Off-Hook after a certain period of time without being associated to any connection. • Remove the ringing when the line goes Off-Hook. This action is performed irrespective of the setting of the keep active flag associated with the "off-hook" event. This action is intended to avoid causing an acoustic shock to the end user. • If a new outgoing call attempt is rejected by the MGC (e.g. due to congestion in MGC) by the return of an error code in response to a Notify command containing an "Off-Hook" observed event, then the RGW/AGW may as an option autonomously apply an appropriate tone or announcement. This will ensure that the end user is not subject to hearing prolonged silence and not knowing the progression of his call attempt. If a tone or announcement is autonomously applied, then the RGW/AGW will then be responsible for performing the subsequent autonomous call cleardown and without any interaction with the MGC. The detailed autonomous call treatment procedures are the same as those defined in ES 283 039-4 [57], for the case where a non emergency call is filtered by the AGW and not admitted to the MGC.
dde9e2552d661348f67d747426d8c5c3
183 002
7.2.2 Alerting
The signal to be used (andisp/dwa or alert/ri) depends on whether or not information need to be displayed to the terminal. See also clause 7.2.4.
dde9e2552d661348f67d747426d8c5c3
183 002
7.2.3 Automatic metering
The pulses in the automatic metering package have to be accurately reflected between the MGC and the MG. In order to achieve that accuracy the following procedures have to be followed. The MG stops the enable metering signal from the automatic metering package in the event of an H.248 control link failure. The MGC is aware that the enable metering signal was stopped if it determines that the H.248 control link is down or if ServiceChange on Root is received indicating disconnected method. The enable metering signals is not reinitiated if the H.248 control link is re-established. The MGC has then the option to reinitialize the enable metering signal and continue charging or not charge for the call anymore. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 79
dde9e2552d661348f67d747426d8c5c3
183 002
7.2.4 Display service
dde9e2552d661348f67d747426d8c5c3
183 002
7.2.4.1 On hook data transmission
If "data transmission during ringing or prior to ringing" as described in EN 300 659-1 [4] is required (e.g. for the calling number display service), then the MGC shall use the andisp/dwa signal without or with the "TAS" parameter. The andisp/dwa signal without "TAS" parameter is used when the default method (i.e. "during ringing" or "prior to ringing" and the corresponding TAS) provisioned (e.g. globally or on a termination-basis) within the MG is to be used. The andisp/dwa signal with a "TAS" parameter is used when the default method provisioned in the MG is to be overridden by the MGC. The indication of "nt" (no TAS) in the TAS parameter informs the MG to apply data transmission during ringing. If "data transmission not associated with ringing" (e.g. for visual message waiting indicator service) as described in EN 300 659-1 [4] is required, then the MGC shall use the andisp/data signal without or with the "TAS" parameter. The former mechanism is used when the default TAS provisioned within the MG is to be used. The latter mechanism is used when the default TAS signal provisioned in the MG is to be overridden by the MGC.
dde9e2552d661348f67d747426d8c5c3
183 002
7.2.4.2 Off hook data transmission:
For Off-Hook display the procedures for shall conform to EN 300 659-2 [5]. If data transmission is invoked after the initial Call Waiting Tone "Subscriber Alert Signal" (e.g. for the calling number display service), then the MGC shall use the andisp/dwa signal without or with the "TAS" parameter. The former mechanism is used when the default TAS provisioned within the MG is to be used. The latter mechanism is used when the default TAS signal provisioned in the MG is to be overridden by the MGC. If data transmission is invoked without the presence of a "Subscriber Alert Signal" (e.g. for the advice of charge service), then the MGC shall use the andisp/data signal without or with the "TAS" parameter. The former mechanism is used when the default TAS provisioned within the MG is to be used. The latter mechanism is used when the default TAS signal provisioned in the MG is to be overridden by the MGC.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3 Specific procedures for ISDN interfaces
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.1 General
Support of ISDN Basic Access, Primary Rate Access and NMDS requires the use of a backhaul mechanism in conjunction with H.248. In particular H.248 shall be used for handling the adaptation of the B channels to RTP media streams, for applying tones and announcements, and for inband DTMF digit collection.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2 ISDN-BA signalling
An AGW or RGW shall support ISDN Q.921 (see ITU-T Recommendation Q.921 [46] User Adaptation (IUA) over SCTP as a backhaul mechanism for transporting D-channel (s-type frames) information to the MGC. It is recommended that the IUA interface identifier (IID) is mapped from the H.248 termination ID. An AGW or RGW shall support either SCTP and IUA (see RFC 2960 [58] and RFC 4233 [12]) or Raw Frame Relay over Generic Routing Encapsulation (see RFC 2784 [38]) as a backhaul mechanism for D-channel p-type frames to the required destination as defined in ETS 300 099 [6]. The Media GateWay (MGW) shall evaluate the DLCI field in the Q.921 message header. For SAPI=16, the Media GateWays (MGWs) shall send the SCTP/IUA messages to the required destination. NOTE 1: For SAPI=0, the Media GateWay (MGW) performs the procedure as described in the first paragraph of this clause. Media GateWays (MGWs) shall support either SCTP and IUA (see RFC 2960 [58] and RFC 4233 [12]) or Raw Frame Relay over Generic Routing Encapsulation (see RFC 2784 [38]) as a backhaul mechanism for D-channel f-type frames to a Frame Relay Gateway to the required destination as defined in ETS 300 099 [6]. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 80 The Media GateWay (MGW) shall evaluate the DLCI field and the Interface Id in the Q.921 message header. For SAPI=16, the Media GateWays (MGWs) shall send the SCTP/IUA messages to the destination, which is pre-provisioned in the MG. NOTE 2: For SAPI=0, the Media GateWay (MGW) performs the procedure as described in the first paragraph of this clause. The LAP-D state machine (including TEI assignment and management procedures) shall reside within the AGW or RGW and shall conform to ETS 300 402-2 [8]. Automatic TEI may only be requested by the terminal equipment (TE). Non-automatic TEI shall be autonomously assigned by the AGW or RGW upon activation of the layer 1 and these values can then be used by either the TE or the MGC. The range of automatic and non-automatic TEI is defined in ETS 300 402-2 [8]. The "point to point" or "point to multi-point" procedures are solely under the control of the MGC and have no impact on the AGW or RGW. The AGW or RGW shall support either permanent activation of the layer 1 or activation of layer 1 on a per call basis and this mode shall be configurable via a management interface. Activation of loop backs within the access digital section (e.g. loopback at the NT1) shall be under the control of the AGW or RGW. When a loopback is applied it will also be necessary for the AGW or RGW to inform the MGC that the ISDN access is unavailable for the presentation of incoming calls. This can be achieved by using the H.248 ServiceChange procedures.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2.1 Discrimination of s-type frames (layer 2 call control signalling)
The discrimination of s-type frames is based on a SAPI value equals to zero.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2.2 Discrimination of p-type frames
The Media GateWays (MGWs) shall evaluate the DLCI field. For the SAPI value equals to 16, the media gateway shall send the X.25 PLP messages over SCTP and IUA to the destination address, which is pre-provisioned for this SAPI value in the media gateway.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2.3 Discrimination of f-type frames
The Media GateWay (MGW) shall evaluate the DLCI field. For the SAPI value range for 32 to 62 inclusive, the Media Gateway (MG) shall send the X.35/X.31 messages over SCTP and IUA to the destination address, which is pre-provisioned for this SAPI-values in the media gateway.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2.4 SCTP Association establishment
The NGN-Packet Handler (NGN-PH) or Frame Handler (FH) plays the role of the Softswitch in a packet mode or frame mode configuration. Therefore, the SCTP association shall be established/re-established by the NGN PH or FH. For signalling (s-type frames) and packet/frame relay (f and p-type frames) two separate SCTP associations shall be used: - for s-type frames, the SCTP association is established between the MG and the Softswitch; - for p-type frames, the SCTP association is established between the MG and the NGN-PH; - for f-type frames, the SCTP association is established between the MG and the NGN-FH. Signalling traffic from one D-channel is transported over one bidirectional SCTP stream within one SCTP association. Signalling traffic from more than one D-channel may be transported over one bidirectional SCTP stream, within an SCTP association. The same may also apply for f- and p-type traffic. NOTE: If, for redundancy, the MG is connected to a secondary Softswitch, there may be also a secondary SCTP association established from the MG to the secondary Softswitch. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 81
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2.4.1 IUA/SCTP for Q.931 call control signalling traffic
The H.248 registration/re-registration between the Media GateWay (MGW) and the Media Gateway Controller (MGC) shall trigger the initiation of the SCTP association between the Media GateWay (MGW) and the Media Gateway Controller (MGC) according to TISPAN TS 102 144 [59]. The H.248 de-registration between the Media GateWay (MGW) and the Media Gateway Controller (MGC) shall trigger the shutdown of the SCTP association between the Media GateWay (MGW) and the Media Gateway Controller (MGC) according to TISPAN TS 102 144 [59].
