GB2367438A - Monopulse radar for automotive cruise control - Google Patents

Abstract

A monopulse radar system comprises a processor 140 and a radar transceiver unit, the transceiver unit including a radar transmitter 102 for transmitting radar energy 112 towards an object 28 and radar receivers 104, 106 for receiving said reflected radar energy 130 from object 28. The processor 140 is arranged to receive signals 144,146 from the radar receivers 104,106 and to calculate therefrom at least an azimuth angle between the radar transceiver unit and said object 28. The radar transmitter 102 has a transmit directional gain pattern represented by a transmit lobe 114 extending along a transmit axis 116 that defines a zero azimuth angle, and the radar receivers 104,106 have receive directional gain patterns represented by two receive lobes 118,120 extending along corresponding receive axes 122,124. Each receive lobe 118,120 is asymmetric about its corresponding receive axis 122,124 and represents a peak in the receive directional gain pattern that falls off more gradually towards the transmit axis 116 than away from the transmit axis 116. The system provides a more efficient radar system for a vehicle automatic cruise-control (ACC), enabling preceding vehicles to be tracked around bends in the road.

Description

optimum Monopulse Radar for Automotive Intelligent Cruise Control Monopulse radar systems can take a variety of other forms, but the basic principle underlying their operation is the reception of target echoes using a radar receiver with directional sensitivity along two diverging axes. Usually, this is accomplished having at least two different receiving antennas, although it would be possible to use just one receiving antenna if this had directional sensitivity, as with a phased array. The radar transmission may be from one or more transmit antennas. All such systems are referred to herein simply as"radar monopulse systems".

The detection of a target is based on the presence of echoes of sufficient magnitude in the receive channels, and the angular position of a detected target is derived from the relative amplitudes, or the relative phases, of the signals in the channels. Monopulse radar systems are often used to measurement angles to objects that reflect the transmitted radar waves. Normally this is done by amplitude comparison of the returned radar waves in the two channels. The signals due to a single target in the two receive channels are used to provide a ratio of the two amplitudes, this being generally called the"monopulse ratio".

One possible use of monopulse radar may be in an adaptive, or an"intelligent", cruise control system for a motor vehicle. Such a radar system may be able to provide target

range information from the time delay of target echoes, with angle measurement being achieved by comparison of the echo amplitudes in two or more receiving channels. The inventors have evaluated a commercially available monopulse radar system in such an application, but found that it does not provide adequate performance. This conventional radar sensor uses two radar receivers, each having a symmetric directional gain pattern, and a single radar transmitter also with a symmetric directional gain profile, to provide range and azimuth angle information in a 100 sector (ISO) up to a maximum range of 135 m.

One of the inventors has performed calculations that simulate the performance needed for intelligent cruise control in a motor vehicle. This assumes the need to track preceding target vehicles around highway bends with a minimum radius of 100 m when the maximum vehicle lateral acceleration would be 0.25 g. These calculation have led to the discovery that azimuth angular coverage needs to be increased to approximately 20 10 . Reduced range performance can be accepted at the limits of the angular coverage. Commercially available low-cost monopulse radar systems cannot provide such coverage.

It is an object of the present invention to provide a more convenient monopulse radar system, particularly for use in a motor vehicle.

According to the invention, there is provided a monopulse radar system, comprising a processor and a radar

transceiver unit, the transceiver unit including a radar transmitter for transmitting radar energy towards an object, and a radar receiver for receiving said radar energy when reflected by said object, in which the processor is arranged to receive a signal from the radar receiver and to calculate therefrom at least an azimuth angle between the radar transceiver unit and said object, the radar transmitter has a transmit directional gain pattern represented by a transmit lobe extending along a transmit axis that defines a zero azimuth angle, the radar receiver has a receive directional gain pattern represented by two receive lobes extending along corresponding receive axes, one receive axis being at a positive azimuth angle and the other receive axis being at a negative azimuth angle, wherein each receive lobe is asymmetric about its corresponding receive axis and represents a peak in the receive directional gain pattern that falls off more gradually towards the transmit axis than away from the transmit axis.

