My code does not produce the correct curve.

Question

u-will-neva-no 2012-6-17

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/41349-my-code-does-not-produce-the-correct-curve

编辑： Sourav Mondal 2014-1-17

Hey everyone, here is a link which shows what formulas I have used, what parameters I have used, and the correct graph that I should be getting (the green line). Here is the direct link http://imageshack.us/g/17/formulaj.jpg/

Here is my code also:

%%Parameters
H = 6.63e-34;%Plancks constant [J.s]
Lambda = 1550; %wavelength of continuous wave laser (units: nm)
C = 3e5; %speed of light [nm/ps]
V = 3e8/1550e-9; %frequency of the propagating light [Hz]
Gamma = 1.2; %[1/W.km] Fiber non-linear paramter:  
D = 17; %[ps/(nm.km)]Chromatic dispersion 
N_s = 10; %number of spans (range of 1-200)
  alpha = 0.2; %Attenuation [dB/km] (convert dB/km <-> (1/km)) 
Alpha = alpha*4.343; %[1/km] 
L = 80;  %fiber length [km]
N_ch = 15;  %number of channels
Delta_f = 28.875e9; %[Hz]N-WDM channel spacing (QPSk = 28.875 GHz, QAM-8 = QAM-16 = 28.4375 GHz)  
  P_tx_db = -8:0.1:3; %[dBm]Launch optical power per channel (convert to W)
P_tx = 10.^( P_tx_db./10).*1e-3; %[W]
I_tx = P_tx./Delta_f;%[W/Hz] % optical launch power density per channel
%%EDFA parameters
G = L*Alpha; %Gain per amplifier 
F_db = 4.5; %Amplifier noise figure [dB] convert to W
F = 10^(F_db/10); % [W]Amplifier noise figure 
%%Sub Equations
L_eff = ((1-exp(-Alpha*L))/Alpha);%[km]non-linear effective length 
B_2 = -(D*(Lambda)^2/2*pi*C);%[ps^2/km]Group velocity dispersion parameter (related to chromatic dispersion)
B = N_ch*Delta_f; %Optical bandwidth [Hz]
I_ase = (G-1)*N_s*F*H*V; % [W/Hz]Amplified spontaneous emission; summation of the white optical amplified spontaneous emission
    SNR_wdm = I_tx_wdm.*exp(-(I_tx_wdm./I_01).^2)./(I_ase + I_tx_wdm.*(1 - exp(-(I_tx_wdm/I_01).^2))); %SNR for WDM
%%Poggiolini equation
%N-WDM equation [11]
I_NL1 = (2/3)^3 * N_s*(Gamma)^2*L_eff*log(pi^2*abs(B_2)*L_eff*(B)^2)*(I_tx).^3 / (pi*abs(B_2)) %[km.w^2/Hz^2]
SNR_wdm = I_tx.*exp(-I_NL1./I_tx)./(I_ase+ I_tx.*(1-exp(-I_NL1./I_tx))) 
figure(2); 
semilogy(P_tx ,SNR_wdm)%plotting optical power against SNR
xlabel('Optical Launch Power [W]','FontSize', 12, 'FontWeight', 'bold');
ylabel('SNR','FontSize', 12, 'FontWeight', 'bold'); 
hold on;

Please note that the sub equations were from a text book and should be correct. I think my error occurs when I convert from dB to W. I get confused when converting units however I have looked over my conversions and seems to be correct.

The conversions I used were: dB/km - 1/km 0.2 dB/km - 4.343*0.2

The other conversion was from dBm to W On the parameter image, it states that dBm should be a matrix (Ptx-launch power). So I used the relationship: dBm - W -8:0.1:3 - 10^(dBm/10)*1e^-3 (The 1e-3 was because the answer gives dBm - mW so I just converted to W).

now im sure ive messed up with the units but there could be other mistakes. To conclude, I am trying to plot the green graph exactly as shown in the attachment.