| Future Wireless Limits: A Quick Look Says 5-15 Hours/month of TV |
| Written by Dave Burstein |
| Thursday, 29 July 2010 18:32 |
 If LTE can deliver 5-12 megabit speeds, does it replace DSL many places? I've confirmed with good engineers the speeds are realistic, if and only if most people don't attempt to watch quality TV over the web. Because wireless networks are shared, they break under very high video volumes or other truly heavy bandwidth demands. That's almost guaranteed not to happen on major networks because the operators control the traffic volume by how they set prices. Something as simple as a cap somewhere between 25 and 75 gigabytes/month should be enough to protect a decent 2015-2020 LTE network. Low caps like AT&T's 2 gigabytes are abusive, but at a level far above that there is a problem.
Glen Campbell of Merrill Lynch calculated bandwidth cost, traffic growth estimates, necessary capex, and the effect different pricing strategies. In an important paper, he concluded there is no crisis to fear. http://bit.ly/bK5s4y. His paper was influential in the broadband plan, who independently came to a similar conclusion. Wireless speeds should be reliably at 5 megabits and above to a traffic load far higher than today but not unlimited in the short run.
Back of the envelope, a 2015 LTE network probably would allow most people to do all the surfing they wanted, social networking, and even dozens of hours of music listening without ruining the economics of a $30-50 service. The current monthly cost per gigabyte is about $2-3, which I'd expect to be under $1 when LTE is mature (2013-2016). 10 gigabytes, costing perhaps $10, would be enough to watch 5-15 hours/month of video.
A few could watch more. These are guidelines because number of towers, frequency, network load and usage patterns will vary widely. Adding an antenna outside each home can nearly double capacity.
These are reasonably well researched but very soft numbers. Don't use them except for very rough estimates.
I asked two dozen senior analysts how much cable/DSL they expected wireless to completely replace by around 2015. In countries like the U.S., with full telco and cable networks, the estimates came in at 5% to 25%. Most homes, I believe, will keep their landline broadband if the companies price appropriately. Countries without landlines will go almost entirely wireless and accept the speed/capacity limits. India has fewer than 4M phone lines for a billion people and wireless will soon dominate data as well as voice. Indonesia and nearly all of Africa have few landlines. Not to despair, DSL folks. No matter how well wireless does, in 2015 DSL and cable will be a $200B business with 500M subscribers. There will still be room for DSL Prime if you're still interested in reading it. |