GB2257866A

GB2257866A - Tracking radar system

Info

Publication number: GB2257866A
Application number: GB9217244A
Authority: GB
Inventors: John William Attwood; Michael Arthur Jones; John Thomas Floyd; Alan James Mitchell
Original assignee: GEC Marconi Ltd; Marconi Co Ltd
Current assignee: BAE Systems Electronics Ltd
Priority date: 1988-07-29
Filing date: 1992-08-14
Publication date: 1993-01-20
Anticipated expiration: 2012-08-14
Also published as: GB2257865A; FR2683914B1; GB8917172D0; DE3943459C2; GB2257865B; GB2258108B; FR2683914A1; NL8915006A; GB8818100D0; IT1245930B; GB9217242D0; GB2258113A; SE9001843D0; GB2257866B; IT9067292A0; SE9001843A0; GB9217248D0; GB9217244D0; GB2257867B; GB2258108A

Abstract

A tracking radar system has a multi-element aerial (1). A receiver (10) processes radio-frequency sum (S) and difference (D) signals derived from the aerial to produce corresponding intermediate frequency sum (S<1>) and difference (D<1>) signals. The oscillator (62) of a phase-locked loop (60) locks on to the frequency of the intermediate frequency sum signal (S<1>), and a phase-sensitive detector (34) compares the oscillator (62) output with the intermediate frequency difference signal (D<1>) to produce an output signal representing the direction of a target relative to the aerial. The phase-locked loop includes a variable gain circuit (300). Responsive to the intermediate frequency sum signal (S<1>), a threshold detector (304) produces a first or second output signal. The first threshold output signal is combined with a target acquisition signal to cause the variable gain circuit (300) to slowly decrease the bandwidth of the phase-locked loop (60). The second threshold output signal causes the variable gain circuit (300) to rapidly increase the bandwidth of the phase-locked loop (60). The invention is particularly concerned with a method of providing AGC in the receiver. <IMAGE>

Description

1 ú_1 5 7 J) 5 0 Tracking Radar Systems The present patent application is

divided out of Application No.8917172.2 filed 27 July 1989.

This invention relates to radar systems and more particularly to staticsplit tracking radar systems, i.e. radar systems in which directivity is found by simultaneous comparison of signals derived from seperate aerial outputs.

In a typical static-split tracking radar system, a target is tracked by means of a multi-element aerial, producing a plurality of radio frequency outputs. These outputs may typically be added and subtracted to produce a sum signal and at least one difference signal, which signals are processed in a multi-channel receiver, and the resulting intermediate frequency outputs are then compared in amplitude and/or phase so as to produce at least one output signal, representing the direction of the target relative to the aerial.

We have previously devised a tracking radar system which comprises an aerial arrangement having a plurality of outputs, means for deriving from the aerial outputs a sum signal representative of the sum of the aerial outputs and a difference signal representatve of the difference of the aerial outputs, a receiver for processing said sum and difference signals to produce corresponding intermediate frequency sum and difference signals, means for comparing the intermediate frequency sum signal with the output of an oscillator in a phase-locked loop and using the resulting signal to control the oscillator frequency so as to cause the oscillator to lock on to the frequency of the intermediate frequency sum signal, and a phase sensitive detector for comparing the intermediate frequency difference signal with the output of the oscillator to produce an output representative of the direction of a target relative to the aerial.

In this way, the phase-locked loop acts effectively as a narrow band filter to select a single intermediate frequency, and thus assists in multiple target discrimination (M), that is, in discriminating between targets having different doppler shifts.

The phase-locked loop may however in certain circumstances prove to be a disadvantage. If for example the missile is tracking a single target and if the target frequency suddenly changes due for example in the case of an aircraft to a sharp turn then the frequency of the intermediate frequency signal sum at the input to the phase-locked loop will move outside the bandwidth of the phase-locked loop. If this should happen rapidly then the phase-locked loop would not be able to follow the incoming intermediate freqency sum signal and the missile would lose tracking and almost certainly miss the target.

The introduction of multiple target discrimination using the phase-locked loop referred to above may introduce problems in the automatic gain control of the missile receiver circuitry. This is because the phaselocked loop will lock to the frequency of one of the targets with a bandwidth sufficient to include that target signal but not sufficient to include any other target signals which may be amplified in the sum channel of the receiver prior to the phase-locked loop. The spectral component of the output of the oscillator in the phase-locked loop will therefore be of greater amplitude than the corresponding component in the sum channel of the receiver because the automatic gain control operates on the total sum signal passing through the receiver. The angular scaling of the desired target signal will therefore be reduced and the present invention provides possible alternative systems to alleviate this problem.

According to the present invention there is provided a tracking radar system comprising an aerial arrangement having a plurality of outputs, means for deriving from the aerial outputs a sum signal representative of the sum of the aerial outputs and a difference signal representative of the difference of the aerial outputs, a receiver for processing said sum and difference signals to produce corresponding intermediate frequency sum and difference signals, means for comparing the intermediate frequency sum signal with the output of an oscillator in a phase-locked loop and using the resulting signal to control the oscillator frequency so as to cause the oscillator to lock on to the frequency of the intermediate frequency sum signal, a phase-sensitive detector for comparing the intermediate frequency difference signal with the output of the oscillator to produce a signal representative of the direction of a target relative to the aerial, and a quadrature phase-sensitive detector connected for comparison of the intermediate frequency sum signal and the T."12 phase shifted output of the oscillator, the output of the quadrature phase-sensitive detector being connected to the receiver for the provision of automatic gain control.

