Earlier this week I ran a post on voice handover from LTE to GSM (or UMTS) and received an interesting comment from a reader who was of the opinion that there is not much benefit coming from such a handover. He said that instead, the industry should make LTE phones in the future that use GSM for voice and LTE for data.
At first I thought, yes, that's what CS fallback is all about, which nobody, me included, really wants. Personally, I don't like the approach because it significantly increases the already long call setup times and falling back to a legacy network for your bread and butter service is not how wireless technology should advance. But then I read the comment again and realized that the author suggested a devcie that could do GSM and LTE simultaneously. An interesting idea!
And it might even be possible as something similar already exists today! Multi SIM phones that have two transceiver units so both SIM cards can be active in different networks and on different frequency bands simultaneously. This shows that devices with two independent transceivers are practical. In the case of the dual transceiver LTE phone there could be one LTE/UMTS/GSM chain and one independent GSM chain that is only used for voice telephony. Both could work with the same SIM card as long as the GSM chain only communicates with the circuit switched GSM network. Should the LTE/UMTS/GSM chain have to fall back to UMTS or GSM, the GSM only chain could be deactivated in favor of a combined treatment of CS and PS. It would make things a lot easier for the network people.
But we are moving towards an IP only world both in the fixed and the wireless telecommunication domain and this approach doesn't fit into this picture. So I keep preferring an IP based voice solution for LTE which has a handover option to GSM and a single transceiver chain in the mobile device. However, I can very well accept that there are other opinions out there concerning this matter so thanks for the comment, it got me thinking!
interesting indeed. I wonder how well those dual-SIM phones worked? I thought there were interference issues that made this much more difficult to make practical and efficient? Cannot recall the details though.
Hi Ed,
Not sure what the probs could be but Samsung keeps releasing such
phones. Here are the latest ones:
http://tinyurl.com/yddmhzo
(sorry, article in German)
Cheers,
Martni
Hi Martin,
I also prefer the IP based voice solution for LTE. It is just a pity that IP was not so spread at it is today when GSM started. But now we have to live (at least for a while) with the GSM legacy networks. But let’s look at the situation today in order to analyze the handover issue for voice. We have 3 types of 3gpp end-users: the voice-only subscriber, the voice+data subscriber and the data-only subscriber.
The voice-only subscriber (like me) is the one that uses his mobile phone for voice. He is in general happy with the mobility. For this user, 3G is totally useless as well as handovers 2G-3G or 3G-2G as he will go for the cheapest mobiles that are GSM-only. And in 99% of the cases, 2G coverage is much better than 3G coverage. Let’s look now at the data-only subscriber, the one that uses a usb dongle: for this one as well voice handover from 2G to 3G or 3G to 2G does not matter as he is just looking at data services with his usb dongle. But of course data handover from 3G to 2G or 2G to 3G matters for him: he would like to take advantage from the EDGE coverage where R99-HSD/UPA is not there. What about the last category? In the last category, the subscriber has a smartphone like an iPhone or a Blackberry. This user is using voice and data services from the same device. In terms of voice services, what is the added value that this user can get out of the 3G network? 3G in fact for voice just bring additional problem: most of the GSM networks today are dual band which make them a bit more difficult to manage with the handovers and 3G adds a third or a fourth layer (some GSM networks have already 3 layers (macro 900, macro 1800, micro cell 900 or 1800). For all optimization engineer, one of the main challenges has been to manage this interworking between 2G and 3G which I think is worth in the case of data as 3G really brings a real value. But for voice? In general the additional layer just worsens the quality. Maybe that the only added value is that you can originate/receive a call while transmitting data. But couldn’t we do that having the 2 networks 2G and 3G independent for voice and a MS with 2 SIMs? Of course those phones are not so common today but I still think that in the future, in the case of LTE, it could be a good transition option until LTE coverage becomes universal. As far as voice service is concerned, 3G coverage just brings headaches to optimisation engineers and does not really improve end-user perception (I would say that in many cases, it degrades it). Maybe that the problem (which is not small) would be the price of such device.
Best regards,
David
Hi David,
There is also the problem of battery drain with the multi-transciever phones. In the future, for example, I feel that the current 3G/Wifi multi-transciever phones will become useless when IP traffic offload is possible near the RAN itself as in femtocells. This will provide excellent IP connectivity without going through the MNOs core network while using the same transciever(3G/4G).
Regards,
Joe