re: Ericsson Stalls on 3G FemtosInterference control between macro and femto layers is tricky but apparently some vendors know how it do it. Seems like Ericcson might be simply running late on this one...
re: Ericsson Stalls on 3G FemtosIt may well be true that Ericsson is careful on 3G femto cells because of its stake in 3G macro cells, but I do agree with them on the interference issue. Even though its "easier" to do network planning with CDMA or WCDMA (than with TDMA, or even OFDMA), it is not trivial. Self-installable base stations are unlikely to make it easier. Although this is unlikely to be a show-stopper, it's the kind of thing people tend to gloss over when talking about the wonders of femto cells.
The main issue remains: How much does a femto base station need to cost to make it an attractive business model?
This is one of the key issues, but it is one that people have put a lot of work into.
Not just vendors (ipAccess, Ubiquisys etc) but the carriers too have done a lot of modelling.
Quite alot of this has been presented in 3GPP and 3GPP2 and is accessible through their websites.
The Femto Forum is working on an 'open' model people can see and examine, but it isn't done yet.
Qualitatively, two arguments make sense: - One of the biggest problems with cdma systems is 'near-far. To the extent femtocells remove a lot of terminals from being 'far' (loud, lots of interference) and make them 'very, very near' and hence quiet that helps the whole network.
- The inbuilding attenuation that makes coverage bad alsop acts well as isolation.
Now, I fully agree that hand-waving isn't modelling and what will matter are the corner cases, but the analysis so far does look promising.
Of course, to make this work does require some sophisticated smarts and RRC in the femtocell - that is the secret sauce of the maufacturers.
The main issue remains: How much does a femto base station need to cost to make it an attractive business model?
This is one of the key issues, but it is one that people have put a lot of work into.
Not just vendors (ipAccess, Ubiquisys etc) but the carriers too have done a lot of modelling.
Quite alot of this has been presented in 3GPP and 3GPP2 and is accessible through their websites.
The Femto Forum is working on an 'open' model people can see and examine, but it isn't done yet.
Qualitatively, two arguments make sense:
- One of the biggest problems with cdma systems is 'near-far. To the extent femtocells remove a lot of terminals from being 'far' (loud, lots of interference) and make them 'very, very near' and hence quiet that helps the whole network.
- The inbuilding attenuation that makes coverage bad alsop acts well as isolation.
Now, I fully agree that hand-waving isn't modelling and what will matter are the corner cases, but the analysis so far does look promising.
Of course, to make this work does require some sophisticated smarts and RRC in the femtocell - that is the secret sauce of the maufacturers.