According to a second aspect of the present invention there is provided a tracking radar system comprising an aerial arrangement having a pluralty of outputs, means for deriving from the aerial outputs a sum signal representative of the sum of the aerial outputs and a difference signal representative of the difference of the aerial outputs, a receiver for processing said sum and difference signals to produce corresponding intermediate frequency sum and difference signals, means for comparing the intermediate frequency sum signal with the output of an oscillator in a phase-locked loop and using the resulting signal to control the oscillator frequency so as to cause the oscillator to lock on to the frequency of the intermediate frequency sum signal, a phase-sensitive detector for comparing the intermediate frequency difference signal with the output of the oscillator to produce an output representative of the direction of a target relative to the aprial, a quadrature phase- sensitive detector connected for comparison of the intermediate frequency sum signal and the 1r12 phase shifted output of the oscillator, and a divider for division of the output of the phase-sensitive detector by the output of the quadrature phase-sensitive detector, the output of the divider being connected to the receiver for the provision of automatic gain control.

Preferably the system includes a low pass filter interposed between the output of the quadrature phase-sensitive detector and the di vi der. The time constant of the filter is preferably greater than 100 Ms.

A radar system in accordance with the invention will now be described, by way of example, with reference to the accompanying drawings of which:Figure 1 is a schematic front elevation of the aerial of the radar system; Figure 2 is a schematic block circuit diagram of a previously devised system; Figure 3 shows in block diagrammatic form a modification to the narrow band phase-locked loop of Figure 2 according to the present invention; Figure 4 shows the phase relationships between the phase shifted output of the voltage controlled oscillator in the circuit of Figure 3 and two target signals; Figure 5 shows the sum signal output and the quadrature phase-sensitive detector output for the circuit of Figure 3; Figure 6 shows in block diagrammatic form circuitry for adding constant amplitude first sidebands to the output of the voltage controlled oscillator of the narrow band phase-locked loop of Figure 3; Figure 7 shows waveforms illustrating the improved scaling of the missile receiver when modified according to Figure 6 with the narrow band phase- locked loop locked to a single target; Figure 8 shows waveforms illustrating the improved scaling of the missile receiver when modified according to Figure 6 with the narrow band phase- locked loop locked to a single target; Figure 9 shows in block diagrammatic form circuitry for adding pseudo sidebands to the narrow band phase-locked loop voltage controlled oscillator of Figure 3; Figure 10 shows in block diagrammatic form circuitry for the elimination of interference on the multiplexing from the multiple target discriminator (narrow band phase-locked loop); Figure 11 shows in block diagrammatic form an alternative form of elimination of interference on the multiplexing frequency in a missile guidance system; Figure 12 shows in block diagrammatic form circuitry for the automatic control of the bandwidth of the narrow band phase-locked loop of Figure 3; Figure 13 shows in block diagrammatic form circuitry for the latching out of the multiple target discriminating narrow band phase-locked loop of Figure 3 and for control of the doppler sweeping; and Figure 14 shows a doppler sweep waveform associated with the circuitry of Figure 13.

The system to be described is part of a homing head for an air-to-air missile. The system is a semi-active one, in which the target is illuminated with radio waves from a source remote from the missile, e.g. from the radar of the aircraft which launched the missile.

Referring to Figure 1, the system includes an aerial arrangement 1, comprising an array of four aerial elements la - 1d, each of which has its own feed antenna and reflector dish. The axes of the four elements la - Id are all parallel to each other, so that when a radio signal is received from a target by the aerial, the resulting output signals from the four elements are all of substantially equal amplitude, but differ in phase, according to the direction of the target relative to the aerial. The aerial arrangement 1 is mounted on gimbals (not shown) so that it can be tilted about azimuth and elevation axes, by means of servo motors (not shown).

Referring now to Figure 2 a radar system is shown which we have previously devised. The four aerial output signals are designated A,. A2, A3 and A4. These output signals are added and subtracted in a suitable comparator circuit 2, to produce three signals S, Del and Daz, as follows:- S = (A, + A 2 + A3 + A4) Del = (A, + A2) - (A3 + A4) Daz = (A, + A3) - (A2 + A4) S is refrred to as the sum signal, and is equal to the sum of the aerial outputs. Del and Daz are referred to as the elevation and azimuth difference signals, ard their amplitudes are measures of the elevation and azimuth error angles between the aerial axis and the target. Because of a 900 phase shift in the comparator circuit 2, the signals Del and Daz are approximately in phase with the sum signals. These three signals, Del, Daz and S pass through a rotating microwave joint on the aerial 1 to the following circuitry.

The two difference signals are multiplexed together in a multiplexer 3, which is controlled by a square wave modulating signal M from an oscillator 4. The multiplexer 3 comprises two phase switches 5 and 6 for phase modulating the signals Del and Daz respectively. Phase switch 5 is driven directly by the output of the oscillator 4, and produces alternate 00 and 1800 phase shifts in the Del signal, in phase with the modulating signal M. Phase switch 6 is driven by the oscillator 4 via a 900 phase shifter 7, and produces alternate 0' and 1800 phase shifts in the Dal signal, in quadrature with the modulation of the Del signal.

The outputs of the phase switches 5 and 6 are added together in a hybrid circuit 8, to produce a multiplexed difference signal D. It will be seen that the signal D, in each cycle of the modulating signal M, passes through the following sequence of values:- D1 = Del + Daz D2 = Del - Daz D3 = -Del - Daz D4 = -Del + Daz The difference signal D is fed to a hybrid circuit 9, along with the sum signal S, to produce two output signals S + D and S - D, which are fed respectively to two channels of a superheterodyne receiver 10. Up to this stage of the system, the signals are all at microwave frequencies (except, of course, the modulating signal M), and therefore the comparator 2, phase switches 5 and 6, and hybrids 8 and 9 are all microwave components and are conveniently constructed using microstripline techniques.

The receiver 10 comprises a first local oscillator 11, the output of which is mixed with the signals S + D and S - D in mixers 12 and 13, to convert them to a suitable first intermediate frequency. The intermediate frequency signals are passed through intermediate frequency amplifiers 14 and 15 to band pass filters 16 and 17. These filters select only a narrow range of frequencies, corresponding to a narrow range of doppler shifts in the radio signal received from the target, i.e. corresponding to a narrow range of relative target-to-aerial speeds. For this reason, the filters 16 and 17 are referred to as speed gates.

