Talk:IP Multimedia Subsystem

From Wikipedia, the free encyclopedia

WP:TEL This article is within the scope of WikiProject Telecommunications, an attempt to build a comprehensive and detailed guide to telecommunications on Wikipedia. If you would like to participate, you can edit the article attached to this page, or visit the project page, where you can join the project as a "full time member" and/or contribute to the discussion.

Contents

[edit] Comparison with UMA/GAN

Could this page use some information on the differences between IMS/UMA(GAN) and how they compete, differ and complement each other?

UMA/GAN is about mobile devices, whereas IMS is about any device, fixed or mobile, connected to a big single network through SIP/IP.

[edit] Optimization

I believe that in the article should be add some info related to ADAMANTIUM project, which is a european research project that aims at optimizing current IMS infrastructure with a Perceived Quality of Service aware adaptation management system. For more info please have a look at http://www.ict-adamantium.eu/


I found this article useful but I don't understand the statement:

IMS is an optimization of the network, and investments for such optimization are questionable

Alastair Rae 16:07, 26 October 2005 (UTC)

ChrisBarron I would explain optimisation as follows. If you first of all just take the words as plain english 'optimisation' can be taken to mean make best use of. The IMS protocol suite will be capable of establishing a session between two devices with a minimum of interconnecting stages. In a telecom network, fixed, mobile or even VoIP, each stage generates a charge and therefore revenue to the carrier. There is something of a disincentive for networks to invest in such optimisation. The converse argument is that the network with the most effective IMS will cause users to migrate more and more devices and sessions towards it and therefore increase overall revenue. At this point we start to get into economics rather than technology - but see [| Hotellings law]-- Chris 00:33, 25 Dec 2005

[edit] Expert writer core dump syndrome

Hi, I looked this page to learn what IMS is. The author obviously is well versed in IMS, but I found the article suffers from "expert writer core dump syndrome". It's very hard to follow if you don't already understand the material. I was thrown off by the first paragraph. In one sentence it says IMS is a VoIP implementation... and then the next says IMS is used to offer multimedia services; multimeda services encompass more than just Voip, so the first sentence is negated by the second. Also what does "network controlled" mean in the last sentence, first para? I think you are trying to pack too much info in for the overall good

Point taken, but these first sentences came out of a real "introductory text". Note that VOIP is more than voice, it's really multimedia-over-ip : audio, video, graphics, data, ... The phone operators really want to merge everything together (all Internet services), even in mobile packages. But it's still about communication, not an Internet terminal or something. "network controlled" means that they still want full control over everything what happens (read: you'll have to pay) -- Bluezy

[edit] no advertising links

Several people have been adding links under the "external links" section, that point to various websites of telecommunications companies. At first they were useful, when they were white-papers, that explained what IMS was. But what's the point if there are a dozen links to similar content ? Wikipedia is an encyclopedia, not a directory listing for advertisement purposes.

Last week, someone tried to add a link that was called "Leading provider of SIP MobileVOIP based Solutions" (I won't mention the name of the company). That was a pure advertisement, and didn't add any value to this page.

I finally decided to remove all links, even the one from my own company (that was pretty neutral), except the 3GPP ones (that define the standard), and the one from LightReading (that present a neutral overview). I even removed the one from my own company. When a good link is presented, that explains IMS or a particular aspect of it, we can accept it. But please, no more marketing papers or products placement.

--Bluezy 21:45, 6 December 2005 (UTC)

[edit] RFC =

Shouldn't the link for RTP's RFC in the "IETF Specs" section point to RFC 3550 instead of 1889 as 3550 obsoletes 1889 ?

Mat 86.203.206.68 12:19, 16 December 2005 (UTC)

You're right, RTP was updated in 2003. But most SIP phones only implement RFC 1998, the changes are not so important for small-scale implementations (see RFC 2550 appendix B). --Bluezy 15:09, 16 December 2005 (UTC)

[edit] I-CSCF x IBCF

Latest issue add: "From Release 7 onwards this function is removed from the I-CSCF and now part of the IBCF (Interconnection Border Control Function). The IBCF is used as gateway to external networks, and provides NAT and Firewall functions (pinholing). " Could you please update also the figure to show the IBCF?

[edit] Image request: Interfaces

The table of interfaces needs a diagram (or at least a very prominent link to one elsewhere) to show where these various interfaces logically reside. At one time (late October '06), I recall there was a diagram immediately above the table but it seems to have been lost. If it was removed because of an inaccuracy or usage problem, it should have been flagged for replacement, not just trashed. DKEdwards 22:59, 15 March 2007 (UTC)

 It is requested that a diagram or diagrams be included in this article to improve its quality.
For more information, refer to discussion on this page and/or the listing at Wikipedia:Requested images.
Diagram was deleted due to the license problem (it was a image scan from a book). I think, that image http://www.tech-invite.com/img/ims/ims.gif can be added, but I'm not sure about copyrights also. Alny 16:40, 19 March 2007 (UTC)

[edit] Categorization

Regarding the removal of IP Multimedia Subsystem from Voice over IP:

Since it is included in VoIP protocols, it should not be included in the parent, unless there is a specific reason.
- TJJFV 00:59, 24 March 2006 (UTC)

[edit] Advantages and Issues

I've been meaning to write why I tagged this section but have been distracted. So here goes:

The entire Advantages and Issues section is terrible. The section begins with weasel words and descends from there. Ignoring the awkwardness, particularly the "Debate" in the second section, there appears to be a complete misunderstanding of context, in terms of where IMS is intended to be used.

