Controlling device of optical amplifier and controlling method

The present invention relates to a control device and a control method for an optical amplifier used in an optical communication system such as a wavelength division multiplexing optical transmission system.

In optical communication systems, as is well known, long-distance transmission is realized by providing optical amplifiers at predetermined intervals. For example, dozens to hundreds of optical amplifiers are installed on an optical transmission line that crosses the Pacific Ocean. Some optical amplifiers use an optical fiber to which rare earth ions such as erbium (Er), placeodium (Pr), and thulium (Tm) are added as an amplification medium. An erbium ion-doped optical fiber (EDF: Erbium Doped Fiber), which provides a wide band, is often used. On the other hand, with the spread of the Internet in recent years, the amount of information transmitted via a network has been rapidly increasing, and further increase in capacity of a transmission system is required. A wavelength division multiplexing (WDM) transmission method has already been put into practical use as one of the technologies for increasing the capacity of a transmission system. In this WDM transmission system, a plurality of signal lights (channels) having different wavelengths are multiplexed and transmitted via one optical transmission line (optical fiber). Therefore, the optical amplifier used in the WDM transmission system is required to be able to amplify a plurality of signal lights at once. In recent years, an OADM (Optical Add-Drop Multiplex) device that can individually increase or decrease each signal light (channel) has also been put into practical use in order to realize a more flexible network configuration, and is transmitted through an optical transmission line. The signal light power is not constant. In this way, if the signal light power fluctuates greatly due to a change in the number of channels (this change naturally occurs even when a failure occurs in some channels), the signal light of other channels cannot be received and demodulated normally. It ends up. Therefore, in the conventional optical communication system, even if the number of signal wavelengths of the WDM signal changes, the output power of each signal wavelength can be controlled to be constant by controlling the amplification gain to be constant in the optical amplifier. It is done. Such a control method is called AGC (Automatic Gain Control). Specifically, the output optical power of the optical amplifier is monitored, and the amplification gain (actually, for example, excitation optical power) of the optical amplifier is controlled so that the output optical power of the optical amplifier becomes constant based on the monitor value. Is done. However, in such a conventional control method (AGC), it takes about several tens of ms from the change of the input optical power to the appropriate control of its influence, so that there is a restriction on increasing or decreasing the signal optical wavelength. It is happening. In addition, if a wire break occurs due to an accident, the signal level of another wavelength will be affected for a long time. Therefore, it is desired that the AGC of the optical amplifier can constantly control the output optical power in a short time by following the change of the input optical power at high speed. However, if the response speed of AGC is increased indefinitely, an oscillation phenomenon may occur. Therefore, as a method for increasing the speed of AGC, for example, Japanese Patent Application Laid-Open No. 9-200145 (hereinafter referred to as Patent Document 1) and Japanese Patent Application Laid-Open No. 7-221737 (hereinafter referred to as Patent Document 2) have been proposed. There is such a technology. That is, first, in the technique described in Patent Document 1, an optical signal to be amplified is input to an optical amplifier (EDF) via an optical fiber having a predetermined delay time, and signal light is input to the input end of the optical fiber. The supply of the excitation light is started between the time when the signal light reaches the EDF and the time when the signal light reaches the EDF (see, for example, paragraphs [0034] to [0039] of Patent Document 1). As a result, it is possible to allow a margin for the time required for the excitation light control by the above-mentioned delay time, and as a result, it is possible to improve the response speed of the AGC while suppressing the occurrence of the optical surge. On the other hand, in the technique described in Patent Document 2, the adjusting light for canceling the change of the input signal light is input to the optical amplifier from the direction opposite to the transmission direction of the signal light, and the optical amplifier responds to the adjusting light. By controlling the excitation light, the output of the optical amplifier is constantly controlled. According to this method, the range of the input optical power of the optical amplifier to be controlled can be narrowed, so that the response speed of the AGC can be easily improved (for example, paragraph [0016] of Patent Document 2]. ~ [0027]). However, in the technique described in Patent Document 1, since an extra optical fiber is provided for delaying the signal light input to the optical amplifier, the characteristics of the optical fiber are deteriorated and the size of the optical amplifier is increased. Become. Further, the technique described in Patent Document 2 requires high-speed control of the regulated light, and also results in an increase in power consumption and heat generation of the optical amplifier due to the regulated light output. The present invention was devised in view of the above problems, and follows the fluctuation of the input power of the signal light at high speed without causing an oscillation phenomenon, an increase in the size of the optical amplifier, an increase in power consumption and heat generation. It is an object of the present invention to provide a control device and a control method of an optical amplifier which can be used.