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.2.4.2 IUA/SCTP for X.25 or frame relay traffic
IUA shall be used asymmetrical as specified in RFC 4233 [12]. The X.25 endpoint, i.e. the NGN-PH shall be a standalone node. It is not the aim of the present document to have two X.25 terminals interconnected via two Access Gateways and an IP domain, but a X.25 terminal connected via an Access Gateway to the NGN-PH.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.3 ISDN-PRA signalling
The RGW and AGW requirements for ISDN-PRA are the same as for an ISDN-BA line as detailed above, with the following exceptions: • "Point to point" procedures are only applicable. • Relaying of p-type and f-type frames is not applicable. Unlike ISDN-BA, any OAM procedures related to the access digital section (e.g. Loopback at the NT1) are handled entirely within the AGW or RGW, where time slot "0" is terminated.
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.4 NMDS
Where the AGW or RGW supports NMDS, there are specific behaviours for the ISDN Basic Access (BA) at both Layer 1 and Layer 2. At Layer 1 permanent activation shall be supported. At layer 2 the range of Automatic TEI values available to be requested by the Terminal Equipment (TE) and TEIs used for the signalling for PSTN ports shall be as defined in EN 301 141-1 [7]. TEIs used for the signalling for the PSTN ports associated with the Network Terminating Node (NTN) shall be operated using "point-to-point" procedures under the control of the MGC. Further details of these Layer 1 and Layer 2 procedures are defined in EN 301 141-1 [7].
dde9e2552d661348f67d747426d8c5c3
183 002
7.3.5 ISDN management
The operation and maintenance of an ISDN-BA access digital section (as defined in ETS 300 297 [47]) and an ISDN-PRA access digital section (as defined in ETS 300 233 [48]) is performed by the MG. The MG controls the access digital section for the whole ISDN access and not individual "B" channels. Therefore the MG shall indicate failure or return to service of the ISDN access digital section to the MGC, via a H.248 ServiceChange. The ServiceChange shall have a TerminationID that specifies the affected ISDN port, but with a wildcarded identifier to indicate that the command applies to the entire ISDN Access. Upon receipt of this ServiceChange the MGC shall assume that the command applies to all the configured "B-channels", as well as the "D-channel", whose signalling is transported via IUA/SCTP. It should be noted that all other H.248 commands (e.g. Add, Modify, Move, Notify, Subtract) sent by the MG/MGC which are associated with the establishment of a bearer shall apply to an individual "B-channel". In addition it is possible to send a ServiceChange identifying a specific B-channel when for example a B-channel is taken in and out-of-service via the MG/MGC Element Management Systems. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 82
dde9e2552d661348f67d747426d8c5c3
183 002
7.4 Specific procedures for V5 interfaces
V5 support is out of scope of an RGW, but may be an optional interface at an AGW.
dde9e2552d661348f67d747426d8c5c3
183 002
7.4.1 General
Support of V5 interfaces (V5.1, V5.2) requires (see note) the use of a backhaul mechanism in conjunction with H.248. In particular H.248 shall be used for handling the adaptation of the 64-kbit/s bearer channels to RTP media streams, for applying tones and announcements, and for inband DTMF digit collection. NOTE: It is recognized that the AGW may also itself terminate the V5 signalling and then map it to IUA (for ISDN) and H.248 (for PSTN).
dde9e2552d661348f67d747426d8c5c3
183 002
7.4.2 V5 signalling
An AGW shall support a signalling gateway according RFC 3807 [53] as a backhaul mechanism for transporting LAPV5 information to the MGC.
dde9e2552d661348f67d747426d8c5c3
183 002
7.4.3 Handling of p- and f-type frames
Handling of p- and f-type frames is according clauses 7.3.2.2 and 7.3.2.3. 8 MG and MGC management and call-independent procedures Generic TISPAN Call-independent procedures are defined in a separate document (TR 183 025 [i.1]), which is an overall description for all ETSI defined H.248 profile specifications, i.e. TR 183 025 [i.1] provides a set of procedures that are available to each profile specification. For this profile, the set of applicable call-independent procedures is primarily given by the supported H.248 Command API capabilities for AuditValue (see clause 5.8.5), AuditCapabilities (see clause 5.8.6) and ServiceChange (see clause 5.8.8), and supported packages (e.g. for overload control), by each profile. In general, all call-independent procedures by TR 183 025 [i.1], as described in of clauses 5.8.5, 5.8.6, 5.8.8 and 5.14, shall be supported. The following clauses providing additional details for some selected call-independent procedures: • MG overload protection by MGC (clause 8.1); • MGC overload protection by MG (clause 8.1); • Monitoring of "quality metrics" for non-Root terminations (clause 8.2); • OAM-driven line tests (clause 8.3); • Measurements and supported H.248 statistics (clause 8.5); and • Conditional reporting capability of H.248 statistics (clause 8.4).
dde9e2552d661348f67d747426d8c5c3
183 002
8.1 Overload control
MG overload control procedures are supported using the Overload Control package (ITU-T Recommendation H.248.11 [19]). In the case of MGC overload there are two mechanisms, the "notification behaviour" and "adaptive rate based" that are available to the MGC in order to regulate the traffic presented to it. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 83
dde9e2552d661348f67d747426d8c5c3
183 002
8.1.1 Notification Behaviour
MGC will give preference to emergency calls and priority lines. Priority lines do not need any special H.248 handling since the MGC has the information and can use regular call setup procedures for those subscribers. Detection of emergency calls requires special handling whilst minimizing the MGC load. The special handling requires support of version 3. At detection of off-hook in the MG a Notify message is delivered to the MGC. The MGC checks its own congestion state in order to determine how to process the call. If the MGC is overloaded then the following procedures may be taken into effect. A Modify command is sent to the applicable termination in NULL context to start dialtone and monitor events which allow detection of emergency calls or non-emergency calls. The MGC uses two digit maps. The first digit map includes only the allowed emergency numbers (i.e. EmergencyDialPlan). The second digit map is used to identify if any non emergency number (i.e. NotEmergencyDialPlan) is dialled and automatically issue congestion tone. The NotifyBehaviour event parameter is set to "NeverNotify" for the second digit map. This ensures that the Notify messages reporting non emergency numbers are suppressed, while Notify messages reporting emergency numbers can progress as in normal conditions. An example of the Modify command is indicated below: Context = - { Modify = aln/1/1/1 { Events = 888 {xdd/xce {DigitMap = EmergencyDialPlan}, ; If emergency report to MGC {xdd/xce {DigitMap = {NotEmergencyDialPlan}, NotifyBehaviour = NeverNotify}}, Embed {Signals {cg/ct}}} ; If any other number issue congestion tone and report to MGC Signals {cg/dt}}} ; This applies dial tone prior to any digits being entered } }
dde9e2552d661348f67d747426d8c5c3
183 002
8.1.2 Adaptive Rate Based
The adaptive rate based mechanism is defined in ES 283 039-4 [57] and allows the MGC to offset the load onto the AGWs during periods of MGC load. This is achieved using the "etsi_nr" package that enables the MGC to specify the "off-hook" rate that it can handle. When the MGC is no longer in overload, then the "Off-Hook" regulation in the AGW can be disabled using the "etsi_nr" package. Further procedural details can be found in ES 283 039-4 [57].
