- sg16-avd - lists.packetizer.com

17 April Mobility Conf call notes
by Guram Paul-LPG019 17 Apr '00

17 Apr '00

Hello mob group Please find below the notes from today's conference call: 1. Representatives from AT&T, Intel, Nokia, Ericsson, Siemens, Motorola and Alcatel discussed MD-017 from Radhika Roy (AT&T) at the conf.call. The use of the word 'network' was clarified in that we shall use it in conjunction with addressing, i.e. 'network point of attachment'. Thus the figure 1 in MD-017 is to be modified to reflect this. The word 'home' in this figure is presently to be kept only for the GK and the Zone even though there was discussion to delete this word completely from the figure. The use of this word 'home' may be omitted from the figure altogether after more discussion later. This figure is to be incorporated in Annex H to show the 'intra-zone' communication scenario. The editor, will also consider to add some text to explain the boxes VLF,HLF and AuF, if these are also included in the new version of Annex H. Also from the email discussion and today's discussion the editor will incorporate text (as he sees fit) in appropriate sections of Annex H. The next version of the Annex will show all the new text added by the editor (with suitable revision marks), and indicate at the beginning of the Annex that the text with no revision marks is the only agreed text so far. This is to ensure that readers of the new version are not misled as to which parts of the annex are the agreed text and which are still for discussion. The editor, Jaakko Sundquist, will endeavour to produce this new version one week before the next conference call which is to be on Tuesday 9 May at the same time. 2. Steve Terrill (Ericsson) will enhance his last contribution MD-016 to explain the reasons behind his proposals. Also he will propose/discuss discovery mechanisms for the VLF. 3. Radhika Roy will produce a contribution to explain why it is necessary to include the word 'home' in the 'network point of attachment', i.e. explain the 'home network point of attachment'. 4. Paul Guram will consult Mr Okubo to see if a separate dedicated reflector can be started for the mobility group's use. 5. A number of contributions have not been discussed yet due to lack of time. From the material in these contributions, the contributors will make proposals to enhance/add/delete text when the editor submits the new version next time. This would be a better way to progress the work and also ensure that no-one's contribution is inadvertently discarded without discussion. Paul

1 0

[Robustness] Conference call for Apr 20
by Terry L Anderson 17 Apr '00

17 Apr '00

The next H.323 Robustness teleconference will be as below. NOTE THAT THIS IS LATER IN THE DAY THAN PREVIOUS CALLS Conference Date: 20-APR-2000 Conference Time: 11:00 AM Eastern US (EDT) 10:00 AM Central 09:00 AM Mountain 08:00 AM Pacific 15:00 UT or GMT Bridge Telephone Number: 630-979-0100 Conference Code: 386137 It would be helpful is those that plan to call in send email to tla(a)lucent.com (NOT TO THE SG16 REFLECTOR) so that we have sufficient ports reserved. -- ------------------------------------------------------------ Terry L Anderson mailto:tla@lucent.com Tel:908.582.7013 Fax:908.582.6729 Pager:800.759.8352 pin 1704572 1704572(a)skytel.com Lucent Technologies/ Voice Over IP Access Networks/ Applications Grp Rm 2B-121, 600 Mountain Av, Murray Hill, NJ 07974 http://its.lucent.com/~tla (Lucent internal) http://www.gti.net/tla

1 0

Re: srcAlternatives and destAlternatives
by Francois Audet 17 Apr '00

17 Apr '00

Cool. You got a good memory... -----Original Message----- From: Paul E. Jones [mailto:paul.jones@ties.itu.ch] Sent: Monday, April 17, 2000 12:06 PM To: ITU-SG16; Audet, Francois [SC9:4K02:EXCH] Subject: Re: srcAlternatives and destAlternatives Francois, This came from a contribution made by VocalTec (AVC-1157) in February 1997. This was introduced as a means of providing some redundancy to the endpoint. Unfortunately, just as with the alternate gatekeeper fields that were added to V2, there was no procedural text to explain the usage of these fields. So for the moment, the usage is undefined. Paul