The filtered signals are passed through further intermediate frequency amplifiers 20 and 21 to mixers 24 and 25, where they are mixed with a signal from a second local oscillator 26, to convert them to a suitable second intermediate frequency. The second intermediate frequency signals are then respectively passed through gain trim and phase trim circuits 27 and 28, and are amplified by second intermediate frequency amplifiers 30 and 31.

The outputs from the amplifiers 30 and 31 are combined in a sum circuit 32 and in a difference circuit 33 to produce intermediate frequency sum and difference output signals S' and D' respectively.

It will be seen that nominally (i.e. assuming that the two channels of the receiver 10 are equally matched in gain and phase difference) the intermediate frequency output S' from the sum circuit 32 is proportional to the radio frequency sum signal S, whilst the intermediate frequency output D1 from the difference circuit 33 is proportional to the radio frequency multiplexed difference signal D. Thus, nominally, the signal D1 passes through the four values D1 - D4, and thus varies periodically in phase with respect to the signal S1, in a symmetrical manner, having an average value of zero. However, since a portion of each of the signals S and D has passed through each of the receiver channels, any mismatch in gain or phase difference between the channels will manifest itself in the output signals S1 and D1. Specifically, if there is any mismatch in gain between the channels, the amplitudes of the four values of signal D1 will be affected by different amounts, so that the average value of signal D1 will no longer be zero, but will be positive or negative depending on whichchannel has the larger gain. Similarly, if there is any mismatch in phase difference between the channels, the phases of the four values of signal DI will be effected by different amounts, so that the average value of signal D1 will again no longer be zero, being positive or negative depending on which channel has the larger phase difference.

The signal S1 is fed to a phase-locked loop 60 comprising a phase sensitive detector 61 arranged to compare the output of the sum circuit 32 with the signal from a voltage controlled oscillator 62. The resultant signal from the detector 61 is integrated in an integrator 63, and is used to control the frequency of the oscillator 62. In this way, the frequency of the oscillator 62 is locked on to the frequency of the intermediate frequency signal from the sum circuit 32. The output signal from the oscillator 62 thus has the same frequency as the output signal from the receiver, but has a much narrower bandwidth. The effect of the phase-locked loop 60 is thus to act as a very narrow band-pass filter for the receiver output signal, the pass-band of this filter depending on the time constant of the integrator 63. This assists in multiple target discrimination, that is in discriminating between targets of closely similar frequencies, especially closely spaced targets and targets in formation.

The output from the oscillator 62 is utilised for feeding to a discriminator circuit 45. This signal is also used for comparison with the output from the difference circuit 33, in the phase-sensitive detector 34, to produce a feedback signal for controlling the gain trim circuit 27. Nominally, the average output of the detector 34 s zero, but if there is any gain mismatch, the output of the detector 34 will develop a DC component. This component is measured in an integrator 35, and is used as a feedback signal to control the gain trim circuit 27, in such a manner as to tend to match the gain of the two channels and thus reduce the output of the integrator 35 to zero.

In order to detect any mismatch in phase, the signal DI i given a 900 phase shift in phase change circuit 36, and is then compared with the signal S1 in a phase-sensitive detector 37. Nominally, the average output of the detector 37 is zero, but if any phase mismatch is present, the output of the detector 37 will develop a D.C. component. This D.C. component is measured in an integrator 38 and is used as a feedback signal to control the phase trim circuit 28, in such a manner as to tend to match the phase differences of the two channels and thus reduce the output of the integral'-or 38 to zero.

When the channels are matched in gain and phase difference, the output of the phase-sensitive detector 34 is proportional to the amplitude of the multiplexed difference signal D, and thus contains information concerning the two difference signals Del and Daz and is an output signal representative of the direction of a target relative to the aerial. To separate this information, the output of the detector 34 is fed to a demultiplexer 39. This demultiplexer is controlled by the modulating signal M from the oscillator 4, delayed in a delay circuit 40 by an amount equal to the overall delay introduced by the receiver 10. The demultiplexer 39 comprises a first phase-sensitive detector 41 in which the output of detector 34 is compared with the delayed modulating signal, to produce an output Eel which is proportional to the amplitude of the error signal D,,,. The demultiplexer 39 also comprises a second phase sensi 1 i ve detector 42 in which the output of detector 34 is compared with the delaypd S modulating signal, shifted in phase by 900 by means of a phase-change circuit 43, to produce an output Eaz which is proportional to the amplitude of the error signal Daz The two outputs Eel and Eaz from the demultiplexer 39 are used as error signals to control the operation of the servo motors (not shown) which tilt the aerial arrangement 1, in such a manner as to tend to reduce the amplitudes of the difference signals Del and Daz to zero. The result of this is to cause the aerial 1 to track the target. The error signals are also fed to the autopilot (not shown) of the missile so as to enable the missile course to be suitably corrected to maintain it on a collision course with the target.

The signal S' is fed to an automatic gain control detector circuit 44, which produces automatLiC gain control signals for controlling the gains of intermediate frequency amplifiers 20,21,30 and 31, in such a manner as to tend to maintain the output of the sum circuit 32 at a constant level.

The output of the voltage controlled oscillator 62 is applied to a frequency discriminator circuit 45, which produces an output signal proportional to the difference between the frequency of the intermediate frequency signal from the oscillator 62 and a predetermined value. This output is integrated in an integrator 46, and the result is used to control the frequency of a voltage controlled oscillator 47, the purpose of which will be explained below.