In other words, the directional gain profile has a peak which is skewed away from the transmit axis. It have been found that such a receive directional gain profile can provide the performance necessary for use in a motor vehicle intelligent cruise control application.

In a preferred embodiment of the invention, the radar receiver is formed from two discrete radar sensors, each of which has a distinct receive directional gain pattern.

The radar sensors may be adjacent the radar transmitter, for example being positioned to the left and right of the

radar transmitter.

The receive directional gain pattern for each radar sensor is preferably essentially triangular in shape. Usually the shape will be such that the directional gain patterns for the two radar sensors are mirror images of each other about the transmit axis.

9. A monopulse radar system as claimed in any preceding Claim, in which the processor is arranged to receive a signal from the radar receiver and to calculate therefrom a range and a range rate between the radar transceiver unit and said object.

10. A motor vehicle with a monopulse radar system for detecting at least an azimuth angle between the motor vehicle and an external object, wherein the monopulse radar system is as claimed in any preceding claim.

Increased angular coverage for the schemes discussed in this proposal is provided by the use of carefully designed radar receiver directional gain profiles in a two-channel monopulse receiving system with a suitably shaped transmit directional gain profile. It is shown that with detection range and angle error specified as functions of direction, there is an optimum combination of directional gain patterns which ensures that the minimum transmitter source power is required.

The invention will now be described by way of example, with reference to the accompanying drawings, in which :

Figure 1 is a schematic drawing of a motor vehicle with a prior art forward directed monopulse radar system, having one radar transmitter and two spaced apart radar receivers; Figure 2 is a plot of a graph showing transmit and receive directional gains against angle for the prior art monopulse radar system of Figure 1; Figure 3 is a plot of a graph showing a monopulse ratio against angle for the prior art monopulse radar system of Figure 1; Figure 4 is a schematic drawing of a motor vehicle with a forward directed monopulse radar system according to the invention, having one radar transmitter and two spaced apart radar receivers ; Figure 5 is a plot of a graph showing left and right receive directional gains against angle for the monopulse radar system of Figure 4; Figure 6 is a plot of a graph showing angle error against angle for the monopulse radar system of Figure 4; Figure 7 is a plot of a graph showing a measure of radar transmit power against angle for the monopulse radar system of Figure 4;

Figure 8 is a plot of a graph showing transmit and receive directional gains against angle for the monopulse radar system of Figure 4; Figure 9 is a plot of a graph showing transmit and receive directional gains against angle for a prior art monopulse radar system similar to that of Figure 1; Figure 10 is a plot of a graph showing the detection range against angle for the prior art monopulse radar system of Figure 9, and a specified range for a desired monopulse radar system; and Figure 11 is a plot of a graph showing the angle error against angle for the prior art monopulse radar system of Figure 9, and a specified angle error for a desired monopulse radar system.

Figure 1 shows a schematic drawing of a motor vehicle 1, referred to herein as an"own"vehicle, with a forward directed prior art monopulse radar system. The radar system has one radar transmitter 2 and two spaced apart radar receivers 4,6, that together form a monopulse radar unit 5. The monopulse radar unit 5 is housed within a front bumper 8 of the own vehicle 1. The transmitter 2 is arranged to project 10 a beam of radar waves 12. The projected beam 12 can be represented by a transmit lobe 14 centered on a longitudinal transmit axis 16 of the own vehicle 1.

One radar receiver 4 is to the right of the transmitter 2, and the other radar receiver 6 is to the left of the transmitter. These receivers 4,6 each have a detection sensitivity that can represented by receive lobes 18,20 directed along corresponding receive axes 22,24 to the right and left of the own vehicle axis 16.

When the transmitted radar waves 12 are reflected 26 off a preceding vehicle 28 in front, reflected radar waves 30 will enter 32,34 the right radar receiver 4 and left radar receiver 6.