1 0

srcAlternatives and destAlternatives
by Francois Audet 17 Apr '00

17 Apr '00

Does anybody have a clue what the srcAlternatives and destAlternatives fields in the ARQ are supposed to be? I did a search on H.225.0 and H.323 and the only occurence of it is in H.225.0 where the fields are described to be "a sequence of prioritized source endpoint alternatives for srcInfo, srcCallSignalAddress, or rasAddress" and "a sequence of prioritized destination endpoint alternatives for destinationInfo or destCallSignalAddress". Not very explicit... ----- François Audet Tel:+1 408 565 5675 <mailto:audet@NortelNetworks.com> mailto:audet@NortelNetworks.com Nortel Networks Fax:+1 408 565 2375 <http://www.nortelnetworks.com/> http://www.NortelNetworks.com

2 1

Re: [H.323 Mobility:] Comments to various mails
by Roy, Radhika R, ALARC 17 Apr '00

17 Apr '00

Hi, Jaakko: Pl. my comments below: Thanks, Radhika > -----Original Message----- > From: Jaakko Sundquist [SMTP:jaakko.sundquist@nokia.com] > Sent: Monday, April 17, 2000 10:34 AM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: Re: [H.323 Mobility:] Comments to various mails > > Hi Radhika, > > Comments embedded... > > > > > Hi, Jaakko: > > > > I am providing a very quick response because we will have the > > conf call > > within 45 minutes. > Same here... > > > Second, the H.323 mobility protocol proposed by AT&T, Alcatel > > or others do > > NOT tie to any topology. The topology shown has come into > > play to provide > > descriptions so that people can understand, and we can say > > all requirements > > are met no matter what the topoloy is. For example, your > > contribution and > > references figures - we need to see whether all toplopies can > > be satisfied > > using the SAME protocol. > > I did not claim that the contributions were tied to any topology. I was > simply questioning the need to define or use the terms: home/visited > network/zone. [Roy, Radhika R] There shall be ONLY reference to the network address [home/visiting/visited/target] because "network" itself does NOT mean anything clearly. It is an OPTION, not mandatory. The contributions explain that what the benefits are: 1. To hide network address as the mobile moves from one place to another if needed. For example, a mobile may communicate with the other party without revealing its present network point of attachment, while the other party may be assuming the mobile is still in its home network point of attachment. The mobile can keep its privacy. (Mobile also allows to keep privacy.) 2. In multimedia communications environment, it may allow, if needed, to treat different media differently. For example, audio and/or video may be directed to the mobile's present network point of attachment, while data may be sent to my home network point of attachment (for storage or other purposes). 3. A service provider may be able to DIFFERENTIATE services based on the home network point of attachment (because a service may be subscribed in mobile's home domain/zone with a certain agreement and mobile's home network point of attachment may provide one of the most IMPORTANT identities to accomplish this). 4. To Facilitate for keeping track of the mobile user's profile based on the home network point of attachment in addition to other parameters stored in the HLF/VLF by the service provider. 5. It may so happen that a mobile may reveal the network point of attachment (e.g., audio) while it may not reveal the network point of attachment for data because of security reason. As a result, the communications may happen like this: Real-time audio and video are coming directly between the calling parities while data can shared from mobile's home network point of attachment without revealing its address. 6. There can other benefits in keeping its "home" network point of attachment known as a reference point (similar "benefits" along the line "home administrative domain"). 7. In the same token, visiting/visited/target network point of attachment is useful to keep references in managing the mobility especially by the service providers. For example, it the target network point of attachment can be predicted or supplied by the mobile user at the time of registration, a service provider can make planning of the resources ahead of the time to provide services before the mobile can move in that zone/domain. With respect to home zone/GK, it is another reference from mobility management point of (similar "benefits" along the line "home administrative domain"). > > > > Third, no contribution (AT&T, Alcatel, or Ericsson) has > > suggested that HLF > > or VLF has to belong to a zone or by couple of zones or otherwise. > > This is exactly why I do not see any point in defining home/visited zone. > Or > if we do define them the home zone should be clearly defined in such a way > that it does not imply that the HLF is a part of the home zone, but home > zones are all the zones that belong to the Home Administrative Domain. [Roy, Radhika R] This is helpful from mobility management point of view because association is made via the GKs. Yes, you are right that home zone belongs to the home adm domain, but not ALL zones of a given adm domain are the mobile's home zone. There is only ONE zone is the mobile's home zone. This distinction is very helpful from mobility management point of view in H.323. Yes, I agree the home zone definition should NOT interfere with the definition of HLF. An HLF may contain the information of many one or many home zones of many mobile users of a given Adm domain. Now HLF is a database behind the GK that keeps the permanent information including mobile's home network point of attachment. > > > > I hope that you will NOT mis-interpret the intend of the main > > idea behind > > the contributions. > > I don't think so and I do think that we have some valuable contributions > for > this teleconference and they will certainly help the progress of the work. [Roy, Radhika R] I appreciate. I am sorry to see that I could not choose the right word. I take my comments off. > -Jaakko