IMS's intended role can be summed up as:

I am Megatelco. I want to migrate my network to 4G (WiMAX, LTE UMTS, etc), I want my customers to be able to purchase integrated devices that work on my network (from me, from others) and "just work". For example, if they buy a cellphone, the cellphone isn't going to "just work" if the customer has to install Skype or Vonage before using it, and organize service provision from third parties. If it, on the other hand, supports the collection of standards Megatelco supports, then all the user has to do is insert the SIM card, and their phone works. It works as a phone. They can sell it as a phone.

In the context of that, most of the questions and answers are either ridiculous or irrelevent or both. Does IMS replace free VoIP? No, it doesn't. Does free VoIP suffer through the existance of IMS? Nope. Does the pre-existance of ISDN or GSM indicate IMS is redundant? Erm, nope, because, GAN aside, neither runs over IP, and GAN is hardly an ideal, complete, integrated, efficient, system.

There's no point/counterpoint necessary here. Either you "get it" and understand the role, or you make irrelevent comparisons to technologies that do not deliver the intended purpose of IMS. You might just as well go into a point/counterpoint about how entirely unnecessary HTTP is when we have UDP packets.

HTTP means we can deliver files in a standard way
Counterpoint: But files are just data. They can be transfered in UDP packets
Countercounterpoint: UDP packets are too small and do not guarantee delivery
Countercountercounterpoint: You can send lots of them and use delivery confirmation packets

(The actual real point is you can't make a web browser or a web server without a standard for transfering pages, and an ISP, OS vendor, and computer maker would find it hard to provide integrated interoperable solutions without having those protocols set up.)

I'll try to rewrite this over the coming month, but if anyone wants to make a start (or just comment on the above) I'd be delighted. Squiggleslash 15:16, 24 January 2007 (UTC)


I agree completely. It was originally a small list of items to help compare with other systems (Advantages & Issues)), but then people started to add counter-points, then counter-counter-points , etc ... Especially what is now the 'Differences' section. It currently looks like something for the Talk-page, but doesn't belong in the main article. --Bluezy 10:57, 26 January 2007 (UTC)

Ok, I did a first attempt at a rewrite. I hope I haven't offended anyone with it, but I have tried to emphasise the context of IMS, which hopefully will reduce the likelihood of these kinds of "debates" appearing again. I tried to keep the same information in, though in many cases the rewording leaves the information there by implication. Any comments would be welcome.Squiggleslash 14:31, 26 January 2007 (UTC)


How about a first sentence that is understandable by a normal human? After that, run with it, but this is a general use encyclopedia so the top level should try to provide clarity. The article is beautiful to behold, but opaque with deeply nested jargon. I am a programmer who gets "Telephony" magazine in the mail for reasons unknown - all jargon, unexplained acronyms, hoped to get a clue here. I did get a clue but there is obviously a culture gap between telco folks and the rest of humanity. C Ross feb 6, 2007

Any better? --Squiggleslash 15:13, 8 February 2007 (UTC)

Squiggleslash - Let me reword your argument slightly differently to reflect an alternative way in which IMS may be perceived? After all - all the potential views should be represented.

I am Megatelco. I want to migrate my network to IMS based systems. I would like to ensure that things only "just work" when they pay us for our own premium service. I would like to ensure services like "vonage", "skype" as well as "youtube" do not "just work" and reduce usage of my more expensive phone tariffs and are bandwidth limited in comparison to our own services. I want to access control 3rd party provisioning and protect our revenue before some aspects of our business model slip away while consumers find cheap or worse still free services. I want to avoid becoming "just a utility".

This could go two ways - it could be used as an enabling platform to make for a richer, smoother user experience, or it could be used as a cynical, competition crushing, monopolistic, cartel forming economic weapon. I will leave it to the success, failure, and market to determine exactly what it works out. Some of us as shareholders win either way. 213.46.174.149 16:36, 9 February 2007 (UTC)

Except that has nothing to do with IMS. If Megatelco wants to block Vonage, Skype, and other third parties, support for IMS has absolutely nothing whatsoever to do with the ability to do so. Your rewording is inaccurate, reeks of paranoid hysteria, and does not belong in a discussion about IMS. I'm amazed you'd even think it would be.

