Recieve Noise Floor Question

The RSSI for FIXED wimax says that measurable rx powers should be in the range of -40 to -123 dBm.

My question is can we acheive a sensitivity of -123 dBm for any combination of bandwidth and modulation scheme in wimax. BPSK 1/2 scheme requires a SNR of 3dB. Hence the noise floor of the receiver should be around -126dBm, if at all -123 dbM sensitivity can be obtained by BPSK using subchannelization.

The noise floor values using the standard -174+10log(BW)+NF is higher than -126 dBm for 1.25 MHz bandwidth.

Can someone please explain which modulation gives a -123 dBm sensitivity. And also how the noise floor should be better, ie < -123 dBm.

-----

Related Answers:

Actually, it is possible to go deeper in case your receiver can sub-divide a channel into smaller chunks, and thus reduce the BW component. However, you can't go infinitely with this concept, and you are limited with subchannelization to one single but whole subchannel, and this concept might make sense only in case you have multiple receivers (for pilots, etc.)
-123 dBm (with 3 dB SNR) stands for 63kHz bandwidth, and it is reduced somewhat in case of a realistic receiver NoiseFigure of ~4dB - down to 25kHz. It is just about enough to encompass a whole subchannel in a 3.5Mhz OFDM channel.
These multiple receivers are realised via serious DSP computing, and it is somewhat different than the original FFT concept of OFDM. What goes around - comes around.

WiMAX don't have BPSK scheme, I think QPSK 1/2 repetition 6 would be most robust MCS in WiMAX.
I don't know the meaning of "measurable". Does this mean "decoderble"?
And, I can't believe the sensitivity of any MS in WiMAX can meet -123dBm.
It's really perfect one.
I think -123dBm only consider AWGN noise and very very good SNR.
Normal situation in field does not show the 10dB SNR(example of good signal) and -123dBm RSSI.
-123 dBm means the edge of cell and SNR would be going to 0dB or negative SNR.
And, required SNR of data burst and preamble or control signals are different also.

Anyway most robust one is QPSK 1/2 repetition 6.

There are some things one must separate mentally when considering any technology, otherwise it all falls into apples-pears kind of confusion. Remember the 20 miles and 70Mbps tantrum? If all calculations are done correctly you'll always get the same result, regardless of the approach. The only necessary thing is to keep your apples together from beginning to the end, OR pick pears and stick with them to the end.
How it works? All communication can be expressed as energy per bit. Energy for all bits over full bandwidth running at some bitrate is total power of a system (apples). It can be seen as power density divided into sub-bands, and you get portions of power over smaller bandwidths with single subchannel as a unit (pears). Whatever approach you pick - end result is the same.
Point with OFDMA is that using small total power divided into small number of carriers you can reach further. To facilitate calculation it is customary to observe a single subchannel's behaviour (pears). Nothing else.

You should find exact meaning of -40~-123 dBm
I think -123 dBm is come from "dBm per minimum data burst unit(It's termed tone in WiMAX)
I don't know exactly what FFT size is used in 1.25MHz BW.
If 128 FFT is used in your system, -134 dBm is calculated from your formula.
And -40 dBm may be come from full BW usage.

[get this widget]