1 0

Re: Should we Break H.323 Model for Mobility?
by Roy, Radhika R, ALARC 17 Apr '00

17 Apr '00

Hi, Steve: Excellent! HLF/VLF: Definitely, I agree that we need to clarify the roles of HLF, VLF, and GKs very clearly because this is the heart of the mobility solution. I guess that we have identified the functions of HLF and VLF. In my contributions, I have explained as follows: 1. Intra-domain: Populate the HLF database through registration [not statically, but dynamically] (although the HLF may not be needed to be used for mobility management). Call flows show how HLF is used. I left the VLF function untouched. However, VLF can be used, if needed, for managing the change in network point of attachments. 2. Intra-Domain/Inter-Domain: Both HLF and VLF are needed. Some call flows are there to explain the role of these entities. You are right that we may have to explain more about HLF and VLF. Is it possible to explain more via short contributions or emails so that we all can agree with you arguments? GK Discovery: A note to Laurent Thiebaut/Nicolas Tran: I understand from the conference call, you might be another efficient mechanism to discover the GK using the DNS method (I hope that the process would be dynamic). Can you please send us a contribution related to this so that we can compare between H.323's GRQ/GCF/GRJ, proposed (mine) MGA, and your proposed DNS method? I have provided explanation via contributions, emails, and conference call why MGA is better than GRQ/GCF/GRJ. We like to see your proposed method. Please note that we may make these three as a part of the H.323 standard. Let people implement whatever mechanism best for them. Best regards, Radhika R. Roy AT&T > -----Original Message----- > From: Stephen Terrill [SMTP:stephen.terrill@ericsson.com] > Sent: Monday, April 17, 2000 1:01 PM > To: Roy, Radhika R, ALARC > Cc: ITU-SG16(a)mailbag.cps.intel.com > Subject: Re: Should we Break H.323 Model for Mobility? > > Hi Rardhika, > > I agree that we need to register with the gatekeeper, however I extend > that to say, the gatekeeper could be at home. > > I think that we need to clarify the roles of the HLF, VLF, GKs etc before > we can move further forward. > > Regards, > > ..//steve > > "Roy, Radhika R, ALARC" wrote: > > > Hi, Steve: > > > > Your statement appears that you are agreeing with the registration with > the > > GK that is the basic foundation of H.323. (However, your contribution > does > > NOT clearly indicate this.) It is a very good news. > > > > Contributions are there what roles to be performed by the VLF, HLF, and > > home/visiting/visited/target GK/zone/domain. We like to see where you > differ > > and why, and your suggestions. > > > > Please also see my subsequent emails what we need to be done to extend > our > > model to be consistent with H.323 as well as how we can reconcile all > > contributions: AT&T, Alcatel, and Ericsson. > > > > Best regards, > > Radhika R. Roy > > AT&T > > > > -----Original Message----- > > From: Stephen Terrill [mailto:stephen.terrill@ericsson.com] > > Sent: Sunday, April 16, 2000 5:00 AM > > To: ITU-SG16(a)MAILBAG.INTEL.COM > > Subject: Re: Should we Break H.323 Model for Mobility? > > > > Hi Roy, > > > > I feel that this email places more assumptions into the role of the VLF, > > HLF, visited GK, and home GK than we have agreed upon. Certainly the > model > > that we have proposed hasn´t stopped registration with the gatekeeper. > > > > I think that we need to agree upon the roles of the particular entities > > befor embarking upon concluding that something breaks the H.323 model. > > > > Regards, > > > > ..