Let's keep the page to a technical discussion, rather than a more general discussion about the behaviour of the telcos when that behaviour has nothing whatsoever to do with the topic at hand. --Squiggleslash 16:51, 9 February 2007 (UTC)

Squiggleslash, I would tend to agree with you that such views are quite hysterical, however, I have heard similar things uttered in the board room of a group who supply technology to a single MegaTelco, which boiled down to rolling out IMS and then traffic shaping Skype to gain revenue. It is not as unlikely as it may have sounded. I don't think IMS is the technology that would do the traffic shaping (at least my understanding is that it does not encompass traffic shaping), but that it could be coupled with Anti-NN (Network Neutrality) technology and lobbying to maximise revenue. All-in-all it could actually work out pretty well for Telco's if they collaborate on it (to stop people tarting and just switching to the few providers who remain neutral). 82.153.19.100 13:05, 10 February 2007 (UTC)

So what is the relevence here? My point was that the bizarre "Point...Counterpoint" thing wasn't encyclopedic and definitely was irrelevent to the article, with the writers missing the point more often than they stayed on-topic. IMS has nothing to do with traffic shaping or operators controlling the network. It's an open standard that anyone can implement. It neither helps nor hinders the blocking of third party VoIP services. It's merely a way to ensure that someone can produce standardized integrated devices for 4G networks that work the way both operators and customers expect.
Or is the point that Megatelco wants to sell VoIP, and even that is wrong? Because it isn't. Megatelco is, today, selling voice services. They want a better platform, a more efficient, more flexible platform, in the near future. Migrating their own network to use IP as an underlying protocol is one way to achieve that. What legitimate interest is served by saying "Yes, you're allowed to upgrade your network, but you can no longer sell the same services"? --Squiggleslash 13:31, 12 February 2007 (UTC)

Please may I point out that the naming of this section betrays a misunderstanding of what IMS is i.e IMS is NOT another form of VoIP! Note also Squigglesquash most of what is written here about Integrated service benefits or criticisms are personal opinions that have little place in a discussion of IMS. We need to be honest: given the changes and discussions on this page, it is clear that IMS is something of a hot issue and I would rather see some of the benefits and/or criticsms of the claims of IMS rather than a side discussion on "integrated service benefits". Examples: often there are claims made that (1) IMS aids service creation (2) IMS helps charging and so on. Critics respond respond indicating why these statements can be challenged. This is what should be in Wikipedia, people should come here get the basic facts understand that there is contention and walk away. Additionally, it should be clear there are commercial sensitivities here, some companies have spent (and are spending) a lot of money developing technology that has hasn't quite caught on, and true some operators are talking about it but in general operators have been ambivalent towards IMS. I don't want to jump in and make all the changes now, i would rather that you guys thought a bit more about this and then we could together do something better here. wirelesspond 13:10, 12 February 2007 (UTC)

IMS is a way of implementing (amongst other technologies) VoIP so it's reasonable it would be compared to other forms.
Either something is a statement of fact or it isn't. If it isn't, then you need to say in what way it is false, and we'll remove the statement from the page. The article is NOT the place for a debate and the way it was worded was terribly unencyclopedic.
I stand by every statement I've made. Every statement I've made I've ensured is as accurate as it can be within the context given. I don't really see contention either, I replaced a section of Wikipedia that had barely anything to do with the topic at hand with relevent text. Do we seriously give room to debates on the article pages especially when those debates have nothing to do with the topic at hand?
As far as the latter comments go, the LTE project hasn't reported yet, and UMB is similarly yet to be finalized. IMS is, for the most part, intended for all-IP networks. With the exception of a handful of WiMAX networks, that aren't the intended audience for this technology, what all-IP networks exist right now? How many 2G and 3G operators are replacing their networks with all-IP networks today? How can something be judged as failing to have "catched on" when nobody expected any serious roll out of the technology yet? Has Mac OS X Leopard similarly been a total flop? --Squiggleslash 13:31, 12 February 2007 (UTC)


Squiggleslash, yes, either something is fact or not. You have titled this "IMS and other forms of VoIP". BUT IMS is not VoIP, it is just as absurd as writing "SIP and other forms of VoIP". Your paragraph 3: (a) The baby has been thrown out with the bath water by your editorial removal of a part of the entry (there are some important points made here by both sides and a synopsis of them would have been very helpful) and (b) your comments, IMHO, about Integrated services are largely tangential here - i do appreciate that you were "trying to keep the peace" and i do not begrudge these opinions (if you want to retain them perhaps you should start a new Wikipedia entry on integrated services or something like that). Your last paragraph is symptomatic of some of your comments/entries here i.e. there is an element of truth in them but they often don't represent the whole picture to the extent that they end up being misleading entries. IMS is not all about tomorrow: 3GPP R5 is about today - and the standard has been available for ages. Nothing stops people putting in CSCF's and HSS's in their networks, they don't need All-IP networks to do this. The commercial fact is that several major vendors are actively selling IMS today for deployment today. Many companies claim IMS is central to FMC *today*. The second half of para 3: I am bewildered, as I am active in and aware of the huge debate going on in the industry at the moment. This should be an entry for an encyclopedia, people have a right to know both the technical and commercial issues. It is perfectly OK to say here that the merits of IMS are not universally accepted (there are several articles and analyst reports on the matter). Wirelesspont 22:20, 12 February 2007 (UTC)