In Figure 1 the own vehicle 1 is moving forwards at a velocity (Vo) 35 centered within a lane defined by a lane marker line 36 and a road edge 38, and the preceding vehicle 28 is moving forwards with a different velocity (Vp) 40 while straddling the lane marker line 36 to the left of the own vehicle axis 16. The radar reflections 26 will therefore be returned preferentially to the left radar receiver 6. This is because the reflected radar waves 30 reach the left receiver 6 along a direction close to the axis 24 of the left receiver sensitivity lobe 20, while for the right receiver, the return direction will be at a greater angle to the directional axis 22 of the right receiver 4.

The radar transmitter 2 and receivers 4,6 are connected 42,44, 46 to a microprocessor-based radar control unit (C) 40 that controls the production of the transmitted radar waves 12, and which receives signals from the receivers 4,6. Differences in the intensity of received signals

enables the control unit 40 to calculate the angle of returned radar waves 30, and hence angular location of the preceding vehicle 28. The transmitted radar waves 12 are pulsed on and off, and the time of arrival of the returned radar waves 30 enables the control unit 40 to determine the distance to the preceding vehicle. Finally, a Doppler shift of the returned radar waves 30 compared with the transmitted radar waves enables the control unit 40 to determine the preceding vehicle speed 40 as compared with the own vehicle speed 35.

The radar control unit 40 may provide outputs of the measured parameters for various purposes. In Figure 1, the control unit 40 provides an output 48 to an intelligent cruise control unit (ICC) 50 that control the own vehicle speed to maintain a minimum safe following distance to the preceding vehicle 28.

When following the preceding vehicle 28 in the same lane of a highway around bends with minimum radius of 100 m, the detection range required directly ahead was calculated to be 134 m for a speed of 161 km/h (100 mph). A radar sensor angular coverage t 100 with a range of 40 m is needed to enable a vehicle to be followed on tight bends at speeds of 64.4 km/h (40 mph) when the vehicle lateral acceleration is taken to be limited by occupant comfort to a maximum of 0.25 g.

Figure 2 shows the directional gain in arbitrary units against angle a as measured from the vehicle axis 16. As shown in Figure 2, the receive directional gains 64,66 for

respectively the right and left radar receivers 4, 6 are about 110 wide. The receive gain curves 64, 66 have peaks 68,70 that are spaced by approximately 10'so that these cross-over on the vehicle axis 16 at just over one-half of the peak level. The transmission directional gain 62 is approximately 150 wide. Each of the receive directional gains 64,66 is symmetrical about its peak 68,70, with a shape that could be formed by a simple horn antenna.

Figure 3 shows a curve 72 for the monopulse ratio p (e) = {R (6)/L (e)} against angle e for the echo amplitudes received from a single target at direction 0. It can be seen that the ratio of the echo amplitudes will give unambiguous indication of direction over a sector bounded by the receive directional gain centres 68,70. It is apparent that a slightly greater sector can be covered before the ratio starts to give ambiguous indications but the regions outside the receive directional gain centres 68,70 are where one of the radar receivers 4,6 has very low response to returned radar waves 30, and this will result in a poor accuracy for angle measurement. Therefore the receive directional gain centres 68,70 normally represent the approximate limits of the angular coverage.

In this case 50% of the sensitivity of each radar receiver 4,6 is unused.

As Figure 3 shows, it is not possible to increase the angular coverage of the present conventional system beyond approximately 140 (I 70) due to the ambiguity in the monopulse ratio. To double the coverage would require a

change in the width of the directional gain patterns and an increase in transmitted power by a factor of 16. This is beyond the capability of current low cost millimetre wave sources and would produce a power density of 8 mW/cm2 at the centre of the transmitter aperture.

It is possible to use alternative schemes which use relative phase information, or the sum and difference of the receive channel signals followed by complex ratio processing, to extract angle indications from the received echoes but when system noise is present the angular accuracy of the alternative schemes, and the sector covered, are virtually the same as for the prior art amplitude ratio method.

The prior art monopulse radar unit 5 has been evaluated on the road, and is capable of providing the required detection range. However, the field of view in which angle measurement can be made is limited to ISo. This is considered to be too narrow for use with many intelligent cruise control applications.

This invention presents a method for increasing the angular coverage of a monopulse radar, with particular emphasis on the practical requirements and constraints imposed by the available transmitter power, radar detection range performance and angular measurement accuracy with this increased field of view.