//steve > > > > "Roy, Radhika R, ALARC" wrote: > > > > > Hi, Everyone: > > > > > > H.323 has some definite architectural notion such as terminal, GK, GW, > MCU > > > and(MC, MP). One of the fundamental corner stones is the GK. The H.323 > > > entities (e.g., terminals, GWs) register with the GK. A GK manages a > zone > > > with the information of all registered users along with other > functions > > > (e.g., bandwidth, resource management, etc). > > > > > > The "registration" with the GK is the fundamental corner stone of > H.323 > > > where an association is made by each user. That is, a GK becomes the > basic > > > source to have the information related to each user. > > > > > > H.323 mobility protocol MUST also follow the same H.323 architectural > > > concept. If this basic notion is not maintained, it breaks the basic > > > foundation of H.323. > > > > > > It appears that H.323 mobility likes to define additional databases > like > > > VLFs and HLFs that will store the information related to the mobile > users. > > > These databases are supposed to help the function of a GK in order to > > manage > > > the mobility related information. > > > > > > If people think that an H.323 will be able to register with a VLF > > directly, > > > does it not mean we are by-passing the basic foundation of H.323 > model? > > Does > > > it not opening the entirely a new set of problems in H.323? It will > create > > a > > > parallel architecture like GK in H.323 what is not needed at all > because > > > solutions have provided that are consistent with H.323 model and there > is > > no > > > need to do this. We probably have to do all over again what we have > done > > for > > > the GKs and BEs. We inviting unnecessary problems that are yet to > imagine > > > where we may end up with. Why do we need to do that if one can do this > > using > > > the consistent H.323 model of GK? > > > > > > In another situation, we have to go step-by-step. First intra-zone, > then > > > inter-zone, and last of all inter-domain. Contributions are there to > > address > > > those problems. > > > > > > The mobile users that move between cells, those cells are expected > > primarily > > > remain the control of a given GK (in some cases it may be two GKs). > These > > > scenarios mostly seem to be intra-zone and inter-zone and may involve > a > > > single service provider. Of course, there will also be the calls > between > > the > > > two adm domains of two service providers. > > > > > > More importantly, solutions are inter-dependent. That is, we cannot > > address > > > the problems of inter-zone without knowing the intra-zone. In the same > > > token, we cannot address the inter-domain problems without knowing the > > > inter-zone problem. Contributions are there. Let us proceed > accordingly. > > > > > > Last of all, I am reminding the mess that we have made in Annex G. Now > > > people are hungry to use the more powerful and efficient Annex G > messages > > > between the GKs. Contributions were brought to do this, but the > proposals > > > had been rejected by a very few. Who is suffering? > > > > > > Best regards, > > > Radhika R. Roy > > > AT&T >

1 0

Re: Should we Break H.323 Model for Mobility?
by Roy, Radhika R, ALARC 17 Apr '00