1. As commented above, IMS is a way to implement (amongst other things) VoIP. No, it's not the same as saying "SIP and other forms of VoIP" because SIP really is just a transport mechanism, whereas IMS standardizes VoIP services (amongst others), that, indeed, is one of the things that distinguishes it from SIP.

2. It would help if you actually talked about what it is you disagree with and described your objections. If you're referring to yesterday's edit, I removed something that had previously only been there "to keep the peace". The comment was bullshit, and needed to be removed. It is undeniable that the existance and implementation of IMS has no effect whatsoever on an operator's ability to block or degrade the service of competing VoIP operators. That's fact, not opinion, and would easily be proven wrong if it were not so.

3. My last paragraph is not in any way misleading. IMS (and FMC) is not about "today", it's being developed for tomorrow. Yes, there are prototype and early releases of the system intended to allow operators of existing networks to put their toes in the water before high bandwidth/low latency networks required for the roll out of more mainstream networks become available; and yes, naturally, those who are forward thinking enough are trying to build it into new devices and technologies to keep them future-proof; but the simple FACT of the matter is that no operator, today, could roll out a 100% all-IP IMS based network if they wanted, because the technologies are simply not ready yet. What's the available bandwidth in an HSDPA/HSUPA network? IMS is a forward looking technology, designed as an early component of what's to come, not a current technology that anyone expects to be mainstream today. Why do you think it's based around IPv6? (And no, the fact some people may be pushing today it isn't something I care about. That's not the goal.)

As far as your last paragraph goes: It is not acceptable to have a debate, let alone a "Point...counterpoint" section, in an encyclopedia. This is what was removed and what you appear to be criticising me for. The current entries, which can be expanded, are truthful and unlike what was replaced are actually relevent. The complaints that were made in the "debate" are utterly irrelevent to this article - IMS is not a plot (whether they have one or not) by big telcos against those without infrastructure, because it can't act that way. IMS does offer advantages for networks that want to operate All-IP networks (which is what both UMB and UMTS release 8 are. Circuit switching is not going to be a choice for them) based upon being a standardized, integrated, technology. The fact the technology replicates ISDN is not a "negative" given the technology is intended to implement the same features in an environment where circuit switching will be unavailable. etc.

Regardless of what the intentions are of mobile operators from a competition standpoint, they're going to want an integrated standard for voice and similar services when they migrate from current circuit switched technologies to (very near) future all-IP systems. That's the context of IMS. That's what it's here for. It may or may not be better or worse than competing standards, but those points have yet to be raised by anyone. --Squiggleslash 13:38, 13 February 2007 (UTC)


1. IMS is not a form of VoIP. IMS does not standardise VoIP. Please point me precisely to the IMS specification which does this. If you don't have easy access to the very large IMS specs please read Camarillo's or Poiksella's books (as a pointer Camarillo's 381 page book on IMS mentions VoIP in three places (pgs 5, 7, 259) and never more than two sentences in each place "VoIP" appears). And never is it claimed that IMS is a form of VoIP or vice versa. Simply put: IMS is a component of a generic architecture for offering services for IP.

2. The whole of the "IMS and other forms of VoIP" section needs to rewritten. The only reason a section is needed here is to indicate specific problems and issues. The problem(s) here started with people making some claims related to IMS hype. Some people objected to these claims and we saw the result ....

3. It seems that we live on different planets. Yes it is true that all-IP is aspirational but this should not detract from the fact that IMS is here today. IMS is being sold as a to do FMC *today*. This is whole reason for the resurgence of the IMS hype. The original hype curve (directed towards wireless opertaors) tailed out as it became clear that benefits such as service creation and charging were not quite what they seemed. Then fixed line operators were sold the idea of IMS for FMC. Hence the new hype curve. This is not new, lots of technologies get hyped. Then either the technlogy peters out or finds some good use. Pop over to 3GSM in Barcelona, this week, and have a look at some of the stands. You will have a great opportunity to understand what is happening in the telco industry (including the IMS space). BTW, if you don't agree that IMS is being sold today then do a google serach and provide an explanation for what is going on in the industry.

4. Your opinion about should be in or not in a encyclopedia is your opinion. This encyclopedia is more than your opinion. I am not asking for a debate in the main entry. But I am asking that if there are various view points they get reflected. I believe some of what was there before was very very relevant (some wasn't). Several of the comments you make above, starting "As far as you last paragraph ..." don't resonate with me. You are too focused on some of the wrong statements somebody/some people have made; not withstanding the wrong statements there were lots of very accurate statements too, as I previously said you have thrown the baby out with the bath water.