dde9e2552d661348f67d747426d8c5c3
183 002
8.2 IP QoS control and monitoring
The Quality of Service (QoS) of network connections can be monitored using the quality alert event of the network package. It is up to the MGC to set the threshold value that will trigger the notification of this event. The threshold value is expressed as a percentage of measured quality loss. The Media GateWay (MGW) does this by taking into account packet loss, jitter and delay, according to a provisioned algorithm. The Quality Alert Ceasing event of the Quality Alert Ceasing package enables the Media GateWay (MGW) to notify the MGC when the network connections return to an acceptable quality.
dde9e2552d661348f67d747426d8c5c3
183 002
8.3 Testing of analog and digital lines
It shall be possible to trigger "metallic" line testing on physical terminations via the MG OAM interface. This profile does not support H.248-controlled line tests (e.g. line tests defined in ITU-T Recommendation H.248.17 [49]). A service change procedure shall be initiated by the MG, when a termination is placed in test. If the line test is required to be performed immediately, then the MG shall issue a ServiceChange with a method of "forced". The MGC shall not attempt to make any calls on the termination and release any existing context on the termination. If the line test is to be performed after release of any current connections, then the MG shall issue a ServiceChange with a method of "graceful". ETSI ETSI TS 183 002 V3.3.1 (2009-08) 84
dde9e2552d661348f67d747426d8c5c3
183 002
8.4 Real-Time Statistics Reporting
dde9e2552d661348f67d747426d8c5c3
183 002
8.4.1 Overview of conditional reporting
Normally a MGC obtains bearer related statistics through periodic auditing of the H.248 statistic descriptor or at the time of deletion of a stream or subtraction of a termination. However, in both cases, there is a time delay from when a reporting condition occurs on a MG (e.g. a statistic threshold being passed) and the MGC learning of the statistic. In many cases, such a delay is of no consequence. However, in some cases, the MGC may require to be immediately informed of a given statistical threshold condition occurring. In this case, the MGC must use the H.248.47 Statistic Conditional Reporting package [56]. This package may be applied to multiple statistics. The MGC should set the reporting thresholds and ranges as appropriate and must specify at least one "condition" for conditional reporting (i.e. the MGC must signal at least one condition per requested packageID/statisticID item). The exact statistics and reporting conditions are determined by operator configuration based on the application/service required.
dde9e2552d661348f67d747426d8c5c3
183 002
8.4.2 Basic conditional reporting
Basic conditional reporting uses the protocol elements of the Statistic Conditional Reporting package version 1. This allows the definition of many, but limited reporting conditions.
dde9e2552d661348f67d747426d8c5c3
183 002
8.4.3 Extended conditional reporting
Extended conditional reporting uses the protocol elements of the Statistic Conditional Reporting package version 2. This package allows in addition • to control whether a timestamp is reported with the detection of the (conditional reporting) events; and • extends the reporting conditions with value-based metric conditions. 8.5 Measurement of Resource Usage and Performance Metrics - Supported H.248 Statistics
dde9e2552d661348f67d747426d8c5c3
183 002
8.5.1 Usage metering
Usage metering is supported by the statistics defined in the Network (nt), RTP (rtp) and RTP Application Data (rtpad) packages. Such statistics are notified to MGC when a termination is subtracted from a context (e.g. at the end of a session) or due to conditional reporting (see clause 8.4). They provide information about: 1) information about resource usage, e.g.: - the duration of the time a termination has been in a context, - the traffic volume, e.g. number of octets sent and received; 2) information about Grade of Service (GoS)/Quality of Service (GoS), e.g.: - the packet delay variation or packet transfer delay. The "number of octets" excludes all transport overhead (see clause E.11.4/H.248.1 Version 3), i.e. IP header is excluded in case of an IP-based H.248 Termination (see clause E.11.5.1.5/H.248.1). ETSI ETSI TS 183 002 V3.3.1 (2009-08) 85
dde9e2552d661348f67d747426d8c5c3
183 002
8.5.1.1 Traffic Volume related Statistics
Figure 5 provides an overview of different traffic volume related statistics. Traffic volume related statistics are only accessible by the nt and rtp packages in profile versions 1 and 2. Profile version 3 provides additional metrics. Application Level Framing Protocol RTP, UDPTL, … L3: IP L2 L4: UDP, TCP, … L1 e.g. audio codec Transport Protocol Network Protocol Application Data Principle protocol stack for H.248 IP Terminations transport overhead reported traffic volume b c d a e rtpad nt & rtp ipocs Figure 5: Overview of supported statistics - Traffic volume related statistics on different protocol layers
dde9e2552d661348f67d747426d8c5c3
183 002
8.5.1.1.1 RTP Case (general)
Ephemeral terminations in this profile using RTP as application level framing protocol. Traffic volume based statistics may be accessed via the rtp package: • packet granularity: RTP packets sent and/or received; NOTE: Packet level statistics could already provide useful volume measurements in case of RTP packets with constant length). • octet granularity: RTP octets send/received statistics are coupled with nt package statistics, i.e. these statistics are also including RTP padding, RTP header information and UDP transport overheads. Such overhead is excluded in the RTP application data specific statistics (see clause 8.5.1.1.2).
dde9e2552d661348f67d747426d8c5c3
183 002
8.5.1.1.2 RTP Case: application data
The RTP Application Data package provides support for explicit octet count statistics concerning media traffic, i.e. the RTP payload based traffic volume.