The design goals for an improved radar monopulse system

were selected as follows :

Table 1 :

Operating parameter Value Comments Wavelength 4 mm 76 GHz frequency Overall loss ratio 4 Processing bandwidth 400 Hz Receiver noise figure 20 dB Maximum range 140 m Target cross-section 1 m2 Transmitter power 10 mW For 100 coverage Mixer drive 2 mW each Transmit antenna loss 0.8 dB Receive antenna loss 0. 8 dB Radiated power 3 mW For 100 coverage Elevation beamwidth 40 12.7 Xls aperture Probability of detection 0.9 Rayleigh fluctuating target False alarm probability 10-6 Polarisation Vertical Figure 4 shows a motor vehicle 101 with a monopulse radar system according to the invention. Features in Figure 4 which are the same as those in Figure 1 are indicated by the same reference numerals, and features of the invention which correspond with those in the prior art motor vehicle 1 are indicated with similar reference numerals incremented by 100. The invention differs from the prior art in a number of ways, which will become apparent below.

In order to appreciate the nature of the invention, it is first necessary to develop expressions for the monopulse radar system as set out below.

The power of transmitted radar energy 112 from the monopulse radar transmitter 102 is taken to be PT and a target preceding vehicle 28 has an echoing area cr at a range r and angle 0. One may use the usual radar equation to obtain the echo powers SL and SR for reflected radar waves 130 received in the two receive channels of the monopulse radar receivers 104,105 as,

where is the operating wavelength and LR is the overall loss ratio for the radar system hence :

C = aX2/ (47tpL (3)

It is now assumed that detection requires that individual detections are obtained in each of the receiver channels and that the False Alarm Probability in each receiver channel is the same, and is written as FAP,. For a target 28 with Rayleigh fluctuation statistics it is straightforward to obtain a simple expression for the Probability of Detection (POD) in one channel in terms of

the received signal-to-noise ratio (SIN) and the result is,

I N POD, = (FAP,)'= (FAP,) (3)

Using equation (3) the overall Probability of Detection for a sequential two-channel monopulse system can be obtained in the form,

N N POD, = (FAPL) . (FAPR) SR+N (4)

For reliable detection performance it is necessary to operate the radar with high signal-to-noise ratio in both receive channels so it can be assumed that the noise in either channel can be neglected in comparison with the corresponding signal and then one has the approximate expression,

N N POD2 (FAPJ. (FAPR) (5)

Since the channel False Alarm Probabilities are the same FAPL=FAPR=FAP, and the overall False Alarm Probability is FAP2 = FAPI2 so equation (5) can be written,

ipL'f-1 PO (N FA FA ' (6) POD FAP. 2 s ; :-+s ; :- = FAP. 2"s ; : + SR (6) POD2 ~ FAP2 2 SL SR = FAP2 2 SL SR (6) POD2 Lp

Equations (1) and (2) may now be used to substitute for 51

and SR to give the final result for the Probability of Detection which is,

ri N (i i f-.-l POD, = FAP2 T, e, LLO, R (e, ( )

The monopulse radar system is likely to be required to provide a fixed Probability of Detection with a given False Alarm Probability for targets at specified ranges which depend on the target angle. For example in intelligent cruise control operation it is necessary to have a high detection range on boresight, i. e. directly ahead, but the detection range can be reduced as the magnitude of the angle from the boresight increases.

Therefore it is convenient to specify the range performance required as f to denote the likely need for different range performance in different directions.

With the above consideration in mind it can be seen from equation (7) that the provision of a required Probability of Detection POD2 with specified False Alarm Probability FAP2 for a target 28 at the limits of the coverage region requires that,

where D (O) represents the specified detection performance for the system.

To determine the accuracy of angle indications provided by the use of the amplitude ratio, it is assumed that in the absence of any system noise the monopulse ratio is p (B) and the inverse function from which the angle is obtained is (} ().