5. If all you are saying was acceptable and agreed then this discussion (peoples comments, our exchange) would not be occuring. True some of the comments may be incorrect, but I don't sense maliciousness in the comments, so we should carefully consider the content and the intent before we sweep them aside.

I could have stepped in and unilaterally made some changes but instead I am inviting you to consider that your view and opinions here may not necessarily be right and are certainly not agreed. I will of course make changes in due course but I wanted to have a discussion first. -- Wirelesspont 12:30, 14 February 2007 (UTC)


1. Your reasoning is specious. Nobody would expect the letters "VoIP" to appear in the specs. IMS was developed so that mobile operators could offer the same services over IP as they do circuit switched networks. The standard doesn't even need to standardize codecs, given those are higher level concepts already defined by implication of what goes over the virtual circuits IMS implements. The primary purpose of IMS to allow mobile operators to offer mobile phones that work over all-IP networks. That is, by definition, voice over IP.

2. It may well do, but you haven't identified an issue other than making the bizarre claim that IMS is not used to implement standardized, interoperable, voice/etc services over IP. All you've done so far is claim I'm "expressing an opinion" and that I went too far in removing a debate from the middle of an encyclopedia article.

3. Perhaps you'd like to reword point 3 so it makes some kind of sense? (OTOH, don't bother, see below)

4. Yes, you are asking for a debate in the main entry. If you're not, you wouldn't be criticising me for removing it. And my opinion about what might be valid for an encyclopedia may be an "opinion", but your pointing that out suggests you have no counter argument. "Oh, what you say is just your opinion anyway, and BTW I was never arguing (what I've been arguing)" Very constructive, well done.

5. I removed a un-encyclopedic debate from the middle of an article. I ensured all the FACTUAL and RELEVENT points made in that debate are still mentioned in the replacement. I disagree with the only other two criticisms you've made - the suggestion that IMS is not in any form VoIP, despite the fact that its entire reason for development was to ensure mobile operators could offer the same standardized services over an all-IP network as they do using their current circuit switched systems, and the suggestion that it's success or failure can be judged today when no 4G networks actually exist and are expected to exist. (As if the success or failure is actually important right now anyway.) If you want to say "It's failed to take off in areas outside of its original purpose", then that probably is a judgable comment (though any inroads into applications outside of 4G services constitutes a measure of success), but to judge the entire thing as a failure when the lower levels of future all-IP replacements to 3G UMTS and CDMA2000 have yet to be standardized is exceedingly premature.

I'm sorry it appears to have pissed two people off, but

  • What was there was inappropriate. I don't care some think that's an opinion, I can't see how a debate is a legitimate part of an article that's supposed to pass on facts. I'm glad you appear to now agree with that, even if you do think it's just my opinion! The replacement covered all relevent issues raised in the "debate" in a factual way.
  • IMS does not in any way help or hinder the world conquest ambitions of large telcos, it just makes 4G viable. No matter what VoIP system you adopt, you can port block or degrade the quality of traffic not matching that profile. You can use exactly the same protocols as Vonage, and make Vonage unusable. IMS does not in any way help or hinder your ability to cut off competitors.
  • IMS is (amongst other things) a VoIP solution for future all-IP networks that replace 3G UMTS (and possibly others.) It provides the virtual circuits needed to replicate the circuit switching of older network technologies. Its primary reason for development is to make voice services work in an all-IP context. It is as much of a VoIP system as IAX, and more of one than SIP. People will, legitimately, make comparisons.
  • Judging success and failure is of questionable merit anyway, and right now would be premature.
  • The context of the "debate" was wrong, as evidenced by the concern that IMS was redundant because ISDN already exists, the allusions to conspiracies by telcos, and others.

I'm sure you can find something that's actually factually incorrect in my rewrite, that's what the "Edit" button is for. Right now though, I'm not seeing it. As it is, I'm beginning to wish I'd just deleted the entire section, rather than trying to make a pigs ear into something informative.

I'll leave you all to get the last word. I'm removing this article from my watchlist. Between the conspiracy theorists and the unconstructive critics this whole thing is just leaving me depressed. --Squiggleslash 13:54, 14 February 2007 (UTC)

Squiggleslash dude, Common man don't be depressed. I'd buy you a beer, you make a LOT of good points. I don't think that everything that you are saying is hogwash - because it isn't. Man, VoIP is not IMS - implied or otherwise. VoIP can run over IMS; end of story, dude. Yeah, I accept that debate should not be part of the main entry; all I am saying is that if some "facts" are not accepted as facts then we need to recognise that in the article - and if there are opposing opinions (where relevant) they need to be indicated (this is not quite the same as debate - simply an indication of opposing view points!). BTW, I agree with you, and repeat again, some of the comments that people made against IMS, in previous versions of this article, were simply wrong like the ISDN one you quote(!) Man I don't want to have the "last word", I wanted us to agree! This article is going to seen by many people, I want to it to make sense, I want it to a be a door which allows others to see that there are some diverse opinions here! -- Wirelesspont 22:44, 14 February 2007 (UTC)


I think we should start over. I have edited most of the content here, and added the first Advantages/Issues section here. But it grew out of hand, and people started to add points, counter-points and counter-counter-points. Really ugly. This is an encyclopedia, not a discussion forum. In my opinion it should be something like 20 lines maximum, like it was originally.

