RU2575303C2 - Method of continuous control of optical fiber disorders - Google Patents

Abstract

FIELD: instrument making.

SUBSTANCE: invention relates to the methods of optical fiber (OF) continuous control and may be used as an algorithm for the software within the protection system controller of FOTS for restricted information. Method of disordered optical fiber continuous control comprises recept, detection of optical signals from a fiber optic line, amplification, integration and analog-to-digital conversion of the received analog electric signals. Before control, define the number of observation readings N, current y_j and previous y_j-1 values of average sample values for N readings from formula: $$y_j=\frac{1}{N}{\displaystyle \sum {}_{i=1}^{N}yi}$$

where i - current value of ADC readings; compare current y_j and previous y_j-1 values of averages sample values while increasing the number of observation readings N to the condition under which $$y_j-y_{j-1}=1EMP,$$

where EMP is a least significant digit value of ADC; set the number of control readings n less than N, then at control at every instant k calculate the sum of squared deviations of readings y_i of L_k mean observation when the number of control readings n from formula: $$L_k={{\displaystyle \sum {}_{i=k-n}^k\left(yi-\frac{1}{N-n}{\displaystyle \sum {}_{i=k-N}^{k-n}}yi\right)}}^2,$$

carry out continuous control of disorders comparing L_k with a predetermined threshold value L_P, if L_k > L_P, shut down transmission of optical signals.

EFFECT: obtained technical result is correlation of the observation time with the start of disorder signal and optimal choice of observation time.

1 dwg

Description

The invention relates to methods for continuous monitoring of optical fibers (OB) and can be used as an algorithm for the software controller of the system for protecting FOSP information of limited access.

The well-known "Method of increasing the probability of detecting the output of radiation from an optical fiber" (RF Patent No. 2349039 dated November 27, 2006, published in BI No. 7 dated March 10, 2009), which is the closest to the claimed method and is therefore selected as a prototype .

The above method consists in receiving, detecting optical signals from a fiber optic line, amplifying, integrating and analog-to-digital converting the received analog electric signals, which are digitally compared with a control value in such a way that if the signal amplitude decreases below the control value by a predetermined value value an alarm is generated. After each number of samples N constituting the jth cycle specified by the timer, the average sample value is calculated

$$Y_j=\left(\mathrm{one}/N\right){\displaystyle \sum _{i=\mathrm{one}}^N{Y}_i,\left(\mathrm{one}\right)}$$

which is remembered and used in subsequent cycles. To calculate the control signal, use the formula

$$F_{J+k}\stackrel{N}{=}{\displaystyle \sum _{i=\mathrm{one}}\left(Y_i-Y_j\right)/Y_i-Y_j/.\left(2\right)}$$

After the end of each j-th cycle, the obtained value Ф _{j is} stored and subtracted from the value calculated in j-1, j-k1, j-k2 ... j-km cycles, the obtained difference is compared with the predefined detection thresholds Ф _o and alarm Ф _t , and if one of the conditions

$$\begin{array}{l}/F_j-F_{j-\mathrm{one}}/\ge F_t,\left(3\right)\\ /F_j-F_{j-k\mathrm{one}}/\ge F_t,\\ /F_j-F_{j-k2}/\ge F_t,\\ /F_j-F_{j-k3}/\ge F_t,\\ .....................\\ /F_j-F_{j-k3}/\ge F_t,\end{array}$$

an alarm is generated. If one of the conditions

$$\begin{array}{l}{F}_t\ge /F_j-F_{j-\mathrm{one}}/\ge F_o,\left(\mathrm{four}\right)\\ F_t\ge /F_j-F_{j-k\mathrm{one}}/\ge F_o,\\ F_t\ge /F_j-F_{j-k2}/\ge F_o,\\ F_t\ge /F_j-F_{j-k3}/\ge F_o,\\ ...........................\\ F_t\ge /F_j-F_{j-km}/\ge F_o,\end{array}$$

then the quantity calculated either in j-k1-1, or in j-k2-1, or in j-k3- ... or in j-km- depending on which from conditions (4) is satisfied. Otherwise, the value used in the j cycle is taken, and k1 / k2, kl2 / k ... km-1 / km≥Ф _t / Ф _o .

The disadvantages of the above method are:

1) lack of correlation of observation time and violation signal. The time of occurrence of the violation signal is not determined, and the beginning of the observation is formed arbitrarily by the timer;

2) non-optimal choice of observation time. The optimum should be considered the time during which the decision is not affected by external and internal interference.

The technical problem to be solved is the creation of a method for continuous monitoring of optical fiber with an increased probability of detecting a violation.

The technical result achieved is the correlation of the observation time with the beginning of the violation signal and the optimal choice of the observation time.