In operation, the system is subjected to high vibration levels from the missile motor. Under these conditions, the first local oscillator 11 tends to be very noisy, i.e. to have significant noise sidebands. This is undesirable, since the performance of the overall system depends critically on the quality of the output of the first local oscillator. In order to overcome this problem, a so-called rear reference phase-locked loop 49 is used. A rear aerial 50 is provided, to receive some of the illuminating radiation from the parent aircraft - one of the sidelobes of the radiating aerial may be used for this purpose - for use as a reference signal. This reference signal is mixed with the output of the local oscillator 11 in a mixer 51, to convert it to an intermediate frequency. The resultant intermediate frequency signal is passed through an amplifier 52 and a band pass filter 53. The gain of the amplifier 52 is controlled by an automatic gain control detector circuit 54, so as to tend to maintain the output of the filter 53 at a constant level. The output of the filter 53 is compared in a phase-sensitive detector 55 with the output of the oscillator 47, and the result is integrated in an integrator 56 and used to control the frequency of the local oscillator 11.

The result of this is that the frequency of the local oscillator 11 is locked in a fixed relationship to the frequency of the reference signal. Specifically, the frequency of the local oscillator is locked to a value equal to the difference between the reference signal frequency and the frequency of the oscillator 47. In this way, the vibration sidebands of the local oscillator 11 are substantially suppressed.

The automatic gain control circuit 54 controls a trigger circuit 57 which in turn controls the time constant of the integrator 56 according to the level of the automatic gain control signal, so as to reduce the bandwidth of the phase-locked loop 49 if the reference signal level drops below a certain predetermined value. In this way, the loop 49 adapts automatically to the level of the reference signal. Thus, when the reference signal received by the rear aerial 50 is strong, the bandwidth of the phase-locked loop 49 is relatively large, giving a significant reduction in the noise sidebands of the local oscillator 11. On the other hand, when the received reference signal level is low, the bandwidth of the loop 49 is reduced, so as to assist in filtering out the reference signal from noise generated in the amplifier 52. Thus, the signal-to- noise ratio of the reference signal is improved, although at the expense of a drop in the ability of the loop to suppress noise sidebands of the local oscillator 11.

In a modification, the bandwidth of the loop 49 may be varied continuously according to the reference signal level, instead of between two discrete values.

As mentioned above, the voltage controlled oscillator 47 which controls the frequency of the first local oscillator 11, is in turn controlled, by way of the discriminator 45 and the integrator 46, by the frequency of the intermediate frequency signal from the output of the receiver 10. This loop is arranged to maintain the frequency of the first intermediate frequency signal (from the mixers 12 and 13) in the centre of the passband of the speed gate filters 16 and 17, as f Ol 1 Ows. If the relative velocity between the target and the missile or the missile and the parent aircraft changes slightly, the doppler frequency of the radio waves received by aerial 1 will change. Thi s will cause a shift in the frequency of the first intermediate frequency signal away from the centre frequency of the speed gates 16 and 17, which in turn will cause a shift in the frequency Of the second intermediate frequency signal at the output of the receiver. This shift will be detected by the frequency discriminator 45, and will produce a change in the frequency of the oscillator 47, and hence in the frequency of the first local oscillator 11. This in turn will produce a change in the frequency of the first intermediate frequency si gnal, and it is arranged that this change is in such a sense as to tend to return this signal to the centre of the passband of the speed gates 16 and 17.

Thus, it will be seen that the effect of this loop is to maintain the frequency of the first intermediate frequency signal in the centre of the passband of the speed gates 16 and 17. As a result, the system tracks the doppler frequency of the target. Thi s permits the system to discriminate between different targets on the basis of differences in their speeds even although they may be separated by too small an angle for sufficient angular discrimination.

The system shown in Figure 2 differs from conventional doppler tracking systems in which doppler tracking is performed by means of an oscillator situated at an intermediate frequency stage of the receiver. In the system of Figure 2 doppler tracking is performed using the first local oscillator 11, which enables the speed gate filters 16 and 17 to be placed at an early stage of the receiver 10. This means that the major portion of the receiver 10 has to deal with only a small range of frequencies, which simplifies the design of the receiver considerably.

The radar system as shown in Figure 2 has the disadvantage that when the phase-locked loop 60 has locked on to a single target signal the bandwidth of the signal output of the voltage controlled oscillator 62 is reduced by the output of integrator 63 so that only that single target is within the bandwidth, all other targets being outside the bandwidth. If the signal from the selected single target suddenly shifts to outside the restricted bandwidth of the phase-locked loop 60 then the signal input from the voltage controlled oscillator 62 to the phase-sensitive detector 34 disappears and the missile loses tracking. In order to overcome this the phase-locked loop of Figure 2 has been modified as shown in Figure 3.

In Figure 3 tne parts 61,62 and 63 of the original phase-locked loop are given the same references and the phase-sensitive detector 34 is also given the same reference number. The output of the phase-sensitive detector 34 is then used as in Figure 2 to control the guidance of the missile for example via phase-sensitive detectors such as 41 and 42. A variable gain circuit 300 is introduced into the phase-locked loop between the phase-sensitive detector 61 and the integrator 63. A quadrature phase-sensitive detector 301 is connected to compare inputs from the intermediate frequency sum input S1 and from afr12 phase shifting network 302 connected to the output of the voltage controlled oscillator 62. The output of the quadrature phase-sensitive detector 301 is connected via a low pass filter 303 and via a threshold circuit 304 to a "change to wide band" input of the variable gain circuit 300. A further "complementary" output of the threshold circuit 304 is connected to one input of a two input AND gate 305 the other input of which is obtained from the target acquisition circuitry (not shown). The target acquisition circuitry gives an indication of the acquisition of a target when for example as the frequency of the radar system is tracked over a range of frequencies a large signal-to-noise ratio is obtained. The output of AND gate 305 is fed to a "change to narrow band" input ol variable gain circuit 300.