17 Apr '00

Hi, Steve: Your statement appears that you are agreeing with the registration with the GK that is the basic foundation of H.323. (However, your contribution does NOT clearly indicate this.) It is a very good news. Contributions are there what roles to be performed by the VLF, HLF, and home/visiting/visited/target GK/zone/domain. We like to see where you differ and why, and your suggestions. Please also see my subsequent emails what we need to be done to extend our model to be consistent with H.323 as well as how we can reconcile all contributions: AT&T, Alcatel, and Ericsson. Best regards, Radhika R. Roy AT&T -----Original Message----- From: Stephen Terrill [mailto:stephen.terrill@ericsson.com] Sent: Sunday, April 16, 2000 5:00 AM To: ITU-SG16(a)MAILBAG.INTEL.COM Subject: Re: Should we Break H.323 Model for Mobility? Hi Roy, I feel that this email places more assumptions into the role of the VLF, HLF, visited GK, and home GK than we have agreed upon. Certainly the model that we have proposed hasn´t stopped registration with the gatekeeper. I think that we need to agree upon the roles of the particular entities befor embarking upon concluding that something breaks the H.323 model. Regards, ..//steve "Roy, Radhika R, ALARC" wrote: > Hi, Everyone: > > H.323 has some definite architectural notion such as terminal, GK, GW, MCU > and(MC, MP). One of the fundamental corner stones is the GK. The H.323 > entities (e.g., terminals, GWs) register with the GK. A GK manages a zone > with the information of all registered users along with other functions > (e.g., bandwidth, resource management, etc). > > The "registration" with the GK is the fundamental corner stone of H.323 > where an association is made by each user. That is, a GK becomes the basic > source to have the information related to each user. > > H.323 mobility protocol MUST also follow the same H.323 architectural > concept. If this basic notion is not maintained, it breaks the basic > foundation of H.323. > > It appears that H.323 mobility likes to define additional databases like > VLFs and HLFs that will store the information related to the mobile users. > These databases are supposed to help the function of a GK in order to manage > the mobility related information. > > If people think that an H.323 will be able to register with a VLF directly, > does it not mean we are by-passing the basic foundation of H.323 model? Does > it not opening the entirely a new set of problems in H.323? It will create a > parallel architecture like GK in H.323 what is not needed at all because > solutions have provided that are consistent with H.323 model and there is no > need to do this. We probably have to do all over again what we have done for > the GKs and BEs. We inviting unnecessary problems that are yet to imagine > where we may end up with. Why do we need to do that if one can do this using > the consistent H.323 model of GK? > > In another situation, we have to go step-by-step. First intra-zone, then > inter-zone, and last of all inter-domain. Contributions are there to address > those problems. > > The mobile users that move between cells, those cells are expected primarily > remain the control of a given GK (in some cases it may be two GKs). These > scenarios mostly seem to be intra-zone and inter-zone and may involve a > single service provider. Of course, there will also be the calls between the > two adm domains of two service providers. > > More importantly, solutions are inter-dependent. That is, we cannot address > the problems of inter-zone without knowing the intra-zone. In the same > token, we cannot address the inter-domain problems without knowing the > inter-zone problem. Contributions are there. Let us proceed accordingly. > > Last of all, I am reminding the mess that we have made in Annex G. Now > people are hungry to use the more powerful and efficient Annex G messages > between the GKs. Contributions were brought to do this, but the proposals > had been rejected by a very few. Who is suffering? > > Best regards, > Radhika R. Roy > AT&T

2 1

Re: [H.323 Mobility:] Comments to various mails
by Jaakko Sundquist 17 Apr '00

17 Apr '00

Hi Radhika, Comments embedded... > > Hi, Jaakko: > > I am providing a very quick response because we will have the > conf call > within 45 minutes. Same here... > Second, the H.323 mobility protocol proposed by AT&T, Alcatel > or others do > NOT tie to any topology. The topology shown has come into > play to provide > descriptions so that people can understand, and we can say > all requirements > are met no matter what the topoloy is. For example, your > contribution and > references figures - we need to see whether all toplopies can > be satisfied > using the SAME protocol. I did not claim that the contributions were tied to any topology. I was simply questioning the need to define or use the terms: home/visited network/zone. > > Third, no contribution (AT&T, Alcatel, or Ericsson) has > suggested that HLF > or VLF has to belong to a zone or by couple of zones or otherwise. This is exactly why I do not see any point in defining home/visited zone. Or if we do define them the home zone should be clearly defined in such a way that it does not imply that the HLF is a part of the home zone, but home zones are all the zones that belong to the Home Administrative Domain. > > I hope that you will NOT mis-interpret the intend of the main > idea behind > the contributions. I don't think so and I do think that we have some valuable contributions for this teleconference and they will certainly help the progress of the work. -Jaakko

