Dear everyone,
I need to plot BER vs SNR/OSNR of a fiber-optic communication system. Which one would be better while setting SNR or OSNR in the system:
Method 1:
Keep signal power constant (assume 0 dBm) and vary the Noise power to have different values of SNR/OSNR. and calculate the respective BER for different values of SNR/OSNR
Method 2:
Keep noise power constant (assume -80 dBm or -100 dBm) and vary the signal power to have different values of SNR/OSNR. and calculate the respective BER for different values of SNR/OSNR.
Which method will be appropriate ?
Regards,
Dhiman Kakati
The first Method is almost the standard one in performance analysis of communication systems. In this case, you could consider the practical range of power and noise in the performance analysis.
No, Method 2 is the best approach. You cannot affect the noise variance in a practical system. It is determined by the noise power spectral density, bandwidth, and receiver hardware. The received signal power can, however, be changed by varying the transmit power and/or the propagation loss.
In order to approach practical systems, I agree with Emil Björnson that method 2 is better.
The noise power is determined by its power power spectral density, there is a typical range for its value. Explicitly, the noise power can be alculated by P_N = F*k*T_0*B, where where F [dB] is the receiver’s noise figure, T_0 is the reference temperature in Kelvin at the receiver, k = 1.3 × 10−23 is Boltzmann’s constant and B [Hz] is the bandwidth.
The received SNR can be further determined given the transmit power and the pathloss as well as the small-scale fading.
Hi Zhang,
Did you referred to Wireless communication scenario. If we consider the Optical Fibre Communication (in place of fading, considering dispersion ) systems how these parameters will effect the system.
In optical communication system one can vary the transmitter power by controlling the current fed to the LED or the LASER diode. As the colleagues hinted, this is a practical way to adjust the required signal to noise ration.
On the other side, one can adjust the signal to noise ratio by varying the sensitivity of the receiver. However such variation needs that one selects an other detector since every detector has its specific sensitivity limited by its intrinsic noise generation. either thermal or shot noise. So, this approach can be used in the phase of selecting an appropriate detector.
However, for theoretical study you can use both scenarios to produce performance design curves.
Best wishes
Its very nice to learn from the high quality professors. With out even reading the comments of honorable Dr Emil, I was thinking about the second method. And the reason is fairly simple that the noise is practically can not be affected. We need to set or vary the transmission power to receive a particular SNR.
Does it really matter in case you have the freedom to apply both methods? If it matters, that is, choosing method A or B, what it then the practical relevance of SNR?
Dear Sir,
It is a simulation perspective discussed here. In Practical systems, no need to define it by the user.
Regards,
Dhiman Kakati
If you have a noise generation instrument, you can use methods 1 and 2, otherwise you can only use method 2.
Dear Dhiman Kakati ,
In the simulation analytic, mostly we can change noise power to see the response of the receiver and know what is the threshold of noise that keeping the system working.
Theory and as math equations, the two methods are active and no error in BER calculation.
Best wishes..
In theory, both methods.
In simulation, method 1.
In implementation and prototyping, method 2.
- Badges
- Science topic
- Similar topics
- Computer Communications (Networks)
- Simulators