This "Idle-Mode Redial" mechanism is a combination of existing Voice call and Video call standards and services. It is a "terminal-centric" solution that requires only minimal additional support from the UTRAN, GERAN and Core Network. There are 3 main components to the solution:
Switching between voice and video during an established call is achieved by one UE releasing the voice (or video) call and then that UE establishing a video (or voice) call with the same destination UE.
During a video call, the radio coverage at one end may degrade such that the video call cannot be maintained. In this case the video call will be released, and, the UE that initiated the video call can offer to its user an attempt to establish a voice call with the other party.
The initial establishment of a video call may be unsuccessful, in which case the A party can automatically establish a voice call.
For all these 3 cases, optimisation of the Man Machine Interface in the UE is possible and desirable.
It is not the intention of this TR to specify the MMI of UEs, however, this TR does describe some MMI actions solely for the purpose of easing the description of the procedures.
The successful case of the "B party initiated switch to video" for UMTS is illustrated in Figure 1. The notation "UE 1" and "UE 2" is used (rather than A and B party) because, if the B party (of the voice call) presses the "switch to video" button, it then becomes the A party of the subsequent video call.
The UEs are in a voice call initiated by UE 1. Following verbal dialogue between User 1 and User 2, they agree that a switch to video is likely to be successful (e.g. because both User 1 and User 2 have video capable handsets, and, they both have "using 3G" indications visible on their screens). The User of UE 2 pushes the "switch to video" button (or uses other MMI that provides the same functionality) and UE 2 stores the CLI of UE 1 for use in step 7.
MSC 2 sends a RELEASE message with the release cause. The cause value in the RELEASE message might be changed by transit networks prior to arrival at MSC 1.
MSC 2 and MSC 1 locally release the Iu connections to RNC 2 and RNC 1 (assuming that there are no other CM connections active). RNC 2 and RNC 1 then release the RRC connections (assuming that their mobiles are not in PMM connected state).
In order to minimise idle tasks performed by the UE, the following should be performed by Operators which preferentially camp UEs on UMTS:
the BSC should include the "Cell selection indicator after release of all TCH and SDCCH" IE in the Channel Release message (see TS 44.018) if the release is done on a GERAN cell and the call was not started on this BSC and the local 2G and 3G cells are in different RAs.
UE 2 initiates RRC connection setup and sends the Setup message for the video call to MSC 2. The Called Party number is the one stored in step 0. The Setup message should carry extra information to indicate that it is the result of a redial attempt.
MSC 2 requests the allocation of radio resources.
If the GERAN does not support CS video, then the MSC 2 shall send an A interface Assignment Request message to the BSC including the Service Handover IE set to "Handover to UTRAN or cdma2000 should be performed" and a Channel Type indicating any value acceptable to the BSC. The BSS handles the Assignment Request message as defined in TS 48.008. The interactions between MSC server and MGW can be minimised if the A interface circuit is not allocated. This can be achieved by setting the Channel Type = signalling. The interactions between the MSC server and the MGW are defined in TS 23.205.
If the GERAN does support CS video then the MSC 2 should send an A interface Assignment Request message to the BSC including the relevant channel type. The BSS proceeds as defined in TS 48.008.
MSC 1 pages UE 1. UE 1 responds by establishing an RRC connection and sending the Paging Response message to MSC 1 within the RRC-Initial Direct Transfer and RANAP-Initial UE messages. MSC 1 then sends the Setup message for the video call to UE 1. The Setup message carries the CLI of UE 2.
UE 1 confirms its capability to handle the video call in the Call Confirmed message and MSC 1 requests the allocation of radio resources.
If UE 1 responded to paging by establishing an RR connection on a GERAN cell, then MSC 1 initiates handover to UTRAN as described in step 8.
If the radio resource allocation was successful, UE 1 sends an Alerting message to MSC 1. MSC 1 sends the ACM message to MSC 2 and MSC 2 sends the Alerting message to UE 2.
