The present invention relates to an optical signal processing device. It finds particular application in optical logic and communications systems, for instance as an optical signal limiter, or as a noise filter.

It is sometimes useful in optical logic or communications systems that the intensity of an optical signal is limited to a maximum value which is at least substantially constant. Such a limited signal could for instance be used as a bias signal, to ensure reproducable peak laser power in pulsed laser systems, to control pulse shapes, as a digital input signal which avoids saturating a photodetector, or to equalise signal amplitudes in a multiplex system. Further, a device which limits optical signal intensities can operate as a noise filter since modulation superimposed on an optical signal can be cut out.

It is known to use semiconductor optical devices in optical logic and signal processing. They are advantageous in that they can be designed to operate at order of low power levels, for instance at about the 10 mW which might be available in optical logic and signal processing. They physically take up little space, operate at wavelengths compatible with those common in optical logic and communications, and potentially can be monolithically integrated with other optical components.

A factor in the choice of materials for optical devices is the fact that silica optical fibres, which are the basis of present optical communications systems, have loss minima at 0.9 μm, 1.3 um and 1.55 μ approximately. Accordingly there is an especial need for devices which

show favourable characteristics when operated using optical radiation in the wavelength range from 018 to 1.65 urn, and especially in the ranges from 0.8 to 1.0 μ and from 1.3 μm to 1.65 μ . (These wavelengths, like all the wavelengths herein except where the context indicates otherwise, are in vacuo wavelengths). Materials which have been found suitable for the manufacture of devices which show such favourable characteristics comprise the III-V semiconductor materials, including gallium arsenide, indium gallium arsenide, gallium alluminium arsenide, indium phosphide and the quaternary materials, indium gallium arsenide phosphides (In Ga, As P, ) . ^' With regard to the quaternary materials, by suitable choices of x and y it is possible to lattice-match regions of different ones of these materials to neighbouring III-V materials in a device while being able to select the associated band gap equivalent wavelength!

A known device for use in conjunction with a laser to shape pulses is a non-linear Fabry-Perot (FP) etalon. A FP etalon is a simple optical cavity which can be fabricated from semiconductor materials, having planar reflecting endfaces of about 30Δ reflectivity, and (optionally) waveguiding properties. Limiting action is achieved by tuning the peak output of the etalon at low signal input to the source laser frequency: as the input intensity rises, the etalon detunes, keeping the transmitted power approximately constant. ^'

Alternatively, passing a signal through a non-linear medium can be used to limit the signal. It is known to use a device wherein the far field of the output signal spreads at increasing input powers. By using an aperture stop, the increasing input powers can be at least approximately balanced out.

Another device suggested but not demonstrated for use as an optical signal li iter is a multiple quantum well guided wave structure, also fabricated from semiconductor materials. Such a device is discussed in "Nonlinear Guided Wave Applications", Optical Engineering 24 (4) 1985, by CT Seaton, Xu Mai, GI Stegeman and H G Winful. Optical limiting action is obtained because either one or both of the bounding media are characterized by a self-defocusing nonlinearity.

However, known devices tend to suffer from one or more disadvantages. For instance, they might only operate at an input signal intensity higher than those suitable for optical logic or communications systems. Another disadvantage, for instance of etalons, can be that the output signal is multistable and tends to show oscillation as a consequence. In the case of a multiple quantum well device, it is likely that even if demonstrated, the devices would be difficult to manufacture, having for instance low layer thickness tolerances.

A semiconductor laser is known for use as a source of optical radiation, or as an amplifier. A laser is commonly based on a wafer grown from the III-V semiconductor materials mentioned above. The layers of the wafer are selectively doped to provide a p-n junction, in the vicinity of which lies an active region. Photons can be generated in the active region by radiative recombination of electron-hole (carrier) pairs under a driving current applied across the junction. The distribution of refractive index in the laser is such as to provide a waveguiding region along which the photons generated are guided. Feedback is provided to the waveguiding region, for instance by reflective end facets of the laser (a Fabry-Perot laser). In alternative versions, feedback may be provided in a "distributed"

fashion (a DFB laser), for instance by means of corrugations in an interface which lies near the active region, or by a grating external to the laser active/ aveguiding regions.

