While HSPA+ and LTE drive data rates higher and higher and have network operators and vendors discussing which is the right way to go, GSM for voice and low bandwidth data applications is unlikely to go away anytime soon. I've speculated in the past about when GSM would be switched off in Europe and elsewhere and wondered if maybe at some point Software Defined Radio (SDR) technology would allow to fold all radio access technologies into the same hardware and into a single digital and a single radio module in the base station.The more I think about it the more interesting such a combined option looks like to me.
With backhaul already converging to IP for GSM, HSPA and LTE, there will be nothing standing in the way from that side of the network in just a couple of years from now. From a handset perspective, GSM might also be the least costly and best technology for the foreseeable future for voice only devices. When I look at my 3.5G mobile stuffed with the latest technology and compare it to the simple GSM phone I use for voice calls only I can not only see a significant difference in size but also in price. After all, a 3G handset does not only have to contain more hardware but the license/patent fees are much higher than for 3G phones. And LTE will further increase the hardware and royalty costs, so there is no break from this perspective, either.
And while LTE and HSPA+ might be optimized for speed, they are definitely not optimized for voice and power consumption when compared to GSM.
A single digital / radio module in the base station would also have another interesting benefit: When only little capacity for GSM voice and GPRS/EDGE data is required in a region the base station could automatically reconfigure itself and use more of the bandwidth for LTE. During busy hour, when voice calls over GSM come close to the capacity of the current configuration, the LTE carrier bandwidth could be reduced and additional narrow band GSM carriers could be fired up within a few seconds. Currently, LTE bandwidths are defined at 1.25, 2.5, 5, 10, 15 and 20 MHz. Maybe not yet fine grained enough but that could be changed in future versions of the standard.
In the backhaul, everything will have converged on IP right up to the MSC Media Gateway and from there the phone call is also sent through the network over IP connections. The H.248 protocol between the Media Gateway and the MSC Call Server is also based on IP, as well as the link to the Home Location Register and all other equipment in the core network. In effect, the once circuit switched GSM network has become fully IP based and only higher layer protocols such as DTAP and MAP are still remaining from the original protocol stack to preserve the super efficient GSM air interface technology for voice.
Unconventional ideas, but who knows what the future holds.
4 thoughts on “Virtual GSM in the Future?”
Isnt this what Vanu mostly perfected? (that is SDR on Intel boxes — i think).
seems like such an elegant solution
Yes, I think SDR might be quite an interesting technology for such multi air interface base stations, too, but I think it can also be done with ‘current’ technology.
Huawei recently announced a product that fits along the lines your post describes. Have a look here.. http://www.3g.co.uk/PR/Sept2008/6647.htm
Thanks for the link, here’s another one that gives additional details and an analysis. Very interesting indeed:
Comments are closed.