Note that I'm not a systems architect, but I work for the largest telecom manufacturer, and I'm one of the core developers.

This was how it was 1 year ago :

Advantages & Issues

Advantages over existing systems

   * The core network is independent of a particular access technology
   * Integrated mobility for all network applications
   * Easier migration of applications from fixed to mobile users
   * Faster deployment of new services based on standardized architecture
   * An end to unique or customized applications, leading to lower CAPEX and OPEX
   * New applications such as presence information, videoconferencing, Push to talk over cellular (POC), multiparty gaming, community services and content sharing.
   * Evolution to combinational services, for example by combining instant messaging and voice
   * User profiles are stored in a central location

Advantages over free VoIP

It's possible to run free VoIP applications over the regular Internet. Then why do we need IMS, if all the power of the Internet is already available for 3G users?

   * Quality of Service : The network offers no guarantees about the amount of bandwidth a user gets for a particular connection or about the delay the packets experience. Consequently, the quality of a VoIP conversation can vary dramatically throughout its duration.
   * Charging of multimedia services : Videoconferences can transfer a large amount of information, but the telecom operator can't charge separately for this data. Some business models might be more beneficial for the user (for instance: a fixed price per message, not per byte); others might charge extra for better QoS.
   * Integration of different services : an operator can use services developed by third parties, combine them, integrate them with services they already have, and provide the user with a completely new service. For example: if voicemail and text-to-speech is combined, a voice version of incoming text messages can be provided for blind users.

Issues

   * Benefits need to be further articulated in terms of actual savings.
   * IMS is "operator friendly" which means that it provides the operator with comprehensive control of content at the expense of the consumer.
   * IMS uses the 3GPP variant of SIP, which needs to interoperate with the IETF SIP.
   * IMS is an optimization of the network, and investments for such optimization are questionable.

-- Sorry man, the "advantages/issues" section that was here a year ago started the problems in the first place - i really hope that what we eventually put here is different from what was here before. The debate is occurring becuase of the misinformed entry. Fully agree that this is an encyclopedia not a discussion forum. Nevertheless, encyclopedia entries should contain facts and they should also indicate, where relevant, that there may be disputes. Some of the arguments that have appeared here and in the article can be disregarded as they are factually incorrect. But some can't and need to be considered for the rewrite. The old write up, mentioned a moment ago, and is being proposed to be reinserted now, has several fundamental flaws; e.g. it is wrong and even absurd to write about "advantages over free VoIP" because IMS does not compete with VoiP, one is an architecture component and the other is an application! Similarily, while not totally inaccurate the entries under "QoS" or "Charging" are misleading in this context (and so on ...). I don't have a lot of time but will come back to this soon and slowly build it up again. Wirelesspont 20:27, 15 February 2007 (UTC)

Wirelesspont, the remark about the old version of the section was from me (bluezy), not from Squiggleslash. And I included it here instead of pasting it in the page again, so that was can improve it, not discussing it again. Note that most of the section came from the Camarillo/García-Martín book, and from marketing material from my company. Hence the section called 'advantages', which might have been better called 'differences. --Bluezy 19:19, 18 February 2007 (UTC)

ok, let's start again. I'll avoid the titles 'advantages' and 'disadvantages' (like in marketing material), just highlighting the differences with other systems. Documenting and informing the public about them is the purpose of Wikipedia, not opening a discussion.

(note: this is still a work in progress !)

[edit] Differences

Here a few of the differences with other phone systems. Note that these differences can also be achieved in other ways. IMS is an implementation of VoIP by the existing telecommunication operators, but is not the only possible one.

[edit] Differences with existing phone systems

  • IMS promises access independence - it can be used over all networks : fixed, mobile or wireless. This means that a user can be reached with a single phone number, redirecting the call to the cheapest or most appropriate network.
  • Interaction with existing phone systems is provided with gateways. This not only allows to call people on non-IMS networks, but can also provide those people with advanced features, without modifying their existing telephone exchanges (which are embedded in IMS). For example, a POTS can receive a videocall by having it redirected to another device, or by storing it into a videomail server. Existing applications like SMS can be integrated with Instant messaging services.
  • IMS is a standardized architecture, which should allow for faster and easier development of services (previously, services had to be developed separately for different phone systems, because of the proprietary architectures).
  • User profiles are stored in a single central location (the HSS), even when shared by different servers, which makes provisioning much easier.

