Eric, about (another) doubling capacity.
Double receivers in handset
Using transmit diversity needs two more output(input) symbol streams after(before) the symbol demux(mux). That is where the so-called double receivers come from. Bascially you double up the repetition blocks, spreading blocks quadrature spreading abd baseband filtering. More ASIC techniqes, with ZIF you reduced the IF blocks and save the space for the receivers.
Base stations - 2 v 4
Same thing for the handsets, without TD, you have 2 receivers for I and Q channel. Now you double it up.
Now let's talk about the so-called double of voice capacity in GSM.
From EFR to AMR, the principle is the same, that is to improve the vocoder to reduce the data rate for voice. Just like the EVRC in CDMA. However there is a fundemental difference between AMR and EVRC, AMR, although multi-rate, is actually a fix rate on different level, it is not based on the speech activity but based on the link condition and is controlled by the networks via signalling msgs. EVRC, on the other hand, is fully utilizing the CDMA environment, that is the link condition is about the same for each user thanks to the power control in CDMA. So with this in hand, EVRC is the real variable rate vocoder that can take more advantage of the speech activity pattern while maintaining a lower coding rate. In AMR, the lower rate is not available to everybody. Therefore the not everybody can use the half rate frames.
So now we are done with the first comparable part. Here comes the second part, in CDMA2000, due to the better power control, the better modulation (coherent detection), you are close to double the voice capacity. On the GSM side, there's nothing to further increase the voice capacity. The main focus is on the data part, that is to increase the data speed by using different modulation scheme. But that won't increase the number of voice users because you are still using the same kind of timeslots and same kind of the multiframes. The increase in 1x is the real increase of resource where GPRS or EDGE is just a better use of existing resource for data.
Now as we see CDMA just did it again, on top of that you got another 70 - 100 % of capacity boost. This time it is using not just the transmit diversity, but also another vocoding technique, SMV, selective mode vocoder. Will the same principle apply to GSM side, maybe, maybe not, Unless you go out and make major changes in the GSM air interface. In WCDMA, there is STTD similar to the STS in CDMA2000. But I don't know transmit diversity is in the agenda in GSM/GPRS/EDGE. And SMV beats AMR hands down
So CDMA went through 3 steps of boost where GSM really had just about one. Even with that one, the increase of capacity is actually not the top reason, improving the voice quality might be more of a factor. There will be certain degree of capacity improvement but to not by much unless you don't care about the audio quality in some emergency situations. |