dde9e2552d661348f67d747426d8c5c3
183 002
8.5.1.1.3 IP Case: network layer data
The IP layer octets count statistics package provides traffic volume statistics on IP layer for IP version 4 or 6. The H.248 ipocs package (see indication in figure 3) is not supported by this profile and previous versions. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 86
dde9e2552d661348f67d747426d8c5c3
183 002
8.5.2 Statistics for reported metering pulses at analog lines
There is a PSTN supplementary service ("Advice of Charge" (AOC)) for reporting metering pulses towards PSTN terminals from network side. The PSTN AOC service is only applicable for H.248 analog line (ALN) terminations and provided by the amet package. The MG records reported metering pulses by: 1) a cumulative statistic (amet/cpc); and 2) an interval statistic (amet/pcslr). ETSI ETSI TS 183 002 V3.3.1 (2009-08) 87 Annex A (informative): Comparison with H.248 ARGW Profile Versions 1 and 2 A.1 General A.1.1 Version 2 Version 2 provides now all the package usage details, which were missing in profile version 1. A few number of further capabilities were added, see table A.1. The profile was also explicitly linked with TR 183 025 [i.1] with regards to the specification of relevant call-independent procedures. A.1.2 Version 3 Version 3 adds a number of new optional packages concerned with statistics and dial digit method detection plus an enhanced VBD capability. See annex A.3 for further details. A.2 Differences between H.248 ARGW Profiles Version 1 and Version 2 Table A.1 provides an overview of the differences between the H.248 ARGW Profiles Version 1 and Version 2. Table A.1: Difference between H.248 ARGW Profile Versions 1 and 2 Topic ES 283 002 (V1.1.3) [52] H.248 ARGW Profile Version 1 ES 283 002 (V2.1.0) [52] H.248 ARGW Profile Version 2 Required H.248 Version H.248 Version 2 IUA/SCTP encapsulation for Q.921 p-/f-type frames in IP domain Raw Frame Relay over Generic Routing Encapsulation for p-/f-type frames transport only IUA/SCTP based transport option in addition Support of V5 access networks Yes, via IUA (for ISDN) and H.248 (for PSTN) Yes, additionally via V5UA Support of MGC overload control Yes, notification behaviour based Yes, additionally adaptive rate based Packages ETSI Notification Rate Package No Yes Statistic Conditional Reporting Package No Yes ETSI ETSI TS 183 002 V3.3.1 (2009-08) 88 A.3 Differences between H.248 ARGW Profiles Version 2 and Version 3 Table A.2 provides an overview of the differences between the H.248 ARGW Profiles Version 2 and Version 3. Table A.2: Difference between H.248 ARGW Profile Version 2 and the present document Topic ES 283 002 (V2.1.0) [62] H.248 ARGW Profile Version 2 The present document H.248 ARGW Profile Version 3 Clarifications for Automatic Metering Statistics — New clause 8.5.2 on "statistics for reported metering pulses at analog lines" Voiceband Data service partial support of V.152 (see note) V.152 compliant service Packages Keepalive Request Package No Yes RTP Application Data Package No Yes Statistic Conditional Reporting Package V1 only. V2 New clauses 8.4.2 and 8.4.3. Digit Dialling Method Information for Extended Digit Maps Package No Yes NOTE: Two exceptions are allowed. ETSI ETSI TS 183 002 V3.3.1 (2009-08) 89 Annex B (informative): Bibliography • ETSI TR 101 183: "Public Switched Telephone Network (PSTN); Analogue ringing signals". • ETSI EN 300 403-1: "Integrated Services Digital Network (ISDN); Digital Subscriber Signalling System No. one (DSS1) protocol; Signalling network layer for circuit-mode basic call control; Part 1: Protocol specification [ITU-T Recommendation Q.931 (1993), modified]". • IETF RFC 2402: "IP Authentication Header". • IETF RFC 3550: "RTP: A Transport Protocol for Real-Time Applications". • IETF RFC 768: "User Datagram Protocol". ETSI ETSI TS 183 002 V3.3.1 (2009-08) 90 Annex C (informative): Change history Date WG Doc. CR Rev CAT Title / Comment Current Version New Version 26-05-08 17bTD101r1 Input draft from rapporteur with identical technical content with v2.1.0 3.0.0 30-05-08 17bTD102r1 001 B Optional Support of Application-level Statistics 3.0.0 3.0.1 30-05-08 17bTD103r1 002 D Clarifications for Automatic Metering Statistics 3.0.0 3.0.1 30-05-08 17bTD104r1 003 B Clarifications for Conditional Reporting of Statistics and support of Package Version 2 3.0.0 3.0.1 30-05-08 17bTD105r1 004 B Update reference for DiffServ Package 3.0.0 3.0.1 30-05-08 17bTD106r1 005 C Support of VBD service 3.0.0 3.0.1 30-05-08 17bTD107r1 006 B NAT Traversal support 3.0.0 3.0.1 30-05-08 17bTD108r1 007 F Comparison between Profile Versions 3.0.0 3.0.1 30-05-08 17bTD146r3 008 B H.248 ARG profile v3 RTP keep-alive packet 3.0.0 3.0.1 June 08 Creation of change history annex and editorial updates by ETSI Secretariat 3.0.1 3.0.2 July 08 Approval by plenary of CRs 001 to 008 3.0.2 3.1.0 26-11-08 19bTD19 009 r1 F Interworking between V.152-compliant and non-V.152 GWs 3.1.0 3.1.1 25-02-09 20WTD116 010 r2 F Addition of Dialling Method Discrimination Package 3.1.1 3.1.2 25-02-09 20WTD139 011 r1 F Clean-up of Annex A 3.1.1 3.1.2 25-02-09 20WTD140 012 r2 F Clean-up of References and kar package 3.1.1 3.1.2 25-02-09 20WTD141 013 r1 F Package usage overview for rtpad package 3.1.1 3.1.2 25-02-09 20WTD142 014 r1 F Check of package usage details for scr package 3.1.1 3.1.2 25-02-09 20WTD115 015 r2 F Further clarification of VBD handling 3.1.1 3.1.2 25-02-09 20WTD143 016 r2 F Autonomous VBD transitions (con't) 3.1.1 3.1.2 10-03-09 CRs 009 to 016 TB approved at TISPAN#20 3.1.2 3.2.0 18-03-09 20bTD117 017 r1 D Editorials in clause 6 3.2.0 3.2.1 18-03-09 20bTD250 018 r2 F Further clarifications to VBD text (VBD clause 6.2.4) 3.2.0 3.2.1 18-03-09 20bTD247 019 r2 F Further clarifications to VBD text (VBD clause 6.2.1) 3.2.0 3.2.1 Publication 3.2.1 3.3.1 ETSI ETSI TS 183 002 V3.3.1 (2009-08) 91 History Document history V1.1.1 August 2005 Publication as ES 283 002 V1.1.3 July 2007 Publication as ES 283 002 V2.1.0 March 2008 Publication as ES 283 002 V3.3.1 August 2009 Publication
a816caab598d1d98df7e0e65e42626c6
182 032
1 Scope
The present document specifies the architecture and functions of a CDN Interconnection system, implementing the requirements defined in TS 102 990 [1].
a816caab598d1d98df7e0e65e42626c6
182 032
2 References
References are either specific (identified by date of publication and/or edition number or version number) or non-specific. For specific references, only the cited version applies. For non-specific references, the latest version of the reference document (including any amendments) applies. Referenced documents which are not found to be publicly available in the expected location might be found at http://docbox.etsi.org/Reference. NOTE: While any hyperlinks included in this clause were valid at the time of publication, ETSI cannot guarantee their long term validity.
a816caab598d1d98df7e0e65e42626c6
182 032
2.1 Normative references
The following referenced documents are necessary for the application of the present document. [1] ETSI TS 102 990: "Media Content Distribution (MCD); CDN Interconnection, use cases and requirements".
