The function of a CLR system is based on the fact that received laser radiation from a target is mixed with a local oscillator radiation (LO radiation). In this way, frequency information from the reflection from the target is recovered and, as a result, e. g. Doppler shift of the target radiation can be measured. Based on this, the speed of a target in the radial direction relative to the CLR system can be calcu- lated. The result thus gives, depending on the type of target, the speed of a solid target or the wind velocity. The principle also allows shot noise dominated detec- tion, which has the maximum sensitivity.

A general description of CLR systems is given by G. W. Kamerman,"Laser Radar" in The Infrared & Electro-Optical Systems Handbook, Vol. 6, Ed. C. S. Fox, Infrared information analysis center and SPiE optical engineering press (1993), which is herewith incorporated by reference. Wind-measuring CLR systems are generally described by R. M. Huffaker and R. M. Hardesty,"Remote sensing of atmospheric wind velocities using solid-state and C02 coherent laser systems", Proc. IEEE, Vol. 84, pp 181-204, February 1996, which is herewith incorporated by reference.

Since the transmitter laser (single frequency), in a pulsed coherent laser radar system, generates a laser pulse, a stable single frequency CW (continuous wave) LO laser has been used up to now to give LO radiation to the necessary mixture with received laser radiation. This known procedure places great demands on the frequency characteristics of both LO laser and transmitter laser. Both must be very frequency stable. It is common to have frequency chirp (frequency changes) of the transmitter laser in the pulse duration, which reduces the accuracy in the frequency measurements.

It would be very advantageous if pulsed coherent laser radar systems could be pro- vided without expensive LO lasers. If also the problem of frequency chirp could be

reduced or obviated, which would allow very accurate measurements of Doppler shift, it would be great progress.

The invention solves these problems by being designed as is evident from the following independent claim. Suitable embodiments of the invention are defined in the remaining claims.

The invention will now be described in more detail with reference to the accompany- ing drawings, in which Fig. 1 shows a first embodiment of a CLR system with a fibre-optic pulse train generator according to the invention, Fig. 2 illustrates examples of generated pulse trains in the arm A4 in Fig. 1, the left diagrams showing individual Gaussian pulses and the right showing the sum of the pulses, which in this example generates a quasi-constant LO effect, Fig. 3 shows the same as Fig. 2 for a fibre length which is three times longer, which causes a greater time separation between the pulses, and Fig. 4 illustrates a second embodiment of a CLR system with a fibre-optic pulse train generator according to the invention.

The basic idea of the invention is that a certain part of the pulsed transmitter radia- tion is deflected and conducted to a fibre-optic ring, from which pulses are deflected revolution by revolution and form a pulse train. The reflected radiation from a certain distance can thus on the detector be mixed with a copy of itself. Signal processing then occurs in the normal way.

There is no special LO laser and problems caused by frequency chirp and limita- tions in time coherence can be eliminated or reduced to a considerable extent compared with prior-art technique. For applications which concern solid targets, defined as the surface reflection totally dominating the reflection of the target, frequency chirp can be fully eliminated when the fibre length L is matched with the distance. When the distance is not matched, the frequency chirp will give a signal widening (spectrally) although to an essentially smaller extent compared with prior- art technique. For distributed targets, such as particles in the atmosphere, restric- tions owing to frequency chirp will remain to approximately the same extent as in

prior-art technique. However, problems caused by frequency instabilities in the LO laser will disappear.

Fig. 1 shows a CLR system with a fibre-optic pulse train generator according to an embodiment of the invention. A laser 1 generates laser radiation pulses. A beam splitter 2 splits the energy of the pulses and conducts a first part to the output unit 12,13 of the system for transmitting laser radar pulses. The output unit is of a known type and can in the usual way comprise a B/4 plate 12 and a beam expander 13. It is also possible for the radiation to be conducted from the laser to the output unit in optical fibres and fibre-optic components.

The beam splitter 2 conducts a second part of the energy of the pulses into a first optical fibre 4, normally via a focusing lens 3. From the first optical fibre, the radia- tion is conducted to the one input on a first fibre-optic coupler 5 with two inputs and two outputs. The coupler connects, via its one output A1, part of the energy of each pulse to a fibre-optic ring, which comprises a second fibre-optic coupler 6 of the same type as the first and a second optical fibre L.

One output A3 of the second coupler conducts part of the energy of each pulse remaining in the fibre-optic ring while the other output A4 discharges pulses in a pulse train with a pulse frequency depending on the time of a pulse passing round in the fibre-optic ring. The pulse train is conducted to a detection device where it is combined with return pulses from targets.

It is a great advantage to use two couplers in the fibre-optic ring compared with using one coupler only. In order to obtain a practically operating device, one wants to have a pulse train out of the optical ring with an effect that decreases only slowly.

This means in an imaginary case involving one optical coupler that the coupling degrees must be such that the first pulse reaching a subsequent detection device will be very great and that there is thus a great risk of damaging the same. This problem is avoided by using two couplers in the ring, see also below for additional reasoning about coupling degrees.

