Will Fixed/Wireless Convergence Push IMS?

Despite its multimedia capabilities the IP Multimedia Subsystem (IMS) hasn't yet gotten a lot of opportunities to show its capabilities in wireless networks. One of the reasons for this is that current circuit switched mobile voice telephony works well and meets user expectations. Things, however, are improving for IMS.

With the introduction of the iPhone, more and more people are getting aware of multimedia and Internet capabilities of mobile devices. Thus, it might well be possible that multimedia enriched voice calls might also soon appear on the radar screen of users. On the network side many operators are upgrading their current 3G networks to 3.5G and with the first WiMAX networks rolled out now and LTE on the horizon there is sufficient bandwidth for such services. Additionally, WiMAX and LTE networks no longer have a circuit switched part and operators need a solution such as the IMS to be able to offer conversational services over their next generation networks. It thus seems inevitable, that the IMS will have a bright future.

In practice, however, things will be a bit more difficult since third party VoIP service providers such as Skype, Vonage and others could try to take a piece of the wireless market in a similar way as in fixed line networks today. After all, the application layer does not care whether IP packets traverse a DSL line or the air interface of a wireless network as long as there is enough bandwidth. From my own experience, SIP and proprietary VoIP services such as Skype work well over 3.5G wireless networks and even Skype video calls have excellent video quality in both directions.

Network operator based IMS systems, however, have a number of advantages over voice services provided by third parties if the play their cards right:

Network operators today sell both a mobile device and voice service. This means that the service works out of the box, no configuration required by the user. With pre-installed and pre-configured IMS applications such as voice, video calling, presence, etc. they have a head start over third party services for which applications have to be installed on mobile devices.

The IMS is also able to request a certain bandwidth for a session from the transport layer. In case there is congestion anywhere in the network, it will be made sure that multimedia sessions are not impacted.

The third advantage, which I think is a major one, is that IMS gives network operators with both fixed line (think DSL, cable) and wireless assets (think HSPA, WiMAX and LTE) the opportunity to converge their voice + multimedia service offerings both in the network and from the users point of view.

In the fixed line world the transition from analog telephony to VoIP over DSL or cable is already in full swing. The incentive for the user to switch to VoIP is usually a lower price for a combined voice service and Internet access over DSL or cable. When combined with mobile voice + Internet access, network operators can offer their clients Internet access + voice (and multimedia) telephony both at home, in the office or while roaming outside with a single device and a single telephone number.

The IMS also allows to have many devices registered to the same telephone number. This is great since at home it might be more convenient to use a dedicated phone, a notebook or even an IMS capable and connected television set to make a voice or video call.

With Voice Call Continuity (VCC) there is even the possibility that a mobile device automatically switches to Wifi when the user returns home or to his office thus reducing the load on the cellular network. Switching to Wifi at home also solves the issue of 3G/4G in-house coverage which in many regions of the world is inferior to 2G coverage due to the use of higher frequency bands.

And finally, the IMS has the capabilities to transfer a voice call from one device to the other. This is quite interesting in scenarios in which the user returns home and then transfers an ongoing voice call from his mobile phone to a television set and adding a live video stream to the call in the process.

It's clear that getting all of this right is not a trivial task. But if network operators want to retain their role as a service provider they have to go beyond what third party service providers could offer over a bit pipe.

As always, thoughts and comments are welcome!

Comments

[Reda]

Hi Martin,
just one point not strictly related to this post: Wireless is not mobile. I think you should have titled this post Fixed/Mobile Convergence. I got a wireless router at home, any device connected to it are not mobile. I personally think that wireless access is equal to radio access and therefore is more suitable to Fixed carries because you don't have mobility. On the contrary, if you have radio access + mobility then you are referring to a mobile operator.
-Reda

[Arvind Padmanabhan]

An excellent article. FMC and IMS have a symbiotic relationship. Packet switched connections are on their way to completely replacing circuit switched connections.

[Dean Bubley]

Martin

I mostly agree with you. I've just published a study on Mobile VoIP and VoIPo3G, and certainly there will be moves to cellular-based packet voice for a variety of reasons (see my blog post).

However there is one significant problem - at the moment, IMS has only got this thing called 'multimedia telephony' designated as a service. There isn't a fully-standardised end-to-end approach to do 'plain vanilla' mobile VoIP.

I can't ever see more than a tiny fraction of mobile telephony realisticalling turning into multimedia calls. Video calls are totally inappropriate in most instances. (recent blog post on this too)

It's therefore not really reasonable to deploy multimedia telephony and then have 99% of traffic as the 'special case' of plain telephony. It makes more sense to optimise for ordinary voice calls first, and then have specific enhancements for video etc.

I also think it's going to be about 3-4 years before the majority of HSPA/LTE networks are VoIP-suitable from the operator's point of view. That's a huge window of opportunity for Skype, fring, Truphone etc.

The gap's much smaller on CDMA, though - EVDO Rev A has VoIP 'designed-in', while with HSPA it's accidental.

Dean

[Martin]

Hello Dean,

Thanks for your comment, as always it is very much appreciated. Concerning built in VoIP with EvDO. First, I'd like to see that and second, there is a reason: While UMTS/HSPA terminals can do circuit switched and voice calls simultanesouly, EvDO devices can not. If you want to do a voice call, data transmission stops. Just like in the good old GPRS days. Hence I guess the urgency to go to VoIP quicker. Could be an advantage for EvDO but I think by the time the market arrives LTE and HSPA have cought up (in case there is catching up to do, I am not sure).

Cheers,
Martin

[Mikalaj]

Well, at the moment the main point of IMS is to get full control over IP traffic, and related services and have a possibility to charge users respectively. Because users want to use mobile internet for cheap, and Network Operators want them to use it, and can give it them for cheap, but don't want users to use skype and similar services instead of paying significant amounts of money for operators and e.g. incredible roaming bills. Almost everything else in IMS currently is theory that is hardly getting a lot of attention in operators business development offices.

[Zed]

IMS seems to be a case of too little, too late. Nobody cares about the proposed services and the use cases are weak. Very few things that real end users care about can't be done with circuit switched voice and a dumb Internet pipe.