[edit] Differences with other VoIP systems

  • IMS promises gateways with existing phonesystems (including [[Plain old telephone service|POTS). While other VoIP systems can provide the same, IMS is implemented by the old telecommunication operators, and they might be better placed to do the integration. Or they could make it more difficult to interoperate, by charging high interoperate fees, which they won't apply to their own IMS customers.
  • IMS is designed to use IPv6, unless Early IMS is used. This might make its introduction more difficult since not every device is ready to use IPv6.
  • IMS networks use the Internet protocol, but that doesn't mean that you can easily use personal computers with soft phones. Telecom operators might only allow certain VoIP software to talk to their systems. Or they might even make sure to block other software like Skype, in order to protect their income.
  • IMS uses the 3GPP variant of SIP, which needs to interoperate with the IETF SIP. This might force changes to existing SIP clients. For example, secure authentication is done using AKA, which can only work with a USIM. This limits a soft phone to the lesser secure HTTP Digest.

(note: this is still a work in progress !) --Bluezy 22:35, 18 February 2007 (UTC)

Bluezy, I understood the remark about the old section was from you and not Sguigglelash - and I understand that you were suggesting that it be used to start a rewrite. I don't think that the old entry is a good basis for a section here. In terms of quotes from Camarill/Garcia Martin - if you are quoting them make it clear where and how. Quoting out of context misleads. In terms of your marketing dept sourced material well i don't want to say much about that other than it is helping to confuse. The main contention that I have and I repeat it again it is simply wrong to suggest VoiP is a form of IMS or IMS is a form of VoIP. This is factually incorrect. One is an application the other is an architecture. Camarillo et al. nor Poiksella et al. nor anyone who is actively working in this 3GPP standards space would say such a thing. The titles "Differences with existing phone systems" and "Differences with other VoIP systems" are just simply wrong or misleading (in this context) and most of the content below those titles is bland stuff that doesn't mean anything or is out of context and so can be misleading. I will get back to the article soon, but won't use the above suggestions as the starting point. Wirelesspont 11:48, 24 February 2007 (UTC)

ok, here goes the application versus architecture debate again. Sigh. Ok, I go back to actually implementing and testing the stuff. And selling. --Bluezy 13:24, 24 February 2007 (UTC)

Wirelesspont, why did you remove the section about Early IMS, and change it in a section about Early IMS Security ? It's not only security that is a problem in early deployments, although it's a very large part of it (I think you mean the lack of IPSEC and AKA). In my experience, most SIP-phones aren't even capable of working in IPv6, use only Http-Digest, do not use DNS NAPTR, do not use registration-subscriptions, etc ... It will be a long time before we'll see real IMS phones and networks, so there is a need for early IMS networks. In my experience with tests and customers demonstrations, the most important aspects are support for IPv4 and http-digest. The rest will come later. See TR 23.981 "Interworking aspects and migration scenarios for IPv4-based IMS implementations (early IMS)". --Bluezy 23:34, 4 March 2007 (UTC)

Bluezy, The original entry said: "There is a need for an IMS version that offers similar features but doesn't require the investments necessary for full IMS. This version has been defined as Early IMS, and has the following advantages:" I also note that in your comment above you refer to TR 23.981 "Interworking aspects and migration scenarios for IPv4-based IMS implementations (early IMS)". (1) The TR (whether v6.4.0 (2005-09) or v5.0.0) is actually called "Interworking aspects and migration scenarios for IPv4 based IMS Implementations". OK, so what is the point? Firstly, 3GPP does not define "early ims". It simply recognises that there will be implementations which do not have, for example, IPv6 support, and that 2G phones may need to use "ims services" without having the benefit of a USIM interface. Is it IMS or is it not? Even TR23.981 quite correctly refers to IPv4 based implementations. (2) It is silly to talk about "advantages" because "early IMS" is an informal term used to describe a transition phase. (3)Wrt to your example above, note that 3GPP specs rule out http digest as an authentication method even for so called "early ims implementations". But admittedly the standards are evolving as there are some moves afoot to include "http digest" as a authentication method in TISPAN. I am not averse to some comments being made about early implementations of IMS lacking certain capabilities and interworking to these implementations will be needed (perhaps in the arch section). Wirelesspont 13:56, 5 March 2007 (UTC)

[edit] Missing interfaces

3GPP TR 24.880 is mentioning the interfaces Sr/Cr which are not in the table. Additionally Mb - the RTP interface between MRFP and endpoint devices - is missing. —The preceding unsigned comment was added by 195.233.250.7 (talk) 10:55, 26 March 2007 (UTC).

done --Bluezy 10:24, 28 March 2007 (UTC)