a816caab598d1d98df7e0e65e42626c6
182 032
2.2 Informative references
The following referenced documents are not necessary for the application of the present document but they assist the user with regard to a particular subject area. [i.1] ETSI TR 102 688-9: "Media Content Distribution (MCD); MCD framework; Part 9: Content Delivery Infrastructures (CDI)". [i.2] IETF RFC 6707: "Content Distribution Network Interconnection (CDNI) Problem Statement", September 2012. NOTE: Available at http://tools.ietf.org/html/rfc6707. [i.3] ATIS-0200003: "CDN Interconnection use case specification and high-level requirements". NOTE: Available at http://webstore.ansi.org/RecordDetail.aspx?sku=ATIS-0200003. [i.4] ISO/IEC 23009-1: "MPEG Dynamic Adaptive Streaming over HTTP (MPEG-DASH)". [i.5] ATIS-0200004: "CDN interconnection use cases and requirements for multicast-based content distribution". NOTE: Available at http://webstore.ansi.org/RecordDetail.aspx?sku=ATIS-0200004. [i.6] ETSI TS 182 019: "Telecommunications and Internet converged Services and Protocols for Advanced Networking (TISPAN); Content Delivery Network (CDN) architecture". ETSI ETSI TS 182 032 V1.1.1 (2013-04) 8
a816caab598d1d98df7e0e65e42626c6
182 032
3 Definitions and abbreviations
a816caab598d1d98df7e0e65e42626c6
182 032
3.1 Definitions
For the purposes of the present document, the following terms and definitions apply: NOTE: Some of the following definitions are from TS 182 019 [i.6] and TR 102 688-9 [i.1]. CDN Interconnection: interconnection between two CDNs, enabling the controlled distribution of content between those CDNs content delivery: act of delivering deployed content to a user Content Delivery Network (CDN): set of functions managing content acquired from content sources, through delivery to the user content acquisition: act of acquiring content from a content source content deployment: act of moving ingested content to one or more network entities, based on content deployment policies content distribution: act of moving content in and between CDNs content ingestion: act of introducing content (and associated data) into the Content Delivery Infrastructure content item: uniquely addressable content element in a CDN. NOTE: A content item is defined by the fact that it has its own Content Metadata associated with it. It is the object of content distribution and request routing operations in a CDN. Example of Content Items are a video file/stream, an audio file/stream, an image file or segmented content together with an associated manifest file. downstream: side of the CDN interconnection that is closest the Consumer logging: recording events related to content items, request routing and content distribution manifest file: file that describes the composition of segmented content metadata: data about content items and CDN network specifics reporting: providing access to recorded events related to content items, request routing and content distribution request routing: exchange of information be two CDNs to aid routing of requests of users for content segmented content: content composed of multiple files, or content composed of multiple streams, or content composed of one or more files and one or more streams upstream: side of the CDN interconnection that is closest to the Content Provider
a816caab598d1d98df7e0e65e42626c6
182 032
3.2 Abbreviations
For the purposes of the present document, the following abbreviations apply: AAC Advanced Audio Coding ALF Asset Location Function AMT Automatic Multicast Tunnelling AS Autonomous System ATIS Alliance for Telecommunications Industry Solutions BGP Border Gateway Protocol CCF Cluster Controller Function CDF Content Delivery Function ETSI ETSI TS 182 032 V1.1.1 (2013-04) 9 CDN Content Delivery Network NOTE: Industry sometimes uses "Content Distribution Network". CDNCF Content Delivery Network Control Function CDN-I Content Delivery Network Interconnection CSF ATIS Cloud Services Forum dCDN Downstream Content Delivery Network DCF Distribution-of-Content Function dDCF Downstream Distribution-of-Content Function dICF Downstream Interconnection Control Function DNS Domain Name System DNSSEC Domain Name System Security Extensions dRCF Downstream Request-routing and Content-control Function DRM Digital Rights Management FLV Flash video GPS Global positioning system HD High Definition HDS HTTP Dynamic Streaming HLS HTTP Live Streaming HTML HyperText Markup Language HTTP Hypertext Transfer Protocol HTTPS HyperText Transfer Protocol Secure ICF Interconnection Control Function IP Internet Protocol MF Manifest File MPD Media Presentation Description MPEG Moving Picture Experts Group MSS Microsoft Media Server OCEAN Open ContEnt Aware Networks RCF Request-routing and Content-control Function REST Representational state transfer RTMP Real Time Messaging Protocol RTSP Real Time Streaming Protocol SDO Standards Developing Organization SLA Service Level Agreement SOAP Simple Object Access Protocol TISPAN Telecommunications and Internet converged Services and Protocols for Advanced Networking uCDN Upstream Content Delivery Network uDCF Upstream Distribution-of-Content Function UE User Equipment uICF Upstream Interconnection Control Function uRCF Upstream Request-routing and Content-control Function URI Uniform Resource Identifier URL Uniform Resource Locator WMV Windows Media Video
a816caab598d1d98df7e0e65e42626c6
182 032
4 High-level overview
a816caab598d1d98df7e0e65e42626c6
182 032
4.1 CDN interconnection services and capabilities
CDNs are in general autonomous networks offering different services to their users. A CDN's primary function is to optimize content distribution and delivery. In addition to this primary function many CDNs chose to implement various other capabilities like content manipulation, digital rights management (DRM), intelligent handling of segmented content and others. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 10 Because of this variation, the present document specifies a basic set of capabilities that every interconnected CDN shall support. In addition to the basic capability set, the present document specifies an extended capability set, describing non-mandatory capabilities available for the CDN interconnection environment. The CDN interconnection shall support a capability exchange mechanism described later in the present document. Requirements defined in TS 102 990 [1] shall apply.
a816caab598d1d98df7e0e65e42626c6
182 032
4.2 CDN interconnection capabilities
Table 4.2.1 provides a table of all the capabilities available for CDN interconnection, grouped into the basic and extended capability set. It also mentions whether those capabilities are mandatory or optional. Table 4.2.1: Capability Overview Table Basic capability set Extended capability set Description Interconnection Control Mandatory Responsible for management of the peering relationship between two CDNs Request Routing Mandatory Capabilities responsible for making it possible for CDNs to direct client requests Content Distribution Mandatory Distribute content to other interconnected CDN Footprint Exchange Mandatory Responsible for announcing the network footprints the CDNs are offering to serve Metadata Exchange Mandatory Used to exchange content related information Content Status Exchange Mandatory Used for distribution of real-time status related to the content Report Exchange Mandatory Used for possibly delayed distribution of comprehensive information gathered during the content delivery process Capability Exchange Mandatory Capability to exchange information about the availability of extended capabilities in CDNs Metadata Defined Reporting Optional Capability of creating reports according to content provider parameters defined in metadata Content Integrity Control Optional Capability to maintain content integrity Content Adaptation Optional Capability to make content processing and adaptation within the CDN Multi-Segment Content Support Optional Capability to handle multi-segment content efficiently (for instance logging segment download sessions per session not just per segment) Content Access Control Optional Capability to define advanced content access control rules Content Security and DRM Optional Capability to use cryptographic technologies to maintain content security and DRM Custom Capability Support Optional Capability to support other (not standardized) capabilities
a816caab598d1d98df7e0e65e42626c6
182 032
4.3 CDN-I basic capability set
Every CDN that wants to take part in CDN interconnection shall have a basic set of interconnection related capabilities. These capabilities give the CDN the ability to properly respond to requests coming from other CDNs. The capabilities included in the basic service set are listed in figure 4.3.1. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 11 Figure 4.3.1: Basic CDN-I capability set
a816caab598d1d98df7e0e65e42626c6
182 032
4.4 CDN-I extended capability set
Whereas the basic capability set is sufficient for basic CDN interconnection many CDNs have additional capabilities that may not be available in all CDNs participating in the CDN federation. The presence of optional capabilities in the CDN federation requires those CDNs that wish to make use of those capabilities to have the capability to exchange information about the availability of extended capabilities in CDNs. This capability is called Capability Exchange and it is mandatory. Some of the capabilities included in the extended capability set are listed in figure 4.4.1. Figure 4.4.1: Extended CDN-I capability set ETSI ETSI TS 182 032 V1.1.1 (2013-04) 12
a816caab598d1d98df7e0e65e42626c6
182 032
4.5 Compliance
A CDN interconnection solution is compliant to the present document if the following points are fulfilled: • All mandatory services and features are implemented as specified in the present document. • If an optional service or feature is implemented, then it is implemented as specified in the present document. • If an optional service or feature is implemented, and the present document specifies multiple options, then it is implemented according to at least one or those options.