The operation of the circuit is as follows:- As the missile is in flight with the radar guidance system "on sweep" looking for the targe:l, i.e. prior to acquisition of a target, the variable gain circa-.it 300 is set to a high gain corresponding to the wideband operCtion of the phaselocked loop. When target acquisition is obtafnied a signal is provided at one input of AND gate 305 from the targe.t acquisition circuitry. Due to the '1112 phase change circuit 302 Viie signals now presented at the input of the quadrature phase-sensitive detector 301 will be out of phase producing an output d.c. level which passes through filter and time delay circuit 303. This (d.c. level is arranged, for a valid target signal, to be in exces.s of the threshold level of circuit 304 thus producing an output sign&..',] on the "signal presenC output of threshold circuit 304. This signal combined with the target acquisition output signal in AND gate 305 p.roduces the "change to narrow band" signal for the variable gain circu-t 300. This signal causes the variable gain circuit 300 to effeCr, a narrowing of the bandwidth of the phase-locked loop and to there' lore lock the phase-locked loop to a single target signal.

If this single target signal should disappear either for examplie by virtue of the target disappearing or rapidly changing direc:t-ion, the output of the voltage controlled oscillator will be in randalm phase with the input to the loop phase-sensitive detector 61 (both signals being substantially noise) and the output of the quadracture phase-sensitive detector 301 will be zero. The "signal prese.mt" output to the AND gate305 will disapppear and an output will be given to the variable gain circuit 300 to change to wideband thus enabl-ing the phase-locked loop to search for a further target signal within ths wider bandwidth.

The changes from wideband to narrowband and from narrowband to wi,,deband are accomplished over different time periods as shown by Figures 3a and 3b. The change from wideband to a narrowband, Figur-e 3a, takes place over a relatively long period of time enabling the Tl,,):)op to lock to the single target signal. The change from narro.--wband to wideband, Figure 3b, takes place over a relatively short peria.d of time - almost instantaneously, since once the target signal 1 has been lost it is imperative to increase the bandwidth as rapidly as possible to search for a further target signal. The circuitry necessary for ensuring the correct changeover characteristics is included within the variable gain circuit 300.

For input signals with a high signal-to-noise ratio the time delay of circuit 303 is increased to for example 0.75 sec. and for low signal-tonoise ratio signals the time delay is decreased to 0.2 sec. This ensures that spurious input signals do not unnecessarily upset the bandwidth of the loop.

With reference to Figure 2 the sum signal to the phase-sensitive detector 34 is obtained from the output of the voltage controlled oscillator 62 and this can result, in the case of multiple targets, in the gain of the receiver being incorrect. This is because the phase-locked loop will lock to the frequency of one of the targets and therefore the spectral component in the output of the voltage controlled oscillator 62 will be of greater amplitude than the corresponding component in the original sum channel since the a.g.c. operates on the total sum signal passing through the receiver. Thus the angular scaling of the wanted target is reduced. This is to say that the automatic gain control system reduces the gain of both sum and difference channels to control the total output of all spectral components of the sum signal. Hence the amplitude of the single component being tracked by the phase-locked loop is reduced. In Figure 3 two alternative circuit arrangements are shown for overcoming this problem.

The first shown in dotted lines is to obtain the receiver automatic gain control signal from the output of the quadrature phase-sensitive detector 301.

The second shown in chain dotted lines comprises dividing the output of the phase-sensitive detector 34 with the output of the quadrature phasesensitive detector 301 in a divider 306 after passing the former through a low pass filter 307. In this way the angular error output from the difference channel is divided by a d.c. value which is proportional to the component of the sum signal which is in phase with the spectral component in the output of the voltage controlled oscillator 62. Thus by dividing the angular error by this d.c. value the scaling will be restored to its correct value. This method also has the advantage that it compensates for the low scaling due to the voltage controlled oscillator not being at the correct average phase with respect to the wanted target signal due to the pulling effect of the unwanted signal.

However, in addition to the DC term the quadrature phase-sensitive detector 301 also has an AC term due to the beat between the signal output of the voltage controlled oscillator 62 and the unwanted signal.

Consider two targets T1 and T2 within the speedgate bandwidth.

Doppler frequency = 10 fn Phase-locked loop damping = 0.7 Target ratio T2/T1 = 5 Phase-locked loop locked to the smaller signal T1 Instantaneous automatic gain control and no valid target lock. The curves will be different lor slow automatic gain control but the basic argument applies.

Figure 4 shows the voltage controlled oscillator phase at the Ti/2 output with respect to the phases of T1 and T2 (in the absence of T2 the voltage controlled oscillator r/2 output will be in-phase with T1). Figure 5 shows the S envelope and the quadrature phase-sensitive detector output. It can be seen that a difference signal D due to T1 will produce a DC term at the angular error phase-sensitive detector of approximately 1/5. cos 42'. D/S and a small AC term due to the voltage controlled oscillator having phase modulation with respect to T1 phase. A di'l"ference signal due to T2 will produce a large AC term due to the difference in frequency from the voltage controlled oscillator main line and a small DC term due to the voltage controlled oscillator phase modulation.

With the employment of analogue divider 306 the AC output due to T2 is divided by the AC term in the quadrature phase-sensitive detector 301 output to produce a potentially large DC term. For the particular conditions taken the two AC terns are almost exactly in-phase thus producing a large dish bias towards the second target. If, however, the quadrature phase-sensitive detector 301 output is filtered by a low pass filter 308 the AC wll be attenuated and the phase lagged. If this lag were 90' then the DC bias would be zero. If the lag were greater than 90', i.e. a multipole filter, then the dish would point away from the second target.

