Example and Results

 

Table: Typical Parameters for various components in CDMA star network

11

Amplifier Gain	G0	30 dB
Amplifier Saturation Power level	Psat	6 dBm
Spontaneous emission factor	nsp	2.0
Amplifier Coupling Loss	LA	3 dB
Optical Bandwidth	Bo	10 GHz
Operating Wavelength	$\lambda$	1550 nm
Quantum efficiency of photodetector	$\eta$	0.75
Receiver temperature	T	300o K
Receiver electrical bandwidth	Be	10 GHz
Load Resistance	RL	100 $\Omega$
Splice loss	Lsp	0.5 dB
Insertion loss of $2 \times 2$ coupler	Li	0.5 dB
Fiber length between user and star coupler	L	1 km
Fiber attenuation coefficient	$\alpha$	0.2 dB/km

 

An example network with the typical parameters is considered. The results have been computed for this network using the models described above. The typical values of the parameters are given in Table . Figure  shows the results for postamplifier. Curve A is for without SOA. Curve B corresponds to saturated SOA and curve C corresponds to unsaturated SOA. The two set corresponds to active encoding and passive encoding cases. The set having better performance corresponds to start with active encoding.

  

Figure: BER vs. P_s for postamplifier case. A- without amplifier, B- saturated SOA, C- unsaturated SOA

11

Since the peak output of CDMA decoder reduces due to effect of gain saturation in preamplifier-1 case, two new threshold values are used. The threshold-1 given by

$$I_{th}(1)=R_o { K P_r G\left ( KP_r \right ) \over 2 }$$ (5)

causes degradation after certain value of P_s(Fig. , curve B). The threshold-2 (Fig. , curve C) given by

$$I_{th}(2)=R_o { K P_r \over 2 } G \left ( { K P_r \over 2 } \right )$$ (6)

gives preformance as good as unsaturated preamplifier-1 case (Curve D, Fig. ). In the above P_r is power received for a pulse at the input of SOA from a single source and P_s is the power transmitted. Curve A in Fig.  is for CDMA star without any SOA. Fig.  shows the gain saturated preamplifier-1 performance for two different P_sat values. Curve A and B corresponds to threshold-1 with P_sat=6 and $10 \; dBm$respectively. Curve C and D are for threshold-2 with P_sat=6 and $10 \; dBm$ respectively. It shows that use of threshold-2 gives more improvement for lower values of P_sat.

  

Figure: A) CDMA star without SOAs; preamplifier-1: B and C) gain saturation with threshold-1 and threshold-2, D) without gain saturation

11

  

Figure: BER v/s P_s for gain saturated Preamplifier-1; A, C for P_sat=6 dBm and B, D for 10 dBm for threshold-1 and threshold-2 resp.

11

  

Figure: BER v/s P_s for preamplifier-2 with gain saturation

11

  

Figure: BER v/s P_S for preamplifier-1, preamplifier-2 and preamplifier-1 with zero ASE noise. No gain saturation is considered.

11

Fig.  gives the results for preamplifier-2 case. Curve A is for unsaturated case. Curve B and C are for gain saturation with uniform and non-uniform multiple access intereference (MAI) respectively. The figure shows that gain saturation improves the performance slightly. The actual peformance will be in between two extremes given by curve B and C.

The results in Fig.  show that preamplifier-1 (amplifier placed after the CDMA decoder) performs poorly as compared to preamplfier-2 (amplifier placed before the CDMA decoder). In computing this, the gain saturation has not been considered. The preamplifier-2 performance is almost as good as that of preamplifier-1 without ASE noise. This implies preamplifier-2 configuration has very small degrading effect of ASE noise. One can also deduce from Fig.  and Fig.  that gain saturation will affect negligibly in the preamplifier-2 case.