Assuming that the echo amplitudes in the channels are nonfluctuating with values SR and SL'and with noise voltages nR and nL, the actual monopulse ratio of the amplitudes when perturbed by noise will be,

Satisfactory detection and monopulse operation is only achieved if the noise voltages are small compared with the echo amplitudes so equation (9) may be written approximately as,

s : : [ (1 nR nL ] [ nR nL J p, =p+5p p (l +). (i-) pl+ L- (io) SR SL SR SL

and hence one concludes that,

SppL-k (11) SR SL

Since the indicated angle is obtained using the monopulse ratio p the error bp in the measurement will produce an

error in angle given by,

60 (12) . oej ao

Only the mean square error in indicated angle is of importance and on assuming that fiR and nL are uncorrelated and of the same power N, which is highly likely since the noises arise in separate but similar receivers, the result obtained is,

5er. p- 4.'- ±-') -7 2 ~ nL P2 N + N o 11 ffl p. [nR s 2 ~ [2. S] (13) LcQj s, s, p 2. S, 2. S,

where pop is a slope factor and the echo powers are m

Sand S. The factors of 2 multiplying the echo powers in equation (13) appear because the mean square for the envelope of each echo is twice the echo power.

From equation (13) one obtains the root-mean-square angle error as,

--. (2. S/N) + (2. S/N (14) Ps pr

Equation (14) shows that the root-mean-square angle error is determined by the slope factor , the monopulse ratio p, and the square root of the sum of the reciprocals of twice the channel signal-to-noise ratios. When the signalto-noise ratios are appreciably different the smaller one

effectively determines the accuracy of angle indication.

Now the monopulse ratio p = ;--) and therefore, L (e)

where R)-R () and L)-L () BO

are the slopes, respectively of the directional gain functions R (G) 118 and L (E)) 120. On substituting equations (1), (2) and (15) in equation (14) the expression for the root-mean-square angle error becomes,

Substituting equation (8) into the above equation gives:

On considering equation (8) it will be noted that specified detection performance for a target at the limits of the coverage region is only achieved by selecting T (O) 114, L (8) 120, R (#) 118 and the other radar parameters appropriately. Therefore the only factor in equation (16)

that can be varied to achieve a specified accuracy in angle measurement is the final term in square brackets. Fortunately the term in square brackets is not affected by the choice of T (O) 114 so the antenna patterns may be selected in the following manner.

Step 1 Assuming that the detection performance is required to be D (#) implies that the term in square brackets in equation (16) should vary with 0 in such a way that the required

variation of angle accuracy, which is written as (r, ) and set out in equation (17), is achieved or bettered.

It should be noted that (r, < f ?) may be regarded as just a function of 0 since r. is a function of 0.

Therefore, the R (O) 118, and L (#) 120 patterns are chosen to ensure that the variation of the term in square brackets makes the root-mean-square angle error less than or equal to that specified in equation (17). Step 2 With the shape of the right and left receive directional

gain patterns R (O) 118, and L (O) 120 chosen in Step 1, finally select the transmit directional gain pattern T (O) 114 and the remaining system parameters to ensure that the specified detection performance is actually achieved, using the modified version of equation (8) which follows.

Of course, the resulting antenna directional gain patterns 114,118, 120 must be physically realisable, and this will have to be considered for the specified detection and angle accuracy conditions. Furthermore the radar performance will only be optimised in the angular sector for which the performance is specified. Outside this sector all the antenna patterns should if possible have negligible directional gain, and as a result the receive directional gain profiles will no longer be symmetrical about their peaks. Example 1 A mathematical model has been developed for the detection range required for an intelligent cruise control system for use on typical curved highways. From this model a required range for a monopulse radar system can be specified by the empirical equation:

where re is in metres and 0 is in degrees.

For the purposes of this example it will be assumed that the root-mean-square angle error required is given by,

I =6. r (20)

The above equations give a root-mean-square error of about 0. 5 for a range of 136 m with an offset angle of 0. 5 and a root-mean-square error of approximately 0. 90 when the range is 45 m with an offset angle of 10 . Thus the lateral position of a preceding vehicle 28 will be determined with more precision as the range decreases with the target 28 at the limits of the coverage region.