If optical radiation is input to the active region of a laser and a driving current applied, amplification of the radiation occurs even when the driving current is below the threshold current necessary for lasing action to occur! The relationship between input and output radiation intensity is non-linear and can show bistability, the output intensity switching rapidly to a new value as the input intensity reaches a relevant switching level.

A feature of a DFB laser is that the input signal will be reflected to an extent dependent on its wavelength. Any passive corrugated waveguide will act as a band reflection filter, for optical radiation of wavelength A free space, whose period Λ is close to the back scattering Bragg condition

-CL S px where p is an integer and η is the

2η effective refractive index of the material of the waveguide. This feature creates a series of stop bands within which optical radiation is substantially reflected rather than transmitted! In a DFB active device, the reflected spectral response is more complicated, showing a strong peak to either side of each stop band, caused by the device gain. (In the present specification, references to a stop band will be to the stop band relevant at very low input signal intensities.)

The optical radiation can be input to a first end of the active region of a laser amplifier and the output taken from a second end of the active region, ie the laser amplifier can be operated in transmission. Alternatively,

the output can be taken from the first end also, ie the laser amplifier can be operated in reflection.

The relationship between input and output radiation intensity is complicated. For instance, the input radiation undergoing amplification reduces the free carrier concentration and hence the gain of the laser. The refractive index varies with gain, and the degree of amplification of the input radiation is dependent on the relationship between the input wavelength and the refractive index. Additionally, temperature will affect the refractive index and both the laser drive conditions and the input radiation affect the temperature. Hence, overall, the interaction of gain, refractive index, driving current and input radiation is difficult to specify for a particular laser.

It has now been found that a DFB laser used as an amplifier can show an optical limiting characteristic.

It is object of the present invention to provide an optical signal processing device suitable for use as an optical limiter in optical logic or communications systems.

According to the present invention there is provided an optical signal processing device comprising a semiconductor DFB laser amplifier operated in reflection, means for coupling an optical input signal to the amplifier, the signal being detuned from an output signal peak to the short wavelength side of the wavelength range relevant to an amplifier stop band, and means for applying a driving current to the amplifier, the input signal reaching at least part of the time an intensity greater than a threshold value above which the output signal intensity of the amplifier is substantially independent of the input signal intensity.

Preferably where the input signal is a binary digital signal, its intensity for a "1" value is always greater than that threshold value.

Optical devices according to embodiments of the present invention have the advantage that the relationship between input and output signal intensities can show an extensive region within which the output signal intensity is substantially independent of input signal intensity.

Further, because the amplifier is operated in reflection, the devices can offer the ability to use samples of higher linear absorption!

Because the devices are active rather than passive, there is also considerable control available over the operating parameters used!

The input-output signal intensity relationship of an amplifier according to an embodiment of the present invention can be used to limit noise in an optical signal since intensity variations of the input signal will not be reproduced. Where the signal is a binary digital signal, noise can be limited not only on "1" values of the signal, where the signal intensity lies above the threshold value, but also on "0" values. This is because if the Gaussian spread of the signal intensity, associated with noise at the "1" value, is reduced, then the decision level selected to differentiate between a "1" and a "0" value can be raised, so reducing the effect of noise on a "0" value at the decision level concerned!

An optical limiter according to an embodiment of the present invention will now be described, by way of example only, with reference to the accompanying figures in which:

Figure 1 shows a schematic representation of the l miter in use;

Figure 2 shows in graph form the output signal of the limiter in response to various input signals of different wavelengths; and

Figure 3 shows the response of the l miter to a high frequency sinusoidal input signal.

Referring to Figure 1, the limiter comprises a DFB amplifier 1 used in reflection. An optical input signal is provided by a tunable laser diode source 2, in combination with an attenuator 7.

A beam splitter 3 mounted between the source 2 and the amplifier 1 deflects a portion of the optical output of the source 2 to an input signal monitor 6, and a portion of the output of the amplifier 1 to an output signal monitor 5. Interaction between the source 2 and the amplifier 1 is prevented by an isolator 8, between the beam splitter 3 and the source 2, and the attenuator 7 is used to modify the output of the source 2 to produce a controllable input signal to the amplifier 1!

