[SOLVED] cdma link budget calculation

[hrushchov]

Hi there,

Need some help - it's about UMTS link budget power calculation and let's say we are using cdma with 512 codes.
How should one calculate the cdma coding gain on the downlink if, for example the BTS is using only one cdma code(from the whole 512) for the transmission? The gain is in regards to the signal power calculation, respectively the SNR.
It is clear that with more coding gain, less power is needed for the transmission, but what is the exact calculation or formula.

Thanks,

 

[tony_lth]

It need simulate. You need set Coding Source, Modulator, channel model, Noise model, phase noise model, demodulator, decoding, then you can get SNR(or BER/ ENR) vs. transmitter power or others.
It's not so easy .

 

[jupitorcuu]

There is an assumption before made such linkbudget calculation for SNR. The base band model will not take effect from Noise figure and IMD by analog component. Then the SNR you can used with signal quality Rho.
Rho=SNR/(SNR+1). We will take the system as linear system if for the linkbudget of whole communication system.
Code domain power as CDP=(1-rho)/#code number. This formula is used in CDMA2k, I think the spreadspectrum system are similar. I think it can solve your problem.

In another word, we want to know what is the signal auto-relation from sending signal to receiving signal.

 

[hrushchov]

Thanks for the replies.
I just wanted to know what is the achieved gain in power, when implementing cdma compared to the case where no such coding is used. That is purely theoretical.

 

[hrushchov]

Hello again,

So here it is what I have figured, correct me if I'm wrong:
The number of codes that we can use depends on the service(voice, video, etc.) we are going to use. Additionally, let us have the assumption that the codes are perfectly orthogonal the gain we are going to achieve is exactly the number of codes in dB.

Any comments oppose to that or willing to correct?
Thanks in advance!

 

[jiripolivka]

Hello again,

So here it is what I have figured, correct me if I'm wrong:
The number of codes that we can use depends on the service(voice, video, etc.) we are going to use. Additionally, let us have the assumption that the codes are perfectly orthogonal the gain we are going to achieve is exactly the number of codes in dB.

Any comments oppose to that or willing to correct?
Thanks in advance!

In ANY radio link, the power budget determines the signal-to-noise ratio in the receiver, before the detector, decoder or whatever non-linear processor is used. This "input" SNR affects the output SNR, or BER, in various ways; in CDMA and other modulation or coding systems , various responses are observed for BER as a function of input SNR.
There are theoretical, and empirical functions of BER/SNR, usually the real function makes ~10x worse the BER that theoretical. Coding and error correction steals a significant part of transmitted data out of the available capacity of a link. Designers have to decide about balancing the cost and complexity, often the only reliable way is using noise-adding link simulators; instead of varying signal level, noise level is easier to control.

 

[hrushchov]

@ jiripolivka,

So after your summery, could the CDMA be considered as a gain? Please consider that in the discussion we are talking about the downlink.

Regards,

 

[jiripolivka]

@ jiripolivka,

So after your summery, could the CDMA be considered as a gain? Please consider that in the discussion we are talking about the downlink.

Regards,

Coded Division Multiple Access certainly has an important gain in satellite downlink application. Coding offers a possibility to utilize better the limited transponder bandwidth by many users without mutual interference. Coding type and including forward error correction improves BER in the channel mainly at low input SNR.

As I wrote, there is no "straightforward" solution. The designer must find a multidimensional balance to obtain an optimal communication. Use a good textbook for details, like JMGomez: Satellite Broadcast Systems Engineering, Artech House 2002.

 

[hrushchov]

Will check the book, thanks for the reference.
The thing was to see under which assumptions is a simple calculation possible.

Thanks to all.