The expressions introduced above can now be combined to indicate the accuracy of angle indications required as a function of offset angle 8, which is,

Adopting the two-step procedure described above, the first step requires that suitable receive beam directional gains 118,120 are found to closely satisfy equation (17).

Therefore, directional gain functions must be found to approximate the relationship,

where for a constant POD of 0. 9 and FAP2 of 10-6 one finds that D (8) = 1/131. 1 and hence the constant

a = 2 DO) = 0. 1747.

A trial and error approach has currently been adopted to find suitably shaped gain functions for angles between sector limits of-100 and +100 and Figure 5 shows the resulting ideal left and right directional gain patterns 164,166, which are almost triangular in shape with respect to a zero gain baseline along the zero gain horizontal axis.

Although maximum performance will be obtained if the optimised receive directional gain patterns 164,166 have no response outside the coverage area, the realisation of 'brick wall'cut off requires an infinite aperture so the design goal is to achieve the least wasted sensitivity outside the area of coverage with the minimum antenna aperture. In practice this means that the receive directional gain patterns shown in Figure 5 will have rounded corners, as shown by lines 165,167. These rounded patterns, however, are still essentially triangular in shape.

It should be noted that the gain patterns may be scaled by any positive factor without affecting the behaviour of the term in square brackets of equation (17). Scaling is essential if the functions found initially have average gains over the coverage region that make it impossible to satisfy the fundamental constraint which requires that the integral of gain over the entire solid angle around an

antenna cannot exceed unity, but scaling with the greatest possible factor is the appropriate approach.

Assuming that the solid angle of coverage required in the example under consideration is 20 (azimuth) x 40 (elevation) the average receive directional gains 164,166 cannot be greater than about 500. This arises because the

(ir 0. solid angle of the radar coverage is 20-4- (--) 2 = O. 002477T 180

and the total solid angle around any point is 4n which is 515 times greater. It is more realistic to have average receive directional gains 164,166 in the region of 250 to allow for those portions of the receive patterns which are outside the region of coverage. In order to approach an individual average value of 250 the gains shown in Figure 5 can be scaled by a factor of 12.5 and this factor will be included in the receive gain values henceforth.

The fact that the angle error in the optimised system will be less than that specified by equation (22) when the gain functions are as shown in Figure 5 is indicated by Figure 6. This shows that the optimised angle error 73 is less than a specified angle error 75 for the intelligent cruise control application.

As described above, the second step of the optimisation procedure is to determine the transmit directional gain, transmitter power and other system parameters needed to provide the detection performance required at the limits of the coverage region as defined by Table 1.

Re-arranging equation (18) gives the result for the gain pattern as,

and on substituting for re using equation (19), with (log FAP2/log POD2) replaced by 1/D (8) the equation obtained is,

It will be assumed that the radar is required to detect a target with an echoing area of cy m,

With an overall loss ratio = 4, and an operating wavelength of 4 mm ( 76 GHz) as defined in Table 1, the value for the constant C becomes C = 2-10-9.

The noise power in equation (24) must correspond to that for amplifiers of appropriate bandwidth B for the intelligent cruise control application, allowing for integration achieved by coherent Doppler processing, and this may be taken as 400 Hz. Typically the noise factor for a receiver operating at the frequency corresponding to the wavelength of 4 mm is around 100 (i. e. 20 dB) and consequently the noise power N

becomes, with T = 290 Kelvin,

N = 50 k T B = 16-10-17

With the use of the system constants just postulated equation (24) now gives the required product of transmitter power times the transmission antenna gain as the pattern which is shown in Figure 7.

The transmission pattern required falls to a very low level outside the 200 sector and therefore it is reasonable to assume that the average gain of the transmission antenna in

the 200 sector takes the realistic maximum value of 500, so the results shown for the product 77 of transmit power and optimised transmit gain in Figure 7 can be used to show that the transmitter power must not be less than,

PT = 11. 5 mW. Figure 8 shows the transmit and receive directional gain patterns 162,164, 166 for the optimised system, with individual average receive gains of 250, and 500 for transmit, over the sector of coverage, are shown in Decibels in Figure 8. It will be observed that the receive beams have an almost triangular shape and have a cross-over level which is nearly 10 dB below the peak levels. In this example, the cross-over should be at least 8 dB below the peak levels.

