4G is pure IP everywhere transported by a pure packet network. WiMAX (Worldwide Interoperability for Microwave Access) just had its WiMax World Conference and is certainly a 4G network with its mobile standard (IEEE 802.16e). For another view, Andrew Seybold says WiMax is not 4G. A Qualcomm ally, he is buying time while Qualcomm’s Jeff Belk makes up his mind what the new topology will be. Agreed WiMax data rate is too low for a next gen standard, has flaws in topology msut battle the incumbent technology. So will Intel’s idea of 4G sell?
- Advanced 3G mobile Internet standards (EVDOa, HSUPA) are already “good enough” data rates compared to the WiMax standard.
- WiMax has a price while WiFi 802.11g is “free”. The upcoming 802.11n “free” standard will offer range and power close to WiMax.
- WiFi is subcellular, WiMAX is a broadcast topology. Cells outrule towers. The problem with towers is that when they serve too much area they becomes congested. Mobile cellular basestations work because they recover spectrum in small “cells”. WiFi has a secret success factor over WiMax. Low power short range WiFi low can create a large number of very small cells in and area and efficiently reduce congestion.WiMax networks do have a clean start. There is no “overlay” on circuit technology which is still in place in 3G transmissions. The key WiMax challenge is to get mobile voice and data to work well together. The issue is that handover of large populations of devices in motion is a serious issue. Hopefully WiMax will do what cellular operators have not – do a better job in call reliability. There is a well known, but embarrassing to see, cell phone commercial. An actor repeats a mantra that many of us say now – “Can you hear me now”. Think about that, he is making sure that their is a cell tower working in the are, that a call has not yet dropped or that the operator has not dialed down your band frequency because of congestion, so you sound like a box of silverware clattering. I appreciate any contender that can clear up communication.
The first WiMax users now appear from the telecommunications time machine. Korean and Japan (always 1-2 years ahead of the US) have something to say now. Korean deployed WiBro, a superset of WiMax, and report that high speed HSDPA (3.6-7.6 Mbps) is “good enough”. They say there was very little reason for subscribers to move to 5-15 Mbps WiMax). They observe that a WiMax handset is silly, but that a WiMax fixed box might work if only because it is transferring data in fresh uncontested spectrum.
So beyond pure IP, what notes will distinguish 4G ? The term ‘Personal Broadband’ is the banner cry. Add in – always on anywhere. But back to reality. 4G will need significant speed difference, topology, and true quality of service. My guess is that 60Mbps down and 30Mpbs up should be a target speed for mobile networks capable of handling all digital IP media. Topology is the rich question. Can a communications designed for bits truly traverse from WAN down to WiFi, Bluetooth, Zigbee and back in meaningful ways? Another topology challenge is to provide multicast (one way realtime bit cast) to millions and still be able to do hi-def peer to peer. Finally an “assured” high quality service protocol is necessary for voice or video communication, steady and stable, realtime verifiable “message streams”. We don’t want a world where you have to wave your arms and say “can you see me now?”