a816caab598d1d98df7e0e65e42626c6
182 032
5 Overview of functional entities
a816caab598d1d98df7e0e65e42626c6
182 032
5.1 Functional architecture for CDN interconnection services
The overall functional architecture for CDN interconnection service is shown in figure 5.1.1. The functional architecture is based on a multi-layer architecture that enable separate functionalities used for interconnection of CDNs that may involve up to 3 different level of functionalities and related reference points required. The first layer is responsible for content distribution, the second for controlling of content and request, and the third for the controlling of the interconnection itself. The CDN-I functional architecture should enable CDNs to agreed on a minimal level of interconnection, depending on available capabilities and their needs. In case of CDN peering both interconnected CDNs may play both the role of upstream and/or downstream CDN, depending of direction of content request. Figure 5.1.1: Functional architecture for CDN interconnection services
a816caab598d1d98df7e0e65e42626c6
182 032
5.2 Functional entities
The Content Delivery Network contains one or more Distribution of Content Function (DCF) that can be grouped geographically or administratively in clusters and contains several delivery nodes (that distribute content to other CDN or to end user) hidden to other CDN. CDN shall contain one or more Request and Content Function (RCF) that process requests related to content distribution control and routing request. A CDN Interconnection Control Function (ICF) is responsible for the management of the interconnection. All CDN interconnection entities may support topology hiding and provide abstraction layer from internal CDN architecture. NOTE: Co-location of interconnection entities with existing CDN entities is possible. For example in case that one of the interconnected CDNs is based on ETSI CDN specification [i.6] there is possibility to collocate ICF with CDNCF, RCF with CCF and DCF with CDF. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 13
a816caab598d1d98df7e0e65e42626c6
182 032
5.2.1 CDN Interconnection Control Function (ICF)
A CDN Interconnection Control Function (ICF) shall manage, create, terminate and exchange CDN networks properties, status report required for CDN interconnection between two or more CDNs (CDN peers). An ICF contains following functionalities: • Footprint Exchange - enable CDNs exchange footprint information. • Capability Exchange - the upstream CDN gets information about capabilities of downstream CDN. This information can be used by Request Control Function. • Network Status Reporting - the upstream CDN gets status of downstream CDN network. This status is used by Request Control Function. If for example downstream network reports problems, upstream CDN will serve request by itself, or via other functional downstream CDN. • Network Logging - the downstream CDN sends logs to upstream CDN network. These logs are important for content provider and also for CDN administration. IP interconnection setup and service agreement between CDN providers shall be realized prior to logical CDN interconnection. Details of these activities are out of scope of the present document.
a816caab598d1d98df7e0e65e42626c6
182 032
5.2.2 Request and Content Control Function (RCF)
The Request and Content Control Function (RCF) is responsible for content control and request routing as well as exchanging metadata related to content control. The RCF contains following functionalities: • Metadata exchange function - Metadata are sent from upstream CDN to downstream CDN. Downstream CDN can then make room for content and can inform upstream CDN about content handling possibility. • Content request function - one CDN can request content from other CDN. Different request routing models can be used (e.g. push/pull model, chaining or redirecting requests). • Content status reporting - The Upstream CDN gets status of content from downstream CDN. This status can be than stored local database. Also events can be invoked on status change. Also Downstream CDN can inform upstream CDN when content status has been changed.
a816caab598d1d98df7e0e65e42626c6
182 032
5.2.3 Distribution of Content Function (DCF)
Distribution of Content Function (DCF) is responsible for distribution of content between CDNs in form of files, streams, metadata. The DCF contains following functionalities: • Transfer of file-based content. • Publication and streaming of stream-based content. • Content metadata distribution (if metadata distributed with content).
a816caab598d1d98df7e0e65e42626c6
182 032
5.3 Reference points
a816caab598d1d98df7e0e65e42626c6
182 032
5.3.1 ICF - ICF (CDN-Ic)
This reference point is between two ICFs and it is used for controlling interconnection peer and transferred over this point information related to CDN capabilities and status, including footprint exchange, capability exchange, interconnection status reporting and network usage/performance logging. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 14
a816caab598d1d98df7e0e65e42626c6
182 032
5.3.2 RCF - RCF (CDN-Ir)
This reference point is between two CCFs and it is used for requesting content and to transfer content related information, including content metadata exchange, content requests and content status reporting.
a816caab598d1d98df7e0e65e42626c6
182 032
5.3.3 DCF - DCF (CDN-Id)
This reference point is between two DCFs. Content files, content streams and content related data (if distributed as part of content) are transferred over this point.
a816caab598d1d98df7e0e65e42626c6
182 032
6 Procedures
This clause specifies CDN-I procedures. Clause 6.1 specifies the 3 main phases of interconnection. The following clauses describe specific procedures related to different capabilities that were agreed during the interconnection phase.
a816caab598d1d98df7e0e65e42626c6
182 032
6.1 CDN interconnection phases
The basic process of CDN interconnection may consist of three basic phases: • Interconnection establishment, during which the CDNs negotiate the interconnection. • Interconnection phase, during which the CDNs are fully interconnected and able to share their resources. • Interconnection release, during which the interconnection between the CDNs is released. The interconnection establishment and release phases may be omitted in systems that prefer manual provisioning and in that case interconnection establishment and configuration is performed statically by both interconnected CDN providers before interconnection phase itself. The procedure related to the interconnection establishment is described in clause 6.1.1, the procedure related to the interconnection phase is described in clause 6.1.2, and the procedure related to the interconnection release is described in clause 6.1.1.
a816caab598d1d98df7e0e65e42626c6
182 032
6.1.1 Interconnection establishment
The interconnection establishment is a procedure in which an uCDN and a dCDN begin with no relationship between each other and proceed by exchanging all the information needed to verify each other's identity and thus establish a secure communication channel between each other. This communication is between the ICFs of the CDNs and spans the first three steps of the procedure. The rest of the phase consists of three sub procedures in which the dCDN informs the uCDN about its capabilities, footprint and starts notifying it about its status. After the uCDN receives the first positive status exchange from the dCDN, it can safely assume that the dCDN is ready to process its requests. The conclusion of this phase is the establishment of communication between dRCF and uRCF and the whole CDN-I relationship moves to its interconnection phase. The interconnection establishment is an optional procedure, which is not used in case both CDN providers agree on static manual pre-provisioning of CDN interconnection, but if dynamic interconnection establishment is supported by both CDNs it should follow one of these procedures. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 15 Figure 6.1.1.1: Interconnection establishment NOTE 1: Prior to interconnection establishment procedure, IP interconnection and service level agreement between interconnected CDN providers and/or content provider are needed. The interconnection process should begin with the interconnection establishment. This interconnection establishment is defined by a procedure consisting of following steps: 1) The uCDN sends an interconnection request to the dCDN. This message is sent between the ICFs of the CDNs and consists of all the information the dCDN needs to decide whether to accept or deny the establishment of peering relationship with the uCDN. This information may include uCDN's CDN identifier, authentication data, required peering parameters and others. 2) After the dCDN receives an interconnection offer from an uCDN it analyses its contents and decides whether to deny it (sending an interconnection deny message and terminating the procedure) or accept it. If it decides to accept the offer then it sends an interconnection accept message that contains information that the uCDN can use to make a final decision about establishing the peering relationship with the dCDN. This information may include dCDN's CDN identifier, authentication data, required peering parameters and others. 3) After the uCDN receives an interconnection accept from a dCDN it analyses its contents and decides whether to deny it (sending an interconnection deny message and terminating the procedure) or confirm it. If it decides to confirm it then it sends an interconnection accept message indicating that the peering can begin. 4) After the interconnection is confirmed by the uCDN then the dCDN shall begin its first capability exchange procedure. 5) After the first capability exchange procedure is finished the dCDN shall begin its first footprint exchange procedure. 6) After the first footprint exchange procedure is finished the dCDN shall begin its first status exchange procedure. NOTE 2: This procedure may be repeated if there are multiple ICFs to be interconnected between the two CDNs. Both the CDNs should then proceed to the interconnection phase.