2 1

Re: [H.323 Mobility:] Comments to various mails
by Roy, Radhika R, ALARC 17 Apr '00

17 Apr '00

Hi, Jaakko: I am providing a very quick response because we will have the conf call within 45 minutes. First, H.323 has the abstraction of network address (or point of attachment). So, all contributions of AT&T and other companies refer to that only. Second, the H.323 mobility protocol proposed by AT&T, Alcatel or others do NOT tie to any topology. The topology shown has come into play to provide descriptions so that people can understand, and we can say all requirements are met no matter what the topoloy is. For example, your contribution and references figures - we need to see whether all toplopies can be satisfied using the SAME protocol. Third, no contribution (AT&T, Alcatel, or Ericsson) has suggested that HLF or VLF has to belong to a zone or by couple of zones or otherwise. All contributions are saying that HLFs and VLFs are behind the GKs. I hope that you will NOT mis-interpret the intend of the main idea behind the contributions. Again, protocols defined is NOT dependent on any topology. The contribution shows how the protocol can satisfy when this is used in the context of any topology that people can think of so that nothing is missed from the protocol. Best regards, Radhika R. Roy AT&T > -----Original Message----- > From: Jaakko Sundquist [SMTP:jaakko.sundquist@nokia.com] > Sent: Monday, April 17, 2000 9:43 AM > To: ITU-SG16(a)MAILBAG.INTEL.COM > Subject: [H.323 Mobility:] Comments to various mails > > Hi all, > > It certainly seems that we have some discussion going on in the reflector. > I > haven't read the mails since thursday and there really has been a flood of > them, so I'll try to gather all my comments in this message. > > Radhika, thanks to our discussions last week I have finally understood > your > intentions on a couple of issues. First the issue of home/visited network. > I > think I now understand what you mean when you say that a zone may include > more than one network (a part of the zone runs on IP and another part on > ATM > for instance), but I still fail to see, why we should define such > concepts, > the H.323 layer should not really be concerned what the underlying network > type is. Furthermore, you made the argument that since H.323 has defined > the > topology of the H.323 systems in terms of network address/zone/domain we > should also define home/visited network address/zone/domain. I do not > agree > on this. So far I have not seen any reason for defining any other of these > concepts than the already defined Home/Visited Administrative Domain. The > only important issue with these home/visited concepts in my mind is the > location of the HLF of the user and I would argue that the entity > administering some part of the network should also own the HLFs in that > part > of the network. Thus defining the Home Administrative Domain of a user as > the Domain that contains the HLF of the user makes sense, but since we are > not binding the HLF to any particilar zone or network we do not need to > define or use the terms home zone/network. > Secondly, I think I understand, albeit somewhat vaguely, now, what exactly > you are meaning with Intra-zone, Inter-zone and Inter-domain > communications. > The difficulty with these terms in the context of Annex H is, however, > that > traditionally these terms have been used in H.323 only when there is a > call > involved. In other words, we haven't talked about e.g. an Inter-zone > registration procedure previously in H.323. The problem when using (at > least) the terms Intra/Inter-zone with the Annex H work is that we have > not > made any assumptions about the relationships of the mobility functional > entities and zones, i.e. a VLF or a HLF is not defined to be a part of any > zone. Thus, when you are talking about Inter-zone or Inter-domain > communications for e.g. the terminal/user registration or location > updating > procedure, what exactly are you meaning? My guess is that the division > that > you are referring to, is based on the previous location of the User, i.e. > in > the same zone or another zone, am I correct? Anyway, I'm quite sure that > we > need to address all the issues that you are thinking of, but dividing the > work in steps defined by the Intra/Inter-zone/domain communications is not > the best approach in my opinion. > > Steve, I think that we need to clarify the issue of home/visited domain > execution of services, so that both of these models are understood by all > the members of the group. Maybe some co-work could be done by us? > > As for the discussion on "not breaking up the current H.323 model", I'll > just say that I share the opinion that all registrations, etc. should be > done via the gatekeeper in order to preserve the old H.323 model. I see no > point in doing the registration straight to the VLF. I'm not quite sure, > what Steve's contribution was all about in this respect, but we'll > probably > be discussing that in the teleconference. > > ------------------------------------------------ > Jaakko Sundquist * > +358 50 3598281 * Audere est Facere! > jaakko.sundquist(a)nokia.com * > ------------------------------------------------