[edit] Application Servers and Applications

My addition of IP Centrex to the list of services generated a back-and-forth reverting session that kind of surprises me. The list of services does not claim that these services are IMS-specific; it is just a series of examples of services that can be implemented by application servers on top of an IMS core. Can I ask that an agreement be reached as to the purpose of the list of applications in the Application Servers section? JanCeuleers 13:14, 11 April 2007 (UTC)

In my opinion the list was not really useful, it added to the length of the article and more served as a bit of a propaganda item (as did the comment that the AS makes it easy to integrate applications in). Suggest that the list should be minimal and really only refer to some specific IMS things. I am fine with, for example, VCC as this is is being dealt with actively in the standards environment. What is the point of adding mms, sms and every call feature one can think off? Then might as well put every mobile application under the sun and make this completely and totally useless. So I am grateful to you for adding IP centrex as that was the "tipping" point which made me go back and think "hey, what are we doing here?". There are also some other issues, e.g. some examples equated the AS with the whole of an OMA enabler such as PoC. PoC could be a AS but is not necessarily so. BTW, some thoughts on rewording this section won't go amiss .... 194.72.92.36 14:04, 13 April 2007 (UTC)
I believe that the point that IMS can support "mundane" applications (such as voice mail and supplementary services such as call forwarding) is important and worth making. It conveys the message (by example) that IMS is not a "fly-by-night" architecture aimed at providing new services that might not remain relevant for very long. It demonstrates that IMS is sufficiently generic an architecture that it can (in time) supplant the legacy rather than having to be implemented as an overlay. I'll grant you that a mere list of such services is perhaps not the best way of making that point. Such a list would certainly not need to aim to be complete either. JanCeuleers 17:14, 13 April 2007 (UTC)
OK. Several fair points here. To address the point we need to state here that IMS may used to provide present services *too*. Lots of patchwork is being done to IMS - the replication is proving to be non-trivial hence all the interesting additional work in 3GPP R7, R8 and also TISPAN and so on. I note that there are several pieces of on-going working to, for example, perform interworking to legacy systems - often non-trivial. 194.72.92.36 13:01, 19 April 2007 (UTC)
  • removing the OSA-SCS reference ? That doesn't make any sense since it's listed in the sentence before that. And if OSA-SCS has to go, then should IM-SSF go too. Why remove one and leave the other ? If the idea is to simplify, then all other references to the 3 types should be removed from the document (like in the table).
  • The phrase "This allows third party providers an easy integration and deployment of their value added services to the IMS infrastructure" is not a reference to the fact it would be easier to develop application servers, since there would be only a single interface that fits all vendors. Something which doesn't exists today. Nobody said that it would be easier than existing servers (it could better or it could be not).
  • and if don't like a list of the different services that can exist in IMS (or in other systems), then it should be added as a separate page. Information is currently spread over many documents, most of which are in Category:Telephone service enhanced features and Category:calling features (but not all, it's out-dated anyway). Note that it's about services, not application servers.

--Bluezy 17:03, 13 April 2007 (UTC)

[edit] What is service layer?

Telecom manufacturers like Ericsson often talk about the service layer. What OSI layer is that? It is mentioned in this article. There should be an article about it. Mange01 14:58, 17 October 2007 (UTC)

Who says that this is an OSI layer ? It's a name for everything that deals with services in the 3GPP model - all the application servers etc ... Not necessarily provided by the same company that gives you the basic connection. A bit similar to Local loop unbundling.
Services offered to whom? Why not write an article on Service layer? Mange01 (talk) 16:16, 24 January 2008 (UTC)

The terms "service layer", "network layer" and "IMS layer" are used by ITU, ETSI, 3GPP/3GPP2 to describe functions of their NGN-switch concepts. For IMS "service layer" is used as another word for application layer, in the ITU slang the term also includes the functions of the IMS layer. Those layers are populated not with network protocols, but with networking devices like SIP switches, SBCs and so on.--Kgfleischmann (talk) 10:26, 26 January 2008 (UTC)


[edit] 'NGN Interconnection' section

(I think that the 'NGN Interconnection' section should be rewritten) --Bluezy 09:55, 18 October 2007 (UTC)

[edit] Relation to Intelligent network services?

Is it correct to say as in the Intelligent network article: "The upcoming IP Multimedia Subsystem (IMS) standards can be seen as a hybrid of intelligent network services and Internet services for cellular multimedia communication." Mange01 15:04, 17 October 2007 (UTC)

No, IMS is a way to utilise packet networks for tecommunications and so replace PSTN (POTS and ISDN) and PLMN. IN added the possible use of external databases for call control decisions in telecommunication switches. ETSI defines the relationship between IMS and IN: "The purpose of the IM SSF is to enable access to IN service logic programs hosted in legacy SCPs". --Kgfleischmann (talk) 10:26, 26 January 2008 (UTC)

This discussion continues at Talk:Intelligent network#IN and IMS. Mange01 (talk) 09:21, 6 February 2008 (UTC)