a816caab598d1d98df7e0e65e42626c6
182 032
6.1.2 Interconnection phase
The second phase is simply called interconnection phase. In interconnection phase shall be both CDNs already fully interconnected and through their interconnection interfaces ICFs and RCFs. Interconnection interfaces are fully useable by both CDNs and agreed capabilities accessible over CDN interconnection. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 16 concurrent subprocedures Figure 6.1.2.1: Interconnection phase The general procedure of the interconnection phase consists of concurrent executions of multiple separate procedures. These procedures are described in other sub-clauses of this clause, beginning with clause 6.2.
a816caab598d1d98df7e0e65e42626c6
182 032
6.1.3 Interconnection release
The interconnection release is a phase that begins when an uCDN or a dCDN decides to release a CDN peering relationship with each other (for a specific ICF-ICF relation, and associated RCF-RCF relations). This means that the procedure has two variants, depending on whether it was initiated by the uCDN or the dCDN. In both cases it concludes the interconnection relationship by cleanly finishing all outstanding actions between the two CDNs. The interconnection release is an optional procedure, which is not used in case CDN providers agree on static manual pre-provisioning of CDN interconnection, but if dynamic interconnection establishment is supported by both CDNs it should follow one of these procedures. The uCDN initiated procedure for this consists of following steps, shown in figure 6.1.3.1. Figure 6.1.3.1: uCDN-initiated interconnection release ETSI ETSI TS 182 032 V1.1.1 (2013-04) 17 1) The uCDN stops redirecting clients to the dCDN. 2) After the last client redirect is finished it sends the release request message to the dCDN. 3) The dICF immediately sends an acknowledgement. 4) After receiving the release request, the dCDN also stops distributing content from the uCDN. 5) The dCDN then waits until all outstanding reports are delivered. 6) The dCDN informs the uCDN about this event using the release confirm message. 7) The uCDN sends back acknowledgment message. After this both CDNs can fully disband the interconnection relationship and release all related resources. The dCDN initiated procedure for this consists of following steps, shown in figure 6.1.3.2. Figure 6.1.3.2: dCDN-initiated interconnection release The dCDN initiated procedure for this consists of following steps: 1) The dICF sends a release request message to the uICF. 2) The uCDN stops redirecting clients to the dCDN. 3) After the last client redirect is finished uICF sends the release accept message to the dICF. 4) After receiving the release request, the dCDN stops distributing content from the uCDN. 5) The dICF then waits until all outstanding reports are delivered. 6) The dICF informs the uICF about this event using the termination confirm message. 7) The uCDN sends an acknowledgement. After this both CDNs can fully disbanded the interconnection relationship and release all related resources. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 18
a816caab598d1d98df7e0e65e42626c6
182 032
6.2 Content distribution
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.1 General
This clause provides procedures for content distribution: • Content distribution control, the initiation of content exchange. • Content exchange, the actually transfer or streaming of content between the interconnecting CDNs.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.2 Content distribution control
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.2.1 General
This clause provides procedures for content distribution control: • Content distribution. • Content deletion.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.2.2 Upstream-initiated content distribution
Figure 6.2.2.2.1 shows the procedure for upstream-initiated content distribution from the uCDN to the dCDN. The two CDNs should be interconnected. NOTE 1: Upstream-initiated content distribution is typically used for the prepositioning of files or streams. Figure 6.2.2.2.1: Upstream-initiated content distribution The procedure has the following steps: 1) The uRCF is triggered to initiate content distribution. It selects a dCDN and makes sure that the capabilities of the dCDN match with the capability requirements of the content that the uCDN wants to distribute. 2) The uRCF selects a uDCF. It may communicate about this with the selected uDCF. 2. 3. 4. 5. 7. 8. 9. dDCF uRCF dRCF uDCF 1. Trigger 6. Content Exchange, see clause 6.1.3 10. Content Exchange, see clause 6.1.3 ETSI ETSI TS 182 032 V1.1.1 (2013-04) 19 NOTE 2: The communication between uRCF and uDCF is CDN internal. It is not specified in the present document. NOTE 3: The uDCF may be located in the Upstream CDN domain or in the Content Provider domain. The latter option saves resources in the Upstream CDN domain. 3) The uRCF sends a request to the dRCF for content distribution. The request contains the contentID of the content item that is to be distributed, the address of the selected uDCF and optionally other information, like a token for authentication purposes, or a DRM flag to indicate that DRM procedures apply (see also clause 6.6.2). 4) The dRCF returns a response, acknowledging the request. The response should contain an address or URI to which future UE request for the content item should be redirected. The dRCF could also reject the request. 5) The dRCF selects one or more dDCFs. It instructs the selected dDCF(s) to perform a content exchange. NOTE 4: The communication between dRCF and dDCF is CDN internal. It is not specified in the present document. 6) The selected dDCF performs a content exchange with the selected uDCF. The content exchange is specified in clause 6.2.3. 7) The dDCF informs the dRCF when the content exchange has completed. NOTE 5: The communication between dDCF and dRCF is CDN internal. It is not specified in the present document. 8) The dRCF informs the uRCF that the content is available for delivery by the Downstream CDN. 9) The uRCF returns an acknowledgement. 10) There may be subsequent content exchanges for the identified content item, where other DCF(s) retrieve the same content item. After step 9, the uCDN can start directing UE requests to the address or URI provided in step 4. NOTE 6: Communication with the UE is not described in the present document. NOTE 7: This procedure can be used in a way that the dCDN retrieves a content item only once, and it takes care of its internal distribution. This procedure can also be used in a way that the same or other dDCFs retrieve a content item multiple times, e.g. because of implemented caching strategies (least recently used, least frequently used, etc). NOTE 8: The dCDN can hide its internal topology towards the uCDN by always using the same limited group of dDCFs to retrieve content items from the uCDN, and performing further content distribution within the dCDN internally. Similarly, the uCDN can hide its internal topology by always using the same limited group of uDCFs to distribute content items to the dCDN.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.2.3 Downstream-initiated content distribution
Figure 6.2.2.3.1 shows the procedure for downstream-initiated content distribution from the uCDN to the dCDN. The procedure is used when the dCDN needs to know from where to retrieve an identified content item from the uCDN. NOTE 1: Downstream-initiated content distribution is typically used for cases where the distribution of content files or streams between the two CDNs is postponed until there is an actual user request. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 20 Figure 6.2.2.3.1: Downstream-initiated content distribution The procedure has the following steps: 1) The dRCF is triggered to initiate content distribution for an identified content item. 2) The dRCF sends a request to the uRCF. There includes the contentID of the content item. 3) The uRCF selects a uDCF. It may communicate about this with the selected uDCF. NOTE 2: The communication between uRCF and uDCF is CDN internal. It is not specified in the present document. 4) The selected uRCF responds to the dRCF. The response includes the location of the selected uDCF and optionally other information, like a token for authentication purposes, or a DRM flag to indicate that DRM procedures apply (see also clause 6.6.2). The uRCF could also reject the request. 5) The dRCF selects a dDCF and triggers it for a content exchange. NOTE 3: The communication between uRCF and uDCF is CDN internal. It is not specified in the present document. 6) The content exchange is specified in clause 6.2.3. 7) There may be subsequent content exchanges for the identified content item, where other DCF(s) retrieve the same content item. After step 6, the dCDN can start delivering the content item to Ues that have been redirected to the dCDN. NOTE 4: Communication with the UE is not described in the present document.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.2.4 Upstream-initiated content deletion