In order to achieve the above object, the control device and control method of the optical amplifier of the present invention are characterized by having the following means. (1) First control means for controlling the gain of the optical amplifier based on the input optical power and the output optical power of the optical amplifier (2) Depending on at least one of the input optical power and the output optical power. Gain control amount variable means for changing the gain control amount of the optical amplifier by the first control means Here, the first control means is the optical amplifier based on the input optical power and the output optical power. It is equipped with a gain control amount calculator that obtains the difference from the target gain of the above and obtains the gain control amount based on the difference, and the above-mentioned gain control amount variable means sets the coefficient for the difference as the input optical power and the output. It may be configured with a coefficient controller that changes according to at least one of the optical powers. Further, the first control means described above obtains a difference from the target gain of the optical amplifier based on the input optical power and the output optical power, and the gain is based on the difference and the integrated value of the difference. A gain control amount calculator for obtaining a control amount is provided, and the above-mentioned gain control amount variable means is a coefficient controller that changes a coefficient for the difference according to at least one of the input optical power and the output optical power. It may be configured in preparation. Further, the first control means described above obtains a difference from the target gain of the optical amplifier based on the input optical power and the output optical power, and the gain is based on the difference and the integrated value of the difference. A gain control amount calculator for obtaining a control amount is provided, and the above-mentioned gain control amount variable means sets a coefficient for the difference and a coefficient for the integrated value to at least one of the input optical power and the output optical power, respectively. A coefficient controller that changes according to the situation may be provided. Further, the control device of the optical amplifier further includes a second control means for feedforward controlling the gain of the optical amplifier based on the input optical power, and the first control means and the second control means described above. It may be configured to control the gain of the optical amplifier by the combination of. Further, when the optical amplifier has a first excitation light source and a second excitation light source, the first control means does not obtain the output light power expected from the first excitation light source. Occasionally, the deficiency determination unit for obtaining the gain control amount for compensating for the deficiency with the second excitation light source, and the first excitation light source for the gain control amount obtained by the deficiency determination unit. It may be provided with a conversion unit that performs conversion so that the coefficient for the difference does not change depending on whether the expected output light power is sufficient or not. Further, the control device for the optical amplifier of the present invention is a control device for an optical amplifier having a first excitation light source and a second excitation light source, and is based on the input optical power and the output optical power of the optical amplifier. A control means for obtaining a difference from the target gain of the optical amplifier and controlling the gain of the optical amplifier based on the difference is provided, and the control means obtains the excitation light power expected from the first excitation light source. When not present, the first excitation is performed with respect to the deficiency determination unit for obtaining the gain control amount for compensating for the deficiency with the second excitation light source and the gain control amount obtained by the deficiency determination unit. It is characterized by having a conversion unit that performs conversion so that the coefficient for the difference does not change depending on whether the excitation light power expected from the light source is sufficient or not. Further, in the control method of the optical amplifier of the present invention, (1) the gain control amount when controlling the gain of the optical amplifier based on the input optical power and the output optical power of the optical amplifier is obtained, and (2) the gain is obtained. The gain control amount is changed according to at least one of the input optical power and the output optical power. Further, the control method of the optical amplifier of the present invention is a control method of an optical amplifier having a first excitation light source and a second excitation light source, and is based on the input optical power and the output optical power of the optical amplifier. In obtaining the difference from the target gain of the optical amplifier and controlling the gain of the optical amplifier based on the difference, (1) when the excitation light power expected from the first excitation light source cannot be obtained, the said The gain control amount for compensating for the shortage with the second excitation light source is obtained, and (2) the excitation light power expected from the first excitation light source is sufficient or insufficient for the gain control amount. It is characterized in that conversion is performed so that the coefficient for the difference does not change from time to time.