As a minimum, it will be useful if mobiles can indicate whether they are using UMTS for their ongoing voice call.
However additional functionality would be useful, specifically:
when using UTRAN, an indication of whether 64 kbit/s video can be supported in the mobile's current location;
when using GERAN, an indication of whether UMTS coverage is available, and, ideally whether the UMTS coverage can support 64 kbit/s video.
RNC 2 is configured such that it knows that the local GERAN does not (or does) support video calls (i.e. does not (or does) support 64 kbit/s conversational QoS on the CS domain).
When RNC 2 detects that the 64kbit/s bearer cannot be maintained any longer (e.g. radio link failure, TS 25.331), RNC 2 sends either an Iu RELEASE REQUEST message or a RAB Release Request message to MSC 2, indicating that the Iu connection should be released
The MSCs send Disconnect messages to the UEs.
In case RNC 2 sent Iu Release Request in step 1 but not a RAB Release Request, MSC 2 will not perform step 3. (i.e. not send Disconnect, which causes that also steps 5 and 6 are not performed), but MSC 2 will immediately proceed with step 7 (i.e. send Iu Release Command).
UE 1 prompts its user to attempt a Redial. When UE 1 gets the Redial confirmation from its user it sends a SETUP message (containing the Redial indication) to MSC 1 to setup a Voice call.
The procedure in 4.3.1.1 can also be used if UE 2 drops out of UTRAN coverage into GERAN coverage. However there are additional complications:
If the UTRAN and GERAN cells are attached to different MSCs then problems are likely to occur due to the time required for the UE 2 to perform Location Area and Routeing Area updating.
In particular, if UE 2 is updating with a new MSC and/or new SGSN, UE 2 will not respond to MSC 2's paging. This will only become known to UE 1 when MSC 2's "no response to paging" timer expires. Typically, an operator configures this timer with values in the range of 8 to 25 seconds. When this timer expires, it is quite likely that UE 2 will have "call forward on not reachable" set and so UE 1's call will be diverted to a voice mail server. To avoid these problems, it seems important that the operator configures the UTRAN to minimise the number of video calls that lead to fallback to GERAN coverage and/or to configure the network such that geographically similar UTRAN and GERAN cells are in the same Routeing Area.
In order to avoid useless handover attempts, the RNCs are frequently configured so that no GERAN neighbour cells are given to the mobile during a video call. This will slow down re-selection from 3G to 2G in the case of total loss of 3G coverage. Conversely, in many other scenarios the lack of 2G neighbour cells might force UE 2 to camp on the 3G cell for long enough for it to receive the paging for the voice call.
Overall, the operator may need to adapt the neighbour cell lists provided in video calls on cell by cell basis and/or dependent upon whether or not combined 2G/3G MSCs AND combined SGSNs are in use.
Additional study may be beneficial on how to 'synchronise' the release of the RRC connection in the RNC and in the mobile when the radio connection is degraded or lost. If synchronisation cannot be guaranteed then timers may be needed to delay the redial attempt by the A party.
If a redial attempt is unsuccessful, it may (or may not) be appropriate to avoid being forwarded to a voice mail server.
One potential solution to this is that the A party signals (e.g. by appending a suffix to the dialled digits) to MSC-A and MSC-A then uses the Call Diversion Not Allowed setting in the Call Diversion Treatment Indicators to inhibit call forwarding. One problem with the use of a suffix in the dialled digits is that the digit analysis in MSC-A usually only examines the leading digits. Conversely, if the UE adds a prefix to the dialled digits, then the call will not be handled properly if MSC-A does not support this functionality.
The following diagram shows the various cases where video call establishment can fail or be unsuccessful. At many points, alternative signalling message flows are shown by the "a" and "b" arrows.
A RAB Assignment Request message is sent from the MSC A to the RNC A, requesting the setup of a RAB for a Video call. The radio bearer is established between RNC A and UE A.