To achieve a technical result in a method for continuously monitoring optical fiber disturbances, which consists in receiving, detecting optical signals from a fiber optic line, amplifying, integrating, and analog-to-digital converting the received analog electrical signals, it is new that before monitoring the number of observation samples N is determined , current y _j and previous y _j-1 values of average sample values for the number of samples N from the relation:

$$y_j=\frac{\mathrm{one}}{N}{\displaystyle \sum {}_{i=\mathrm{one}}^{N}yi\left(5\right)}$$

where i is the current value of the ADC samples, compare the current y _j and previous y _j-1 values of the average sample values with an increase in the number of observation samples N to the condition under which

$$y_j-y_{j-\mathrm{one}}=\mathrm{one}emR,\left(6\right)$$

where EMP - ADC discharge unit Jr., setting the number of control samples n is smaller than N, then at the control at each time k computes the sum of squares of deviations samples yi by the average sample value L _k when the number of control samples n from the relation:

$$L_k={{\displaystyle \sum {}_{i=k-n}^k\left(yi-\frac{\mathrm{one}}{N-n}{\displaystyle \sum {}_{i=k-N}^{k-n}}yi\right)}}^2,\left(7\right)$$

carry out continuous monitoring of violations by comparing L _k with a predetermined threshold value of L _p , if L _k > L _p , the transmission of optical signals is disabled.

A new set of essential features makes it possible to increase the probability of detecting optical fiber disturbances due to the correlation of the observation time with the beginning of the violation signal and the optimal choice of observation time.

Figure 1 presents a diagram of the implementation of the control method.

The proposed control method is implemented as follows. Received optical signals are amplified, detected and integrated. As a result, the input optical signals in the form of a superposition of a constant signal, noise, and bias due to internal and external influences are fed to the input of the controller. The input analog signal, which is proportional to the transfer coefficient between the optical poles of the FOTS, is subjected to analog-to-digital conversion (ADC).

The ADC sampling period is t _d , which is determined by the upper frequency of the input signal in accordance with the Kotelnikov theorem:

$$t_d\le \mathrm{one}/2f,\left(8\right)$$

where f is the upper frequency in the spectrum of the input analog signal.

The number of sampling levels in amplitude K (ADC bit depth = log ₂ K) is determined by the ratio of the average input signal level U to noise

$$K~U/\sigma ,\left(9\right)$$

where σ is the mean square noise value.

After this, the time during which the signal can be considered stationary is determined. For this, the average sample value is calculated by the formula

$$yi=\frac{\mathrm{one}}{N}{\displaystyle \sum {}_{i=\mathrm{one}}^N}yi,\left(10\right)$$

where N = T _n / t _d is the number of ADC readings corresponding to T _n.

The value of N increases until the condition

$$y_j-y_{j-\mathrm{one}}=\mathrm{one}emR,\left(\mathrm{eleven}\right)$$

where emr is the unit of the least significant bit of the ADC. The obtained value of N will provide maximum sensitivity of the detection method.

Divide the amount of N in two parts (figure 1). The second part n is the number of samples by which the violation signal is detected and allocated. The first part (N-n) is the number of samples from which the average sample value is determined, from which subsequent deviations are measured (blue line in Fig. 1). It is advisable that the value of n be as large as possible, since the accuracy of the calculation depends on the number of samples.

After that, at each time moment k, the parameter L _{k is} calculated by the formula (green curve in Fig. 1):

$$L_k={\displaystyle \sum {}_{i=k-n}^k}{\left(yi-\frac{\mathrm{one}}{N-n}{\displaystyle \sum {}_{i=k-N}^{k-n}}yi\right)}^2.\left(12\right)$$

The obtained value is compared with the threshold value of L _p (Fig. 1 is a red line). If

$$L_k\ge L_P,\left(13\right)$$

It is believed that a fiber optic line violation signal has been detected.

Claims (1)

A method for continuously monitoring optical fiber disturbances, which consists in receiving, detecting optical signals from a fiber optic line, amplifying, integrating, and analog-to-digital converting the obtained analog electrical signals, characterized in that the number of observation samples N, the current y _j and the previous one are determined before monitoring y _j-1 values of average sample values for the number of samples N from the relation:
$$y_j=\frac{\mathrm{one}}{N}{\displaystyle \sum {}_{i=\mathrm{one}}^{N}yi}$$

where i is the current value of the ADC samples, compare the current y _j and previous y _j-1 values of the average sample values with an increase in the number of observation samples N to the condition under which
$$y_j-y_{j-\mathrm{one}}=\mathrm{one}emR,$$

where emr is the unit of the least significant bit of the ADC, the number of control samples n is less than N, and then during the control at each time moment k, the sum of the squared deviations of the samples yi from the average sample value L _{k is} calculated with the number of control samples n from the relation:
$$L_k={{\displaystyle \sum {}_{i=k-n}^k\left(yi-\frac{\mathrm{one}}{N-n}{\displaystyle \sum {}_{i=k-N}^{k-n}}yi\right)}}^2,$$

carry out continuous monitoring of violations by comparing L _k with a predetermined threshold value of L _p , if L _k > L _p , the transmission of optical signals is disabled.

Images