The “Next Generation Mobile Networks Alliance has many of the major mobile operators as members and is “a forum to share, assess, and drive aspects of mobile broadband technologies”. The focus is on LTE & EPC (Evolved Packet Core) and its evolution.The first “NGMN Innovation day” was held i Stockholm on September 15th. Quite a number of prominent researchers from academia and industry made presentations on cognitive radio systems and spectrum issues. It seems that the “hype” with cognitive radio and cognitive systems continues. As a footnote, personally I have difficulty to see what is “cognitive” (as opposed to adaptive control) in most architectural proposals
Anyhow, one of the highlights of the workshop that really puts the finger on where research should go was the one by Yoshihisa Kishiyama from NTT DOCOMO. He made an analysis how the 500-1000 fold increased capacity matching projected user need to 2020 should be achieved. He accounted for a 10 fold improvement in spectrum efficiency (questionable, I think with the energy constraints), 2 times more spectrum (yep, indoor, no problems) and 25-50 times more base stations and off-loading (i.e. even more base stations). I think Kishiyama-san is right in his analysis … and the focus is now clear, we need to get concepts to bring massive amounts of base stations out there at a cost that is 25-30 times lower than today. I think this is feasible – these base stations are mostly indoors, they are deployed like WLAN access points (i.e. very robust plug-and-play installation). In a sense network complexity has to become lower than today if this is going to work. Yes, devices/access points can have more complex electronics and algorithms (self-organizing), but these can hardly be reliant on complex interaction and cooperation with other devices.
At the closing pane, I also asked the question “when will the cost level for deploying infrastructure reach high enough to force you (operators) to cooperate?”, or rather enter a coopetition mode like in physical transportation, grocery’s and gasoline ? There are no parallel railway track – instead “operators” together provide the service everywhere and the customer can choose which offer to pick “on the fly”[1]. Unfortunately only mumbling answers .. our conditions are very special etc. On the other hand, there are few markets where technology is more standardized, so there are hardly any technological barriers [2]. Certainly, if the whole business valuation is based on how many subscribers you “own”, its a difficult proposition.
Two items still puzzle me after the workshop – who can afford 1 Gbit/s high mobility (access point range is likely to be a few 10th of meters) and why do we need 60 GHz. Didn’t we abandon that idea 15 years ago ?
Ref [1] Johan Hultell Jens Zander, Jan Markendahl, Competition and Cooperation in Wireless Multi-Access Networks, Book chapter in “Cognitive Wireless Networks “ , September 2007
Ref [2] Niebert N Schieder A Zander J Hancock R (ed.s), “Ambient Networks – cooperative mobile networking for the wireless world”, Wiley, April 2007