Figure 6.2.2.4.1 shows the procedure for upstream-initiated content deletion in the dCDN. NOTE 1: Upstream-initiated content deletion is typically performed at the request of a Content Provider to make sure that content is no longer available after a point in time. 2. dDCF uRCF dRCF uDCF 1. Trigger 6. Content Exchange, see clause 6.1.3 3. 4. 5. 7. Content Exchange, see clause 6.1.3 ETSI ETSI TS 182 032 V1.1.1 (2013-04) 21 Figure 6.2.2.4.1: Upstream-initiated content deletion The procedure has the following steps: 1) The uRCF is triggered to initiate content deletion. It should stop the request routing for the content item to the dCDN. 2) The uRCF sends a request to the dRCF to delete content. The request includes the contentID of the to-be- deleted content and a time-out parameter. 3) The dRCF returns an acknowledgement. 4) There are dCDN-internal communications and action to delete the identified content item. The dCDN should stop accepting new delivery requests for the content item. After the time-out, specified by the time-out parameter, it should actively stop/release all still on-going deliveries of the content item. If the content item is a stream, then clause 6.2.3.4 applies. NOTE 2: The dCDN-internal communications and actions are not specified in the present document. 5) The dRCF conforms the successful deletion of the content to the uRCF, see clause 6.2.2.5. After completion of this procedure, there shall be no further content exchanges for the deleted content.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.2.5 Downstream notification of content deletion
Figure 6.2.2.5.1 shows the procedure for downstream notification of content deletion. Figure 6.2.2.5.1: Downstream notification of content deletion The procedure has the following steps: 1) The dRCF is triggered to send a notification of content deletion. The dRCF may be triggered when all instances of a particular content file or stream are deleted in the dCDN. The dRCF shall be triggered if the content deletion is the result of an upstream-initiated content deletion as specified in clause 6.2.2.4. 3. 2. dDCF uRCF dRCF uDCF 1. Trigger 2. 3. dDCF uRCF dRCF uDCF 5. Notification of content deletion, see clause 6.1.2.5 1. Trigger 4. Content deletion, see also clause 6.1.3.4 ETSI ETSI TS 182 032 V1.1.1 (2013-04) 22 2) The dRCF sends a notification of content deletion to the uRCF. The notification includes the contentID of the deleted content item. 3) The uRCF returns an acknowledgement.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.3 Content exchange
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.3.1 General
This clause provides procedures for content exchange. The following types of content may be exchanged: • Files. • Streams. • Segmented content.
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.3.2 File transfer
Figure 6.2.3.2.1 shows the content-exchange procedure for file transfer. Figure 6.2.3.2.1: File transfer The procedure has the following steps: 1) The dDCF is triggered retrieve a file. The trigger includes information that enables the dDCF to identify and locate a selected uDCF. 2) The dDCF sends a retrieval request to the uDCF. 3) The uRCF authenticates the request and returns the requested file. NOTE: There may be more messages exchanged between the dDCF and uDCF, depending on the used protocol(s).
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.3.3 Stream set-up
Figure 6.2.3.3.1 shows the content-exchange procedure for stream set-up. 3. 2. dDCF uRCF dRCF uDCF 1. Trigger ETSI ETSI TS 182 032 V1.1.1 (2013-04) 23 Figure 6.2.3.3.1: Stream set-up The procedure has the following steps: 1) The dDCF is triggered to set up a stream. The trigger includes information that enables the dDCF to identify and locate a selected uDCF. 2) The dDCF sends a request to the uDCF to set up the stream. The request includes the contentID of the requested stream. 3) The uDCF authenticates the request and starts sending the requested stream to the dDCF. 4) The uDCF confirms to the dDCF that the stream has been set-up. NOTE: There may be more messages exchanged between the dDCF and uDCF, depending on the used protocol(s).
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.3.4 Stream release
Figure 6.2.3.4.1 shows the content-exchange procedure for stream release. Figure 6.2.3.4.1: Stream release The procedure has the following steps: 1) The dDCF is triggered to release a stream. 2) The dDCF sends a request to the uDCF to release the stream. 3) The uRCF stops sending the stream to the dDCF. 4) The uRCF confirms to the dDCF that the stream has been stopped. NOTE: There may be more messages exchanged between the dDCF and uDCF, depending on the used protocol(s). 3. 2. dDCF uRCF dRCF uDCF 1. Trigger 4. X 3. 2. dDCF uRCF dRCF uDCF 1. Trigger 4. ETSI ETSI TS 182 032 V1.1.1 (2013-04) 24
a816caab598d1d98df7e0e65e42626c6
182 032
6.2.3.5 Segmented content
Segmented content is content composed of multiple files, or content composed of multiple streams, or content composed of one or more files and one or more streams. The composition of segmented content is described in an associated manifest file (MF). NOTE 1: The term "manifest file" has several synonyms, depending on the technology used. MPEG DASH [i.4], for example, uses the term Media Presentation Description (MPD). Figure 6.2.3.5.1 shows the content-exchange procedure for segmented content. Figure 6.2.3.5.1: Segmented content The procedure has the following steps: 1) The dDCF retrieves an MF from the uDCF, using the procedure of clause 6.2.3.2. The uCDN may decide to keep some URLs in the MF empty, in order to prevent the dCDN from retrieving the associated segments. NOTE 2: There may be several reasons why an uCDN would want to do this. If the dCDN is a mobile CDN, then it makes no sense to populate the dCDN with high-bitrate high-definition (HD) segments, that would never be delivered by the dCDN anyway. There may be cost considerations, where the uCDN would outsource the delivery of popular segments to the dCDN, and deliver the less popular segments from the uCDN itself. Especially for video clips, the early parts are much better watched than the later parts. Finally, the uCDN may want to deliver some segments itself for monitoring and logging purposes. NOTE 3: An MF with some empty URLs would typically only be used for content distribution and not be delivered to UEs. 2) The dDCF parses the manifest file and decides to initiate further content exchanges to retrieve the identified segments. 3) Further content exchange, see clause 6.2.3. 4) Optional further content exchange(s), see clause 6.2.3. NOTE 4: Whereas this clause shows how an MF (Manifest File) is distributed as a special type of content item, the MF could also be handled in a different way depending on bilateral agreements between the uCDN and dCDN. For example, depending on the type of MF (containing relative URLs, absolute URLs with redirection, or absolute URL without redirection), the MF could be distributed like any generic content item, the MF could be distributed like a special type of metadata, or the MF is not distributed to the dCDN at all. dDCF uRCF dRCF uDCF 3. Content Exchange, see clause 6.1.3 2. Manifest file 4. Content Exchange, see clause 6.1.3 1. Content exchange for manifest file, see clause 6.1.3.2 ETSI ETSI TS 182 032 V1.1.1 (2013-04) 25
a816caab598d1d98df7e0e65e42626c6
182 032
6.3 Request routing
a816caab598d1d98df7e0e65e42626c6
182 032
6.3.1 General
This clause provides procedures for request routing between two interconnected CDNs.