FIG. 1 is a block diagram showing a configuration of a main part of an optical amplifier according to a first embodiment of the present invention. FIG. 2 is a block diagram showing the configuration of the control unit shown in FIG. FIG. 3 is a diagram showing the relationship between the required feedback coefficient for the input optical power and the convergence limit according to the present embodiment. FIG. 4 is a block diagram showing the configurations of the feedback control unit and the feedback coefficient control unit shown in FIG. FIG. 5 is a block diagram showing a first modification of the feedback control unit and the feedback coefficient control unit shown in FIG. FIG. 6 is a block diagram showing a second modification of the feedback control unit and the feedback coefficient control unit shown in FIG. FIG. 7 is a block diagram showing a configuration of a control unit of an optical amplifier according to a second embodiment of the present invention. FIG. 8 is a block diagram showing the configurations of the feedback control unit, the feedback coefficient control unit, and the feedforward control unit shown in FIG. 7. FIG. 9 is a block diagram showing a first modification example of the feedback control unit, the feedback coefficient control unit, and the feedforward control unit shown in FIG. FIG. 10 is a block diagram showing a second modification of the feedback control unit, the feedback coefficient control unit, and the feedforward control unit shown in FIG. FIG. 11 is a block diagram showing a configuration of a main part of the optical amplifier according to the third embodiment of the present invention. FIG. 12 is a block diagram showing the configuration of the control unit shown in FIG. FIG. 13 is a block diagram showing a modified example of the control unit shown in FIG. FIG. 14 is a diagram showing the relationship between the excitation light control value and the excitation light power in order to explain the excitation light power control (without correction). FIG. 15 is a diagram showing the relationship between the excitation light control value and the excitation light power in order to explain the excitation light power control (with correction) by the control unit shown in FIG.

[A] Explanation of First Embodiment FIG. 1 is a block diagram showing a configuration of a main part of an optical amplifier according to a first embodiment of the present invention. The optical amplifier 1 shown in FIG. 1 is an optical demultiplexer 2, It is composed of 3,6, an erbium-added optical fiber (EDF) 4, a gain equalizer 5, an optical sensor 7,9, an excitation light source 8, and a control unit (control device) 10. Here, the optical demultiplexer 2 branches a part of the WDM signal light (main signal light) received from the optical transmission line 20 and outputs one to the EDF4 side and the other as a monitor light for a power monitor. The optical demultiplexer 3 is for combining the excitation light (pump light) supplied from the excitation light source 8 with the main signal light from the optical demultiplexer 2. .. Further, the EDF 4 amplifies the main signal from the optical combiner 3 by the above-mentioned excitation light, and the gain equalizer 5 equalizes the gain (gain) of the amplified output of the EDF 4 to the gain of each wavelength of the main signal light. For flattening (flattening), the optical duplexer 6 branches a part of the equalized output of the gain equalizer 5, outputs one to the optical transmission line 30 on the output side, and outputs the other to the optical transmission line 30 on the output side. It is output to the optical sensor 9 as monitor light for a power monitor. On the other hand, the optical sensor 7 receives the monitor light branched by the photodetector 2 and generates an electric signal according to the received light amount (power), and the optical sensor 9 similarly receives the light. It receives the monitor light branched by the demultiplexer 6 and generates an electric signal according to the received light amount (power), and all of them are composed of, for example, a photodiode (PD). That is, the above-mentioned optical sensor 7 functions as an input light measuring means for measuring the input light power of the EDF 4 functioning as an optical amplifier, and the optical sensor 9 functions as an output light measuring means for measuring the output light power of the EDF 4. .. Further, the excitation light source 8 is for generating excitation light for EDF4, and is composed of, for example, a laser diode (LD). Then, the control unit 10 controls the excitation light power of the excitation light source 8 based on the power measurement results (input light power and output light power) by the above optical sensors 7 and 9, and constantly controls the gain of the EDF 4 ( AGC: Automatic Gain Although it is for controlling), in the present embodiment, a device for improving the response speed of AGC control has been devised so as to sufficiently follow a sudden fluctuation of the input optical power. Specifically, the control unit 10 of the present embodiment includes a feedback control unit 11 and a feedback coefficient control unit 12, as shown in FIG. 2, for example. Here, the feedback control unit (first control means) 11 feedback-controls the gain (that is, the excitation light power) of the EDF 4 based on the input light power and the output light power monitored by the optical sensors 7 and 9, respectively. The control amount is