As can be seen from Figures 7 and 8, the directional gain pattern of the radar transmitter is centered and symmetric

along the transmit axis, and the transmit directional gain pattern substantially falls within I 2. 5 of the transmit axis. These cross-over and width features are very different from those which normally would be adopted for the prior art two-beam monopulse system described above, if it were intended to adapt this to give angle measurements within a sector of 20 .

The receive directional gain 174,166 should ideally be zero beyond I loo for maximum efficiency. This antenna configuration provides I 100 azimuth coverage, with 140 m range against a 1 m2 target using 11.5 mW of transmitter

power. power.

The relative performance of the monopulse radar systems according to the prior art and to the invention will now be considered.

Suitable receive and transmit directional gain patterns shapes will be assumed for the three antennas of the prior art radar monopulse system so that this may cover the same solid angle as the optimised system according to the invention. It will be assumed that the average gains for the three antennas of the conventional system are similar to those specified for the receive directional gains of the optimised system described in Example 1. The radar parameters selected for use in Example 1 will also be adopted, and in particular the transmitter power will remain as 11.5 mW.

Figure 9 shows right and left receive directional gains 265, 266, as well as a transmit directional gain 262 for the conventional antennas to be used in this comparison. It may be observed that the individual transmit and receive directional gains 262,264, 266 are symmetrical about their respective peaks 274,268, 270 and are each about 180 wide with the receive directional gain peaks 268,270 separated by approximately 15 .

The range performance now may be found by using the following modified version of equation (8),

and on inserting the desired monopulse radar parameter values set out in Table 1, and the directional gains for the conventional system as illustrated in Figure 9, the conventional detection range 80 is found to be as indicated in Figure 10. This drawing also includes for comparison the desired or specified detection range 82.

The conventional monopulse radar system does not achieve the desired range performance, but the optimised monopulse radar system described above in Example 1 does in fact achieve this performance.

Equation (16) is now used to determine the root-mean-square angle errors provided by the conventional system at the

specified ranges for angles in the sector. The results for the conventional angle error 84 are shown in Figure 11. This drawing also includes for comparison the desired or specified angle error 84.

The conventional monopulse radar system does not achieve the desired angle error performance, but the optimised monopulse radar system described above in Example 1 does in fact achieve this performance.

The comparison between the conventional monopulse radar system and the optimised monopulse system according to the invention shows that when the target is near to the axis of the radar transmitter 16, the conventional system fails to provide for an intelligent cruise control system the necessary range performance and root-mean-square angle error.

The performance on the axis of the radar transmitter is an important requirement for intelligent cruise control applications and the specified value of angle error can only be achieved by increasing the transmitter power to nearly 13.7 times the value needed by the optimum system. With this higher transmitter power the range on axis will become approximately 180 metres.

It can be stated that a conventional non-optimum monopulse radar system may be very deficient in performance in some aspects of operation, particularly in the efficient use of transmitter power, and unwanted superior performance appears in other aspects.

Optimum two beam monopulse radar performance in range, angle error and field of view coverage is possible with the limited transmitter power available from current millimetre-wave radar systems by careful shaping of the transmit and receive directional gain patterns using the methodology contained in this report.

In conclusion, the invention provides a method for trading off radar transmitter power, range performance, angular coverage and aperture size in a representative monopulse radar system in order to optimise performance and avoid over-performance in areas of modest interest. The conventional approach to the selection of antenna patterns for two-beam monopulse radar systems provides excess performance in range at the expense of angular coverage or other parameters. Analysis and simulation have shown that optimisation of the transmit and receive directional gain profiles is able to offer increased angular coverage whilst still meeting the required specifications for range, angular error and transmitter power.

Traditionally military monopulse radar systems have been employed for accurate angle tracking for fire control and missile tracking/guidance radar systems, and in these applications the choice of receive directional gain cross over at about the-3 dB points has been driven by the requirement for maximum angular accuracy on boresight.