If the UE is using a GERAN that does not support CS video calls, the MSC sends an Assignment Request with an encoding that prompts the BSC to initiate a handover to UTRAN.
(Failure case) The video call fails because of lack of radio resources on the A side. MSC A sends a Release Complete message (e.g. with cause #47, Resources unavailable, unspecified) to UE A. The RRC connection is released. UE A can then prompt the user to initiate a voice call to UE B.
(Failure case) The video call fails because MSC B or the transit network do not support video calls. MSC B sends a REL message to MSC A. MSC A sends a Disconnect message (e.g. with cause #79, Service or option not implemented, unspecified) to UE A. The RRC connection is released. UE A can then prompt the user to initiate a voice call to UE B.
(Failure case) The video call fails because UE B does not support video calls. UE B sends a Release Complete message with cause #88 (Incompatible destination) to MSC B. MSC B sends a REL message to MSC A. MSC A sends a Disconnect message (e.g. with cause #88, Incompatible destination) to UE A. The RRC connections are released. UE A can then prompt the user to initiate a voice call to UE B.
(Successful case) The RAB Assignment Request message is sent from MSC B to the RNC B, requesting the establishment of a RAB for a Video Call. The radio bearer is established between the RNC B and UE B. RNC B responds to MSC B with a RAB Assignment Response message. Following the allocation of the radio resources, UE B sends an Alerting message to MSC B.
(Failure case) The video call fails because of lack of radio resources on the B side. MSC B sends a Release Complete message to UE B. MSC B sends a REL message to MSC A. MSC A sends a Disconnect message to UE A (e.g. with cause #47, Resources unavailable, unspecified). The RRC connections are released. UE A can then prompt the user to initiate a voice call to UE B.
(Failure case) UE B is alerting the user of UE B. The user of UE B may choose not to accept the video call and either Reject the call or indicate User Determined User Busy. UE B then sends a Disconnect message (with cause #21, Call rejected or cause #17, User busy) to MSC B. MSC B sends a REL message to MSC A. MSC A sends a Disconnect message (with cause #21, Call rejected or cause #17, User busy) to UE A. The RRC connections are released. UE A can then prompt the user to initiate a voice call to UE B.
Many other cause values can be returned to UE A in the Disconnect/Release Complete messages. It is a matter of UE implementation as to whether or not these cause values are used to prompt the user to initiate a re-dial.
If (following a failed video call establishment attempt) the UE prompts the user to initiate a Redial, and, the user requests the Redial attempt, then, the resulting SETUP message should carry extra information to indicate that it is a redial attempt.
There are many reasons (User Determined User Busy, Barring of all incoming calls, an existing active video call) why a video call and a re-dialled voice call might both be unsuccessful. It would be useful if the rejection for the video call carried information that helped UE A to determine whether or not a subsequent voice call might succeed.
Can users of video phones be provided with a full suite of supplementary services, e.g. can automatic forwarding to a video mail box when not in UMTS coverage be performed? Analysis of the current Call Forwarding specifications (TS 22.082, TS 23.082) shows that MSC B should reject the call rather than forward it (because the UE is reachable but is not busy and has not been offered the chance to reply).
Once a video call has been stored by the mailbox, it can only be retrieved from the mailbox when the UE is in coverage that allows for a 64kbit/s bearer. As long as the UE is not in such a coverage area it cannot receive the video call from the mailbox.
One potential solution to this is that the mailbox provides a voice call comprising only the audio part of the video call. An alternative is that the user deactivates all call forwarding for video calls (c.f. CS data calls).
This appears difficult to achieve in an automatic manner. However, given the privacy issues associated with automatic switching from voice to video, this does not appear to be a serious deficiency.
For manual switching, good indications of UTRAN coverage, and, its ability (or not) to handle 64 kbit/s video would be useful. This is the same requirement as described in clause 4.2.2.1.