The amplifier 1 is a DFB ridge waveguide laser, 300 urn long, with a coupling coefficient times length of 2.4, comprising InP with an InGaAsP active layer. Threshold current at room temperature is 57 mA and the emission wavelength 1525 nm. The diode has an active cross section

2 of 0.4 ^ιm .

The source 2 is a grating - tuned external cavity laser which provides a single-mode signal. This laser is an anti-reflection coated ridge waveguide laser, tunable in the range from 1450 to 1580 μm inclusive, again comprising InP with an InGaAsP active layer.

The isolator 8 is provided by two isolating devices, giving together 60 dB isolation. Maximum coupled powers from the source 2 to the amplifier 1 of a few hundred μW can be obtained, as deduced from the resultant photocurrent induced in the amplifier 1. The beam splitter 3 comprises a simple uncoated glass slide, and a fast PIN-preamp combination (not shown) provides temporal resolution of 100 p sees to allow switching speed measurements by directly modulating the tunable source 2.

A method of operating the switch will now be described, and results discussed.

The amplifier 1 is driven by a driving current so as to produce a material gain in the amplifier 1 of 0.95 times the lasing threshold gain. The source 2 is then driven to produce a series of input signals for the amplifier 1. Each input signal increases then decreases steadily in intensity, starting at zero, and is detuned, from a cavity resonance of the amplifier 1 at low input intensities, by a characteristic amount. All the input signals have a wavelength shorter than the stop band of the amplifier 11

Referring to Figure 2, the output signal intensity I of the amplifier 1 tends to show bistability at low values of input signal intensity I.. However, as I. increases beyond those low values, I becomes substantially insensitive to I.. (As shown on Figure 2, both signal intensities I and I. are normalised using a scaling intensity I 1)

The amount by which each input signal is detuned, from s cύ. an output peak on the short wavelength/of a stop band, corresponds to a single-pass (ie non-reflected) phase change of 0106τr, 0.13ιr, 0.19*, 0.26τr, and 0.32τr respectively! It can be seen that as the detuning increases, the threshold intensity value of I. above which I is substantially independent also increases. However, at most the normalised threshold intensity value is just below 0.011 The scal ng intensity I_ has a

5 2 value of about 8 x 10 W/cm , which over the active cross section of the amplifier 1 corresponds to a factor 8 x 10 ⁹ x 0.4 x 10 ^'12 W, or 3.2 x 10 ^'3 W. Thus the highest normal sed value of I. corresponds to an absolute value of about 30 μW.

The results described above are relevant to an input signal whose intensity varies relatively slowly. It is important to avoid operating conditions whereunder the amplifier 1 shows a spiked, resonant response as might occur with a high frequency sinusoidal input signal. Referring to Figure 3, if the input signal has a period of the order of the carrier recombination time (about 1.7 nsecs), such a spiked response 9 is predicted. This type of response is understood tostem from changes in the material refractive index of the amplifier 1 due to changes in the optical input intensity.

To investigate the presence of the spiked response, the output of the amplifier 1 was calculated for four different input signals, an effectively steady state signal (Graph (a)), and sinusoidal signals whose periods corresponded to 4, 8 and 12 times the carrier recombination time respectively (Graphs (b) to (d) in that order). In the case of each sinusoidal signal, the spiked response 9 could be observed.

To avoid the spiked response 9, an oscillating input signal whose intensity is greater than a threshold value associated with a spiked response should be used. In the case of a binary digital signal, the intensity of the input signal in the case of a "1" value should be so selcted. In embodiments of the present invention, a

« threshold value of less than 100 μW, and even of less than 40 μW, should be achievable.

In the embodiment of the present invention described above, the period of the grating selected for distributed feedback was second order wth regard to the wavelength of the input signal. That is, it was of the order of /2 μm. It is predicted that a lower order grating would produce a stronger coupling coefficient /n the amplifier and so reduce the input signal intensities at which the device can be satisfactorily operated.