calculated. Specifically, for example, the control amount is calculated by the following formula (1). LD_{out out}= α × (P_in× GP_{out out}) ... (1) LD_{out out}: Excitation light power P_in: Input light power P_{out out}: Output light power G: Target gain Therefore, the feedback control unit 11 uses the input light power P as shown in FIG. 4, for example._inMultiplier 11-3 that multiplies and the target gain G, and the output optical power P from the multiplication result_{out out}It is composed of a subtractor 11-2 for obtaining the difference by subtracting, and a multiplier 11-3 for obtaining the excitation light control value by multiplying the obtained difference by the feedback coefficient a. That is, these multipliers 11-1, subtractors 11-2 and multipliers 11-3 differ from the target gain G of the optical amplifier 1 (input) based on the input optical power and the output optical power of the optical amplifier 1. Optical power P_in× Target gain G-Output optical power P_{out out}), And the gain control amount is calculated based on the difference. It functions as a gain control amount calculator. Here, in order to perform feedback control with a quick reaction, the coefficient (feedback coefficient) a in the above equation (1) may be increased, but if this coefficient a is increased indefinitely, the output of EDF4 oscillates. It will not converge. In particular, EDF4, which is often used as an optical amplifier, tends to oscillate due to its characteristics, and the feedback coefficient a cannot be set large enough to sufficiently follow sudden fluctuations in input optical power. However, by considering the transient phenomenon of EDF4 with respect to the change in excitation light power and setting the feedback coefficient a according to the input / output light power, faster control can be realized without oscillating. For example, in EDF4, the relationship between the input optical power, the coefficient (convergence limit) a that causes the oscillation state, and the coefficient a that can secure sufficient performance (high-speed response performance) is shown in FIG. As shown in FIG. 3, it can be seen that it has a characteristic that it is difficult to oscillate where the coefficient a needs to be increased. Therefore, in the present embodiment, the feedback coefficient a is monitored by the optical sensor 7 by the feedback coefficient control unit (gain control amount variable means) 12, and the input optical power P is monitored._inControl (variable) according to. Specifically, as shown in FIG. 4, the function f of the required feedback coefficient a shown in FIG. 3₁It is controlled by (function arithmetic unit 12-1). Here, the function f of the required feedback coefficient a shown in FIG.₁Can be expressed, for example, by the following equation (2). f₁= b × Convergence limit ... (2) In this equation (2), "b" is a constant between 0 and 1, which is determined by how fast response performance is required. Further, the "convergence limit" can be expressed by the following equation (3), for example, when the EDF4 has a gain of about 20 dB. Convergence limit = c / P_in ... (3) Therefore, the function f₁Is f₁= bc / P_in ... can be expressed as (4). However, "P" in the above equations (3) and (4)_in"" Represents the input optical power, and "c" represents the constant determined by the EDF characteristics and gain. When the input optical power is small, if this equation (4) is used as it is, the feedback coefficient a becomes large, the response is too fast, and the arithmetic circuit becomes complicated. Therefore, the feedback coefficient a is constant where the input optical power is small. Or change the value of "b". Further, when the above control units 11 and 12 are implemented in a small logic circuit or the like, the above function f₁It becomes difficult to realize the division of, but for example, the function f₁Can be realized by implementing as table format data with the required memory such as RAM. With the above configuration, in the optical amplifier 1 of the present embodiment, the input optical power P of the optical amplifier 1 (EDF4)_inAnd output light power P_{out out}Are measured by the optical sensors 7 and 9, respectively, and the measured input optical power P is measured by the feedback control unit 11._inAnd output light power P_inDifference from target gain G based on (input optical power P_in× Target gain G-Output optical power P_{out out}) Is obtained by the subtractor 11-2, and the feedback coefficient a is multiplied by the multiplier 11-3 to obtain the excitation light control value. At this time, the feedback coefficient a is the feedback coefficient control unit 12 (function f).₁) Input light power P_inIt is variable according to the above, and becomes a necessary and sufficient value below the "convergence limit", and the AGC of EDF4 can be speeded up without causing an oscillation phenomenon. In particular, in this example, since the feedforward control by the feedforward control unit 13 is also performed, a higher speed AGC can be realized. Therefore, it is possible to follow the fluctuation of the input power of the signal light to the optical amplifier 1 used in the WDM transmission system at high speed without causing an oscillation phenomenon, an increase in the size of the optical amplifier, an increase in power consumption and heat generation. , The output power fluctuation of the optical amplifier 1 can be suppressed. Specifically, when the number of wavelengths (channels) used as