1 0

[Robustness] Minutes and Next telecon is April 20, 11AM EST USA
by Maureen Stillman 17 Apr '00

17 Apr '00

Minutes of H.323 Robustness Teleconference - April 12, 2000 Prepared by Maureen Stillman, Nokia Next Teleconference: Scheduled for April 20, 2000 11AM EST USA hosted by Lucent Action items from 4/6/2000 teleconference: 1. Post meeting minutes to mailing list 2. Jorg Ott will provide failure scenarios for gatekeepers for the Robustness Framework document. 3. Review Terry Anderson's H.225 and H.245 state machine document. We need to begin to determine what state information should be kept and how it should be restored under different failure scenarios. 4. Review IETF's SCTP and DDP protocols. Randy Smith and Qiaobing Xie of Motorola will join us on 4/20 to answer questions about these protocols. Attending the 4/6/2000 teleconference: Maureen Stillman maureen.stillman(a)nokia.com Sasha Ruditsky sasha(a)tlv.radvision.com Paul Jones paul.jones(a)ties.itu.int Terry Anderson tla(a)lucent.com Rex Core, Lucent (please send e-mail address) Mohash and Tim of Trillium (please send e-mail address) Timothy Sipsey Timothy.Sipsey(a)Aspect.com Summary of Teleconference: Discussion of SCTP and DDP SCTP is about to become an RFC and can be referenced in the appropriate time frame for H.323v4, but DDP will not be ready in time to be referenced. An advantage of SCTP versus TCP: Using TCP, a socket on one end or another can fail. SCTP knows about alternate sockets and can perform fail over. Debate concerning Annex E and DDP: During this debate several issues/requirements were highlighted: 1) Endpoints need to incur a minimum amount of overhead in supporting robustness protocols 2) Amount of extra network traffic generated for robustness purposes must be minimized 3) We must support robustness procedures that include conference calls 4) We need to support restoring state information "in intermediate points" that is for gatekeepers, etc. We need to examine the overlap between DDP and Annex E. Should we use both or just one or the other? If we just use Annex E, then there is no concept of a pool of resources that can handle fail over. DDP also has a built in mechanism to store state and gives us a place to store check pointed state information. If we don't use this mechanism, then we have to invent one. We also need to reestablish state information at the intermediate points (i.e. at gatekeepers) as well in the event of a failure. DDP supports a mechanism for this. We need to have minimum complexity at the endpoints. There is a concern about the amount of overhead incurred at the endpoint to implement DDP. Randy Stewart will supply us with some numbers on the size of a representative DDP implementation. If we use Annex E as the transport, then we need to have end-to-end acknowledgements. We could use Annex E to guarantee delivery of the messages (hop by hop) and use DDP for resource pooling and fail over. Robustness protocols need to generate a minimum of network traffic to perform these functions. There is also a concern that robustness for conferencing be supported. For example, if an H.245 message is sent that establishes that the floor was transferred, then if this message gets lost, then the conference call has a problem. We are still debating the issues at this stage and need further discussion to come to some conclusion. -- maureen Maureen Stillman Member of Scientific Staff Voice: (607)273-0724 x62 IP Voice Networks Fax: (607)275-3610 127 W. State Street Mobile: (607)227-2933 Ithaca, NY 14850 e-mail: maureen.stillman(a)nokia.com www.nokia.com

2 1