The action of the receiver automatic gain control is similar although the effect is reversed for the conditions considered. When the sum signal S envelope is at its greatest the difference channel gain is reduced assuming fast automatic gain control but the phase-locked loop signal S' is constant, thus the scaling is reduced, similarly the gain is increased when the sum signal S envelope is low. This will cause the AC beat due to a difference (D) signal in-phase with the S component at W2 to have its positive peaks (corresponding to T2 in-phase with voltage controlled oscillator) to be reduced and the negative peaks to be increased giving a negative output whereas T2 alone would have given a positive scaling (D and S in-phase) thus the dish points away from T2. The effect of slow automatic gain control is to delay the phase of the automatic gain control sinewave by 90' producing a positive bias, i.e. towards T2 as the unfiltered S envelope leads the angular phase-sensitive detector beat by 420. For the conditions considered the automatic gain control effect will be much smaller than the analogue divider effect due to the relative amplitude of the AC terms involved but this may not always be true.

The filter in the quadrature phase-sensitive detector output must not have a time constant much longer than 100 ms in order to preserve the receiver ability to react to changes in relative target amplitude.

A further modification to the circuit of Figure 2 which may be used to reduce the variation in scaling which occurs when the multiple target discriminator or narrow band phase-locked loop is in operation is shown in Figure 6, in which constant amplitude sidebands are added to the multiple target discriminator voltage controlled oscillator output which have the same phase relationship to the multiple target discriminator output as the original sum signal sidebands have to their carrier. This results in a change of scaling of 1.15 to 0.9 for a carrier locked multiple target discriminator and 1.02 to 0.64 for a first sideband locked multiple target discriminator as against 1.0 to 0.0 and 0. 0 to 0.64 for the multiple target discriminator alone.

The block diagram is shown in Figure 6 to which reference is now made. The circuit of Figure 6 produces a modified multiple target discriminator sum signal to be used in place of the output signal from the voltage controlled oscillator 62. Parts performing the same or similar functions to those in Figure 3 are given the same relerence numerals.

A balanced mixer 600 produces the first upper and lower sideband by mixing the multiple target discriminator voltage controlled oscillator output with the output of a second voltage controlled oscillator 601 at the nutation frequency, this voltage controlled oscillator is locked to the first upper and lower sideband of the sum signal by the second phaselocked loop comprising mixer 600, voltage controlled oscillator 601, an integrator 602 and a linear multiplier 603. A proportion of the output from the balanced mixer 600 is added to the oscillator 62 output in a combination circuit 604 to produce a composite intermediate frequency sum signal (modified multiple target discriminator sum signal) with pseudo first upper and lower sidebands which have a constant amplitude.

As the nutation voltage controlled oscillator 601 does not have to track the sum signal rates its bandwidth can be made small so as to reject sidebands from other target sum signals.

So far it has been assumed that the multiple target discriminator including voltage controlled oscillator 62 is locked to the main line of the sum signal but it can easily be locked to a sideband, in this case the balanced mixer 600 output will produce lines either side of that sideband.

AnalysiS of the method.

Let the input signal be ASin C (w-p)t +] +Sin (wt) +BSin E (w+p)t +Y] The 500 kHz voltage controlled oscillator output will then be - Cos (wt) and the 64 Hz voltage controlled oscillator Cos E pt + 3 then the balanced mixer output is 4COS 1 (W-P)t - 1 1 + Cos 1 (W+P)t + 1 1} The nutation phase-locked loop phase-sensitive detector output d.c.

terms are which must be zero.

Therefore 114 { A Sin [+ j 1 + B Sin E V- 1 1} Cos (f) { A Sin () + B Sin (Y) + Sin (j) A Cos () B Cos (V)} = 0 1 = Tan-' B Sin (V) + A Sin ().................... (l) B Cos (y) - A Cos () The output from the balanced Mixer is then Cos Ept+fl Cos M) = 1 Cos E (W-P)t-f 1 + Cos E (W+P)t + j] The new multiple target discriminator sum is then - K Cos E (w-p)t - 11 + Cos (wt) + K Cos [ (w+p)t + f] (2) The main phase-sensitive detector of the receiver must be changed to a multiplier as li!iiiting of either the input or the new multiple target discriminator signal will remove the sidebands. (This statement assumes that the nutation modulation is either zero or w radians).

The new sum signal must be phase shifted as before, before being multiplied with the difference signal of - - H{A Sin [ (w-p)t +1 + C Sin (wt) + B Sin E (w+p)t -YI j The resultant output is E {AK Cos (w+ + C + BK Cos (W-J) (3) With the multiple target discriminator phase-locked loop locked to a single target the difference signal assuming that the nutation modulation switches between o and is E 1) + 4 Sin Cos (mqt + m6) Sin (w-t) (4) :FC m 2 whe re 0 is the angle over which the nutation phase is zero 5 is the phase angle of the nutation waveform m the nutation harmonic number.

From (3) this gives a receiver output of E 2 Sin (g) Cos (A + j) + ( - 1 Tr 2 Tt + K2 Sin g Cos (S + TC 2 From (1) 1 = -4 Also the phase of the multiple target discriminator voltage controlled oscillator 62 will change byTt radians when the sum carrier changes by Te radians. This also results in the nutation voltage controlled oscillator 601 changing by Tr radians to remain locked to the first upper and lower sidebands which results in no phase change of the pseudo sidebands although there will be a transit phase shift as the nutation voltage controlled oscillator changes phase. This gives the receiver output as - E {K 4 Sin (g 9) + ( - - 0 - 1) T1 Tr 2 Tr ......# 000 o.o (5) In Figure 7 the receiver scaling is shown plotted against for K 1/ J2_.

With the multiple target discriminator phase-locked loop locked to a first sideband E { K ( - 1) Cos (6+f) + 2Sin W+K 1 Sin ()Cos Tr rr Q):r From (1) In this case the phase of this multiple target discriminator voltage controlled oscillator 62 does not change phase but as both the lines either side of the line the multiple target discriminator is locked to change phase by ITdegrees the receiver output becomes - EfK ( - 1) + 2 Sin + K 1 (Sin (6) TC TT:R In Figure 8 the receiver scaling is shown plotted against jK for K = llF2.