signal light fluctuates, the output power fluctuation of each signal light can be reduced. As a result, WDM optical communication that is more stable than before can be realized. (A1) Explanation of the first modification FIG. 5 is a block diagram showing a first modification of the control unit 10 (feedback control unit 11 and feedback coefficient control unit 13) described above. In FIG. 5, the feedback control unit 11 is compared with the configuration shown in FIG. However, in addition to the multipliers 11-1, 11-3 and subtractor 11-2 described above, the integrator 11-4, the multiplier 11-5 and the adder 11-6 are further provided. The other components (those having the same reference numerals as those described above) are the same as or similar to those described above unless otherwise specified, and the same applies to the following modifications. Here, the adder 11-4 integrates the difference obtained by the subtractor 11-2, and the multiplier 11-5 is the integrated value for the above difference obtained by the adder 11-4. Is multiplied by a predetermined coefficient, and the adder 11-6 adds the multiplication result of the multiplier 15 and the multiplication result of the multiplier 11-3 to obtain the excitation light control value. .. That is, each component 11-1 to 11-6 of the feedback control unit 11 of this example is the input optical power P of the optical amplifier 1._inAnd output light power P_{out out}Difference from target gain G based on (input optical power P_in× Target gain G-Output optical power P_{out out}), And it functions as a gain control amount calculator that obtains the gain control amount of the optical amplifier 1 based on the difference and the integrated value of the difference. Even with such a configuration, the same effects as those of the above-described embodiment can be obtained, and a more stable AGC can be realized as compared with the above-mentioned embodiment. (A2) Explanation of the Second Deformation Example FIG. 6 is a block diagram showing a second modification of the control unit 10 (feedback control unit 11 and feedback coefficient control unit 13) described above, and is shown in FIG. 5 in FIG. Compared to the configuration, the feedback coefficient control unit 12 uses the function f described above.₁Function f in addition to (function arithmetic unit 12-1)₂(Function arithmetic unit 12-2) and its function f₂The difference is that the output of is multiplied by the output of the integrator 11-4 by the multiplier 11-5 of the feedback control unit 11. Where the above function f₂Is the function f mentioned above₁It is a function that only differs in the value of "c" in (depending on EDF characteristics, gain, amplifier configuration, etc.). That is, in this example, in the feedback control unit 11, not only the difference obtained by the subtractor 11-2 but also the coefficient to be multiplied by the integrated value of the difference obtained by the integrator 11-4 is also a function f.₂By input light power P_inIt is designed to be variable according to. Therefore, in this case as well, the same effects as those of the above-described embodiment and the first modification can be obtained, and a faster and more stable AGC can be realized as compared with the one shown in FIG. [B] Description of the Second Embodiment FIG. 7 is a block diagram showing a configuration of a control unit of an optical amplifier according to a second embodiment of the present invention, and the control unit 10 shown in FIG. 7 is the one shown in FIG. In comparison, the feedforward control unit 13 and the adder 14 are further provided. Here, the feedforward control unit (second control means) 13 is the input optical power P monitored by the optical sensor 7._inThe control amount (excitation light control value) at the time of feed-forward control of the excitation light power is obtained according to the above, and the adder 14 adds the excitation light control values obtained by the control units 11 and 13 respectively. It is supplied to the excitation light source 8 as an excitation light control signal. Specifically, in this case, for example, as shown in FIG. 8, the feedback control unit 11 is configured to include the same multipliers 11-1, 11-3 and subtractor 11-2 as those described above, and provides feedback. The coefficient control unit 12 has the same function f as described above.₁It is configured with (function arithmetic unit 12-1), and the feedforward control unit 13 further has a feedforward function f.₃It is configured with (function arithmetic unit 13-1). The feedforward function f₃May be a known function used for feedforward control. That is, the control unit 10 of the present embodiment is configured to control the gain of the optical amplifier 1 by the combination of the feedback control unit 11 and the feedforward control unit 13. This makes it possible to further increase the speed of the AGC of the optical amplifier 1 as compared with the first embodiment without causing an oscillation phenomenon. Even when the feedforward control unit 13 is used in this way, the feedback control unit 11 is replaced with the subtractor 11-, as shown in FIGS. 9 and 10, respectively, in the same manner as in the configuration described above with reference to FIGS. The integrated value of the difference obtained by 2 may be used for feedback control, or the coefficient to be multiplied by the integrated value in such a configuration is also a function f.