The detection device can, according to Figs 1 and 4, comprise a third fibre-optic coupler 8, in which case the pulse train can be conducted via a third optical fibre 7 to one input of the third fibre-optic coupler and return pulses from targets via a fourth optical fibre 9 to another input of the same coupler. In the example in the

Figure, the radiation is deflected there by a beam splitter 11 and focused into the fourth optical fibre by a lens 10. In the third coupler, the two radiations are combined and the result is detected by means of detectors D1, D2 at one or both outputs of the third coupler 8.

The detector signal can be generally written as wherein 9 is the sensitivity of the detector, PLO is the effect of the local oscillator, Pter is the received effect, fLo is the frequency of the LO laser, ftar is the frequency of radiation reflected by the target and po is a phase difference between LO radia- tion and target radiation. It is evident from the formula that in the solid target appli- cation, any frequency dependencies on the time of the transmitter laser will be compensated for by the fact that the frequency dependency of the LO laser is exactly the same.

The length of the optical fibre in the fibre-optic ring, the second optical fibre L, determines the time interval between the pulses, see Fig. 1. L can easily be adapted depending on what application and laser source are involved. The coupling degree of the couplers 5 and 6 determines the pulse effects. Moreover, the pulse duration, the fibre length and the coupling degree of the couplers determine how the effect varies with time.

The output effects in the fibre arms A2 and A4 are given by wherein PA2 and PA4 are the effect in the arms A2 and A4, respectively, c is the speed of light, L is the optical length of the ring, C1, C2, C3 and C4 are coupling degrees, to is the starting time, n is an integer which indicates how many revolutions the radiation has made through the ring and N indicates how many revolutions the

radiation is allowed to be in the ring before it is dumped, which in Fig. 4 is marked with D. See also below.

Figs 2 and 3 illustrate examples of generated pulse trains in the arm A4 in Fig. 1.

Gaussian pulses have here been assumed. However, the invention is not limited to Gaussian pulses, but other forms of time are also possible.

The left diagrams show individual Gaussian pulses and the right diagrams show the sum of the pulses. In the example, the beam splitter 2 splits 100 W of the trans- mitter radiation and the couplers 5 and 6 are 99: 1 coupleres.

By choosing 99: 1 coupleres, where 99% of the radiation which is conducted through the first optical fibre 4 to the first coupler 5 is connected to the output A2 and 1% is connected via the output A1 to the fibre-optic ring and of this radiation, to the second coupler 6,99% is connected via A3 remaining in the fibre-optic ring and on) y 1% is discharged via the output A4, an essentially constant level of the pulses in the pulse train is obtained in the third optical fibre 7. This occurs by, in each passing of an coupler in the optical fibre-optic ring, the major part, 99%, remains in the ring and on 1% is discharged as pulses in the pulse trains in the arms A2 and A4. If the pulses in the pulse trains are sufficiently close, as in Fig. 2, a quasi- constant LO effect is generated.

It should be noted that the first pulse which is connected to the output A2 has a high effect, which in an imaginary case with only one optical coupler in the ring could damage a subsequent detection device.

The third coupler 8 can in many cases suitably be a 50: 50 coupler which mixes the radiation from the two inputs in equal proportions which are presented with equal strength on the two outputs. In other cases, other proportions may be preferred, such as 80: 20.

For efficient utilisation of the available laser effect, it is convenient to use single mode fibres for the wave length at issue. If multimode fibres are used, there is risk that the radiation splits between different modes, which can result in signal loss.

Moreover, control of the polarisation is important. A transmit/receive switch included in the system is often based on polarisation. In Figs 1 and 4, the transmit/receive switch consists of a polarisation-dependent beam splitter 11 and a X/4 plate 12.

This configuration results in only one linear component of received radiation being conducted to the detector. The LO radiation must then have the same polarisation state for a maximum signal. The polarisation can be controlled in various known manners. Of course, polarisation-preserving fibres can be used. However, it is also possible to use a less expensive fibre and carefully monitor the"winding"of the fibre in the system. Finally, it is possible to use a polarisation modulator together with a less expensive fibre. The polarisation modulator can be both a manual mechanical modulator and an electronically controlled modulator.

Pulses which are rectangular in terms of time are ideal since any time overlap between the LO pulses may cause interfering signals at specific frequencies.

A plurality of variants of the invention are conceivable, of which a few are shown in Fig. 4. In order to determine whether the target moves towards or away from the CLR radar, an acousto-optical modulator or some other component for frequency shift of the radiation can be connected to the first 4, third 7 or fourth 9 optical fibre.

Moreover, a fibre-optic switch S can be arranged in the fibre ring to interrupt the circulation of the radiation in the fibre-optic ring after a predetermined period.

It is also possible to use switches S to connect in parallel other optical fibres L, Ln of different length in the fibre-optic ring. By using these switches, different fibre lengths in the ring can easily be chosen, which results in a simple choice of the pulse repetition frequency in the pulse trains.

Finally, a plurality of fibre-optic rings of the stated type can be connected in succes- sion. If, for example, one more ring is connected to the arm A2, one more LO with almost the same effect/time dependency as the first is obtained. This can be usable in applications where two measuring directions are of interest and when stationary optics (not scanning) is used. If further measuring directions are of interest, it is possible to provide the same number of LO as measuring directions by connecting additional fibre-optic rings in succession.