The nutation phase modulation was taken as switching between o and Tcradians. This results in equal amplitude sidebands either side of the carrier with equal and opposite phase angles. If the amplitude-of the two lines either side of the multiple target discriminator voltage controlled oscillator 62 are not equal but have equal and opposite phase angles the pseudo sidebands will have the right phase angles for demodulating the difference signal (i.e. the case of the first sideband locked multiple target discriminator and an improvement will be obtained, but if these two lines have equal amplitudes and phase angles the pseudo sidebands will be TC/2 radians different from the above result and no improvement will result.

In general if the nutation modulation is not what was assun--d the improvement in scaling will be less than that shown but should not be worse than for the multiple target discriminator alone. The range of the voltage controlled oscillator 601 should be as small as possible and in a typical example a range of from 61 to 67 Hz is chosen to cover the possible nutation frequency range. This will enable the loop to pull into lock quickly with a small enough bandwidth to reject other target lines produced from multiple targets. As it is required to track the two nutation lines either side of the multiple target discriminator locked line a sinusoidal output is desirable to avoid other nutation or target lines influencing the voltage controlled oscillator phase angle.

If pseudo sidebands are added to the output of the multiple target discriminator voltage controlled oscillator 62 within the multiple target discriminator phase-locked loop, the change of natural frequency and damping with nutation duty ratio is reduced resulting in a reduction in the number of times the multiple target discriminator transferred lock to other target sum signals which have frequency lines between the locked sum line and the lines either side of it. In Figure 9 the block diagram of such a system is shown. During the locking process it may be preferable to switch out the pseudo sidebands on the multiple target discriminator voltage controlled osci 11 ator.

An improvement in scaling changes with nutation duty ratio can thus be obtained by the addition of a second phase-locked loop at the nutation frequency together with a balanced mixer. The variation in scaling with the proposed system assuming the nutation modulation switches the phase of the signal between o and Ir radians will be with a K value of 11,r2; 1. 15 to 0.9 for a carrier locked multiple target discriminator and 1.02 to 0.64 for a first sideband locked multiple target discriminator. These figures without this system will be 1.0 to 0.0 and 0 to 0.64 and if the best scaling is obtained by picking the best multiple target discriminator lock condition that is carrier or first sidebands lock 1.0 to 0.54.

The pseudo sidebands can as described be added to the multiple target discriminator voltage controlled oscillator within the multiple target discriminator phase-locked loop resulting in a reduction in the variaton in loop bandwidth with nutation duty ratio and a possible reduction in the number of changes of tracked target in a multiple target situation.

A further problem which can arise with the system of Figure 2 is that low frequency beats can occur between harmonics of the multiplexing frequency and multiples of the doppler difference frequency between two targets and also multiples of the nutation frequency. The resulting low frequency disturbances produce a spurious acceleraton demand and degrade the performance of the multiplexed receiver. The circuit of Figure 10 has been designed to detect the low frequency beats and switch the multiplexing frequency to avoid coincidence with the doppler difference and nutation frequency.

Referring now to Figure 10, parts which perform the same or a similar function to the system of Figures 2 and 3 are given the same reference numerals. The output of the multiple target discriminator quadrature phase-sensitive detector 301, which has frequency components equal to the doppler difference and nutation if present, is fed to a further phase sensitive detector 1001 via a filter 1002, the other input to the phasesensitive detector 1001 being the multiplexing waveform. When multiples of these signals frequencies approach coincidence the low frequency disturbance passes through a filter 1003 and switches the multiplexing frequency from for example 40 Hz to 35 or 45 Hz to avoid coincidence. The switching earths the filter output for a short period after the bistable has changed state to prevent the over-swing of the disturbance switching the multiplexing frequency back into coincidence.

As an alternative to the arrangement of Figure 10, it is possible to use the detected sum signal input as the input to the phase-sensitive detector 1001 for comparing with the muliplexing signal as shown in Figure 11. The output of the automatic gain control detector 44 (see Figure 2), which detector is shown as connected directly to the receiver sum signal, is connected to filter 1002. A threshold detector 1004 is connected to the output of filter 1003 and thus the multiplexing frequency is only changed when the output of the filter exceeds the threshold level.

A further problem which arises with the use of the narrow band phaselocked loop as described with reference to Figure 2 above is that the narrow bandwidth is of the order of 20 - 30 Hz and it is found that the spectrum of a target return may widen up to 80 Hz if the target pulls a sudden high acceleration laterally (due to its glint spectrum widening). A constant bandwidth narrow bandwidth phase-locked loop tends to have a reduced tracking ability on such a target and for acceleration above a certain value this can cause a loss of lock. This in turn causes an interruption in guidance and can result in a large miss distance or poor target discrimination. The network of Figure 12 detects the widening of the target spectrum return and widens the narrow band phase-locked loop bandwidth appropriately for the duration of the transient condition thus preventing loss of lock.

Referring now to Figure 12, the parts of the network performing the same or similar functions to the system of Figures 2 and 3 are given the same reference numerals.

The sum signal is fed to the narrow band phase-locked loop via phasesensitive detectors 61 and 301 and also to a further adaptive phaselocked loop. The adaptive phase-locked loop comprises a phase-sensitive detector 1200, a variable gain circuit 1201, a loop integrator 1202 and a voltage controlled oscillator 1203.

c The sum signal is also fed to a quadrature phase-sensitive detector 1204 the other input of which is the output of voltage controlled oscillator 1203 phase shifted 11/2 by phase shifter 1205. The d.c. output of phasesensitive detector 1204 is compared in a comparator 1206 with a fixed reference bias. The output of comparator 1206 is used to control the gain of variable gain circuit 1201 and is fed as control input to a bistable threshold circuit 1207. The output of bistable circuit 1207 controls the bandwidth of the narrow band phase-locked loop by controlling the gain of variable gain circuit 300.