₂It may be configured to be variable by. In either case, the AGC of the optical amplifier 1 can be performed more stably and at high speed. [C] Description of Third Embodiment FIG. 11 is a block diagram showing a configuration of a main part of an optical amplifier according to a third embodiment of the present invention, and the optical amplifier 1 shown in FIG. 11 is the one shown in FIG. In comparison, a plurality of (here, two) excitation light sources 8-1 and 8-2 are provided, and an optical amplifier for inputting the excitation light generated by the excitation light source 8-2 to the EDF4 from behind the EDF4. The difference is that the wave device 3'is provided between the EDF 4 and the gain equalizer 5. The control unit 10 in this case also has a feedback control unit 11 and a feedback coefficient control unit 12 similar to those shown in FIG. 2. In this case, the feedback control unit 11 is shown in FIG. 12, for example. Like the function f₄(Current value calculator 11A-7), function f₅(Current value calculator 11B-7), function fc (conversion calculator 11B-8), limiter 11A-8, 11B-9 are provided. Multiplier 11-1, subtractor 11-2, multiplier 11-3, function f₁Are the same as or similar to those described above. Where the above function f₄(Current value calculator 11A-7) obtains the current value for driving one (first) excitation light source 8-1 from the excitation light control value obtained by the multiplier 11-3 as described above. And the limiter 11A-8 is this function f₄When the current value obtained in (1) exceeds the allowable range of the excitation light source 8, the current value is kept within the allowable range (below the maximum value). However, in this embodiment, the limiter 11A-8 is a function f.₄When the current value obtained in (1) exceeds the permissible range, the current value obtained by subtracting the above maximum value from the current value that should be output corresponds to the shortage of the excitation light power by one of the excitation light sources 8-1. It is supplied to the function fc (conversion calculator 11B-8) as a current value (hereinafter referred to as insufficient current value). In other words, this limiter 11A-8 is a shortage that obtains a gain control amount to make up for the shortage with another excitation light source 8-2 when the expected excitation light power cannot be obtained from the excitation light source 8-1. It functions as a decision-making unit. Then the above function f₅(Current value calculator 11B-7) is a function f₄Similarly, the current value for driving the other excitation light source 8-2 is obtained from the excitation light control value, and the function fc [conversion calculator (conversion unit) 11B-8] is this function. f₄With the current value obtained in the above and the above-mentioned insufficient current value supplied from the limiter 11-8, the shortage of the excitation light power by one excitation light source 8-1 is supplemented by the excitation light power of the other excitation light source 8-2. This is to correct (convert) the insufficient current value so that the feedback coefficients a for both excitation light sources 8-1 and 8-2 do not change. Specifically, the output (current value) I of this function fc (conversion calculator 11B-8)₃Is the function f₅Output of I₁, I set the shortage current value₂Then, for example, it is expressed by the following equation (5). fc = I₃= I₁+ d × I₂ ... (5) This equation (5) is a function f if there is no undercurrent value.₅The current value obtained in is the drive current value of the excitation light source 8-2 as it is, and in other cases, the function f₅It means that the driving current value of the excitation light source 8-2 is obtained by adding the insufficient current value corrected so that the feedback coefficient a does not change to the current value obtained in. Note that "d" in this equation (5) is selected so that the feedback coefficient a is the same for "state 1" and "state 2" shown in FIGS. 14 and 15 (it can be easily obtained by actual measurement). it can). In this way, the oscillation limit can be made the same in "state 1" and "state 2". In FIG. 14, the maximum outputs of the excitation light source 8-1 and the excitation light source 8-2 are the same, and the output powers (excitation light power) of the excitation light source 8-1 and the excitation light source 8-2 are 2: 1. The relationship between the excitation light control value (driving current value) and the excitation light powers of the excitation light sources 8-1 and 8-2 when the above function fc is not corrected is shown. , FIG. 15 shows the relationship between the excitation light control value (driving current value) and the excitation light powers of the excitation light sources 8-1 and 8-2 when the correction by the above function fc is performed. That is, when the correction by the above function fc is not performed, as shown in FIG. 14, in "state 1", each excitation light power represented by the solid lines 20 and 21 increases linearly as the excitation light control value increases. However, in "state 2", the excitation light power (driving current value) of one excitation light source 8-1 shown by the solid line 20 is maintained at the maximum value by the limiter 11A-8, and the other excitation light source 8 shown by the solid line 21 is maintained. It can be seen that the excitation light power of -2 increases linearly with the same inclination as in "state 1" until it reaches the maximum value. On the other hand, when