« 3G and 4G Wireless Is Private - DSL Is for Sharing | Main | The Web Server For Your Pocket Gets Released By Nokia Labs »



Hi Martin,

Interesting entry. It seems though that all these higher speeds are being pushed by manufacturers (because they want to sell new equipment). The biggest cost for an operator is the transmission. Operators still count their transmission in terms of E1 links which they usually lease and cost a LOT of money. How you are going to get 1000Mbps back to the core network? And who wants 1000Mbps worth of bandwidth anyway? Although the technology is quite impressive to read about I don't think it has much value in a real world environment.. Just my opinion..


Hello Mark,

you mention an interesting point. Even with HSDPA today, using 2 MBit/s E-1 copper links to the base station is quite a stretch as they are very expensive.

I recently talked to some guys who deal with backhaul transport. For WiMAX, LTE and also for HSDPA (in the forseeable future) there's a radical change in store. Instead of E-1 over copper, IP links (e.g. Metro Ethernet) will be used with much faster link speeds than current E-1's. This won't work over copper cables however, so the base station needs to either get a fiber link (expensive again if not already there...) or needs to be connected via mircowave (already done today but usually only to bridge E-1's).

Concerning the use of the bandwidth. Well, maybe a single person won't need 1 GB/s. However, a base station severs potentially hundreds of simultaneous users which brings you back to reality very quickly. Or, let me put it in the words of an optical network architect whom I met recently: "5 MBit/s is not broadband" :-)



It is only half way true that LTE is mainly pushed by operators. In fact operators have realized that end users prefer flat rates over complex tariff models and that the introduction of flat rates causes a tremendous increase in traffic (duh!). With current networks, however, they have to pay more for each additional bit they transfer. Not only in backhauling but also in capacity licenses at the RNC or several other network components.
For them every increase in traffic means additional costs. Not a very comfortable situation if you want or need to introduce flat rates to stay competitive. Thus, operators are in urgent need for a network technology that can bring more bits to more end users at less cost.
To find those networks they have even jointly founded a company - the NGMN Ltd. (http://www.ngmn.org) which searches for the best and most cost efficient next generation network. Manufacturers indeed propose LTE as the possible answer to what the NGMN Ltd is searching for as LTE significantly reduces the cost-per-bit.
It is doing so not only by an all-IP backhaul concept but also by a simplified flattened architecture, an increased spectrum efficiency and the fact that existing spectrum can flexibly be re-used causing only few changes in the existing antenna sites which are another cost factor next to backhaul.


The calculation isn't exactly right, with S-OFDMA, and probably increased subcarriers number, the spectral efficiency can be increased. so the base is no longer 2xnn, and 1Gbps is achievable, which is also demoed by Samsung and some other companies.


Hello Bruce, I don't quite understand yet how spectral efficiency is increased by using more bandwidth!? This just increases the number of subcarriers that can be used. This however does not increase spectral efficiency. So I am still missing something.


The comments to this entry are closed.

My Photo

The Books to this Blog

Secure Hotel Wi-Fi Sharing

My Pictures on Flickr

  • www.flickr.com
    martin.sauter's photos More of martin.sauter's photos


  • Clicky
    Clicky Web Analytics