In operation the adaptive phase-locked loop adapts its bandwidth in proportion to the input signal spectrum width by alteration of the output level of comparator 1206 and hence alteration of the gain of variable gain circuit 1201. When the comparator output exceeds a given threshold set by the bistable threshold circuit 1207 the bistable element is set and the set output of the bistable alters the gain of variable gain circuit 300 by a predetermined amount. This predetermined amount widens the bandwidth of the narrow band phaselocked loop by an amount sufficient to enclose the spectrum of the target signal. In the above example the bandwidth is enlarged from 30 to 80 Hz.

A yet further problem which may arise with the use of a multiple target discriminating narrow band phase-locked loop is that it may not be able to track all types of target signals. The inventive solution to this problem is to latch out the multiple target discriminating narrow band phase-locked loop and also to recommence the doppler sweep (which of course is normally stopped after target acquisition has been obtained and the narrow band phase-locked loop is operative) at the correct frequency and in the required "direction" of frequency increase or decrease.

Referring now to Figure 13, the parts of the receiver of Figure 3 and Figure 2 which perform the same or similar functions are given the same reference numerals. In Figure 2 and Figure 13 therefore the output of the voltage controlled oscillator 62 is fed via a discriminator 45 to an integrator 46.A switching circuit 1300 is connected to the output of the discriminator, the output of switching circuit 1300 being connected to a first control input of a sweep generator 1301 the output of which is used to further control the voltage controlled oscillator 47. A futher control input to sweep generator 1301 is provided by the "change to wideband" output of threshold circuit 304 (Figure 3) which output is also used to set a latch 1302 the output of which is used to inhibit AND gate 305 preventing any further change to narrow band.

The operation of the circuit of Figure 13 explained with reference to Figure 14 is as follows:- The missile receiver circuitry after activation (usually by a time lapse arrangement after firing) commences a doppler sweep for the target. In Figure 14 the tim.e lapse is shown as from 0 to t, and the doppler sweep is coi-,i-.)enced at tl, The target is assumed to be acquired by the receiver circuitry at a time t2 at which time the doppler sweep is stopped at the target frequency. The normal pattern for-the doppler sweep is shown continued in dotted lines, the intermediate frequency of the receiver being swept over a range of frequencies and if no t-rget is found the sweep being recoffnenced from a starting frequency.

The target is now successfully tracked from a time t2 to a time t3 during which time period the multiple target discriminating narrow band phaselocked loop is activated. The narrow band phase-locked loop is assumed to lose track of the target at the time t3 and the target is assumed to have moved in frequency in an "increasing frequency" direction. The doppler sweep is commenced in this direction at the time t3 or soon thereafter and to sweep in an increasing frequency direction as shown between t3 and t4 where the target is again "acquired" by the receiver circuitry.

The direction in which the doppler sweep is recommenced is determined by the switch 1300 as follows:- When the target moves out of the narrow bandwidth of the narrow band phase-locked loop the discriminator 45 will be attempting to alter the froquency of voltage controlled oscillator 47 (see Figure 2) to compensate for this movemient. The discriminator 45 output will therefore be either a maximum positive or a maximum negative voltage depending on the direction of frequency movement of the target signal and this is detected by the switch 1300 which then switches to a position indicating in which the direction the frequency of the target was moving when the narrow band phase-locked loop lost lock. The output of the switch 1300 is then used to give a direction of sweep control signal to the sweep generator 1301 which sweep generator is activated by the change to wideband signal from the threshold circuit 304. In this embodiment the change to wideband signal is used to set the latch 1302 which provides an inhibit input to AND gate 305 (Figure 3) to prevent the subsequent operation of the narrow band phase-locked loop. Thus the missile will continue to track the target using the relatively wide bandwidth provided by the speed gate filters 16,17 (Figure 2).

Claims

1. A tracking radar system comprising an aerial arrangement having a plurality of outputs, means for deriving from the aerial outputs a sum signal representative of the sum of the aerial outputs and a difference signal representative of the difference of the aerial outputs, a receiver for processing said sum and difference signals to produce corresponding intermediate frequency sum and difference signals, means for comparing the intermediate frequency sum signal with the output of an oscillator in a phase-locked loop and using the resulting signal to control the oscillator frequency so as to cause the oscillator to lock on to the frequency of the intermediate f requency sum si gnal, a phase -sensiti ve detector for compari ng the intermediate frequency difference signal with the output of the oscillator to produce an output signal representative of the direction of the target relative to the aerial, and a quadrature phase-sensitive detector connected for comparison of the intermediate frequency sum signal and the '1r12 phase shifted output of the oscillator, the output of- the quadrature phase-sensitive detector being connected to the receiver for the provision of automatic gain control.

2. A tracking radar system comprising an aerial arrangement having a plurality of outputs, mans for deriving from the aerial outputs a sum signal representative of the sum of the aerial outputs and a difference signal representative of the difference of the aerial outputs, a receiver for processing said sum and difference signals to produce corresponding intermediate frequency sum and difference signals, means for comparing the intermediate frequency sum signal with the output of an oscillator in a phase-locked loop and using the resulting signal to control the oscillator frequency so as to cause the oscillator to lock on to the frequency of the intermediate frequency sum signal, a phase-sensitive detector arranged for comparing the intermediate frequency difference signal with the output of the oscillator to produce an output signal representative of the direction of a target relative to the aerial, a quadrature R phase-sensitive detector connected for comparison of the intermediate frequency sum signal and the 1-r12 phase shifted output of the oscillator, and a divider for division of the output of the phase-sensitive detector by the output of the quadrature phase-sensitive detector, the output of the divider being connected to the receiver for the provision of automatic gain control.

3. A tracking radar system as claimed in Claim 2, including a low pass filter interposed between the output of the quadrature phase-sensitive detector and the divider.

4. A tracking radar system as claimed in Claim 3, wherein the time constant of the low pass filter is greater than 100 ms.