the correction is performed by the above function fc, as shown by the solid line 21 in FIG. 15, in "state 2", the excitation light source 8-2 has a larger inclination than the inclination in "state 1". It can be seen that the excitation light power increases and the shortage of the excitation light power of the excitation light source 8-1 is supplemented by the excitation light power of the excitation light source 8-2. Note that the limiter 11B-9 shown in FIG. 12 is within the allowable range when the output (current value) of this function fc exceeds the allowable range of the excitation light source 8-2, similarly to the above limiter 11A-8. This is to keep it below the maximum value. With the above configuration, in the control unit 10 (feedback control unit 11) of the present embodiment, until the excitation light power of the excitation light source 8-1 reaches the maximum value (during "state 1" shown in FIG. 15). , The excitation light powers of both excitation light sources 8-1, 8-2 are the functions f, respectively, as shown by the solid lines 20 and 21 in FIG.₁It is controlled based on the excitation light control value (output of the multiplier 11-3) obtained by using the feedback coefficient a variable according to the input light power. On the other hand, after the excitation light power of the excitation light source 8-1 reaches the maximum value (during "state 2" shown in FIG. 15), the excitation light power of the excitation light source 8-1 is maintained at the maximum value. The shortage due to this (the shortage current value obtained by the limiter 11A-8) is corrected by the function fc and the function f.₅By being applied to the current value obtained in, the excitation light power of the excitation light source 8-2 has the same feedback coefficient a in "state 1" and "state 2" as shown by the solid line 21 in FIG. Moreover, it increases so as to make up for the shortage caused by the excitation light source 8-1. As described above, according to the present embodiment, when a plurality of excitation light sources 8-1, 8-2 are used for the optical amplifier 1, and the excitation light power of one excitation light source 8-1 is insufficient, another excitation light source is used. Even when the excitation light power of 8-2 is used to make up for the shortage, stable AGC can be carried out without apparently eliminating states with different oscillation limits and causing an oscillation phenomenon. Further, in this case, the feedback coefficient a is made variable according to the input light power by the control unit 10 (feedback control unit 11) common to each excitation light source 8-1, 8-2, so that the excitation light source 8-1 It is not necessary to provide a separate control unit for each of 8 and 2, and high-speed AGC can be realized at low cost. In the above-mentioned correction function for the shortage of excitation light power, for example, as shown in FIG. 13, the feedback coefficient a is fixed (function f).₁It may be applied to the existing feedback control unit 11. It is also possible to apply it to the control unit 10 provided with the feedforward control unit 13 described above in the second embodiment. Further, in the above-mentioned example, when the excitation light power of the excitation light source 8-1 is insufficient, the shortage is compensated by the excitation light power of the excitation light 8-2, but of course, on the contrary, the excitation light source 8-2 The same applies to the case where the excitation light power of the excitation light source 8-1 is used to make up for the shortage. Further, in the above example, the excitation light sources 8-1 and 8-2 are arranged on the input / output side of the EDF4 so as to sandwich the EDF4, but even if they are arranged on either the input side or the output side. Good. Furthermore, the number of excitation light sources is not limited to two, and three or more may be arranged. [D] Others The present invention is not limited to the above-described embodiments and modifications thereof, and can be variously modified and implemented without departing from the spirit of the present invention. For example, in the above-mentioned example, the case where the feedback control unit 11 performs control using the difference or the difference and its integral value has been described, but the control using the combination of these and the differential value of the difference is performed. You can also do it. Further, in the above-described example, the configuration in which the feedback coefficient a in the feedback control unit 11 is made variable according to the input optical power of the optical amplifier 1 has been described. For example, the output optical power or both the input optical power and the output optical power have been described. It may be made variable according to.

As described above, according to the present invention, when controlling the gain of the optical amplifier, the gain control amount is made variable according to at least one of the input optical power and the output optical power of the optical amplifier, so that the signal light It is possible to follow the input power fluctuation of the above at high speed without causing an oscillation phenomenon, an increase in the size of the optical amplifier, an increase in power consumption and heat generation, and stable optical communication can be realized. Therefore, its usefulness is considered to be extremely high in the field of optical communication.