This conventional design criterion may not be applicable to an automotive radar.

The solutions proposed here make use of amplitude and/or

phase weighting of the directional gain patterns. Phase weighting is a practical option as this can be achieved by the insertion of low cost dielectric lens in the radar path to the receivers. Amplitude weighting, other than the naturally occurring cosine distribution from a horn antenna, would be more difficult to implement and would be better avoided if possible unless a phased array antenna is adopted.

Any increase in the field of view of a radar, whilst maintaining range performance, will increase the potential number of targets that can be detected by the radar.

The use of a simple two-beam monopulse radar system may have advantages over multibeam staring arrays when the complexity of feeding multiple beams and the power requirements of multiple mixers is considered, switched multiple beams systems may reduce the power requirement but then introduce the additional complexity of interrupted target measurement time.

A significant point is that this is an'antenna only' method of system optimisation and therefore the required directional gain profiles could be implemented with antennas ranging from simple horns, quasi optic lens antennas, planar array antennas or full phased array antennas.

This invention provides a general methodology for optimising two-beam monopulse radar performance in respect of detection range, sector coverage, angular error and

transmitter power.

In particular, the invention described herein uses radar range requirements for a practical motor vehicle intelligent cruise control system, including a realistic specification for the root-mean-square error in azimuth angle measurement. It has been shown that for realistic range and angle error specifications, it is possible to define optimum directional gain patterns to achieve the necessary performance requirements within a minimum transmitter power. It should be emphasised that the solutions presented provide optimum performance that can be achieved only by modification to the antenna directional gain patterns, and a modest adjustment to transmitter power.

Aperture dimensions and field distributions across the apertures are the only features that can be adjusted to modify the directional gain patterns. The results presented here have been calculated using aperture distributions that are considered to be practically realisable and have been adjusted to provide near optimum directional gain patterns.

This invention also provides a convenient monopulse radar for a motor vehicle, with an increased field of view. This is achieved by selecting appropriate radar transmit and receive directional gain profiles, whilst retaining the basic performance parameters, with modest change to transmitter power, compared with prior art radar monopulse systems.

Claims (12)

Claims :

1. A monopulse radar system, comprising a processor and a radar transceiver unit, the transceiver unit including a radar transmitter for transmitting radar energy towards an object, and a radar receiver for receiving said radar energy when reflected by said object, in which the processor is arranged to receive a signal from the radar receiver and to calculate therefrom at least an azimuth angle between the radar transceiver unit and said object, the radar transmitter has a transmit directional gain pattern represented by a transmit lobe extending along a transmit axis that defines a zero azimuth angle, the radar receiver has a receive directional gain pattern represented by two receive lobes extending along corresponding receive axes, one receive axis being at a positive azimuth angle and the other receive axis being at a negative azimuth angle, wherein each receive lobe is asymmetric about its corresponding receive axis and represents a peak in the receive directional gain pattern that falls off more gradually towards the transmit axis than away from the transmit axis.

2. A monopulse radar system as claimed in Claim 1, in which the radar receiver is formed from two discrete radar sensors, each of which has a distinct receive directional gain pattern.

3. A monopulse radar system as claimed in Claim 2, in which the radar sensors are adjacent the radar transmitter.

4. A monopulse radar system as claimed in Claim 2 or Claim 3, in which the two radar sensors are positioned to the left and right of the radar transmitter.

5. A monopulse radar system as claimed in any of Claims 2 to 4, in which the receive directional gain pattern for each radar sensor is essentially triangular in shape.

6. A monopulse radar system as claimed in any of Claims 2 to 5, in which the directional gain patterns for the two radar sensors are mirror images of each other about the transmit axis.

7. A monopulse radar system as claimed in Claim 5 or Claim 6, in which the receive directional gain patterns for the two radar sensors cross over at least 8 dB below the peak levels of the directional gain patterns for the two radar sensors.

8. A monopulse radar system as claimed in Claim 7, in which the transmit directional gain pattern substantially falls within I 2. 5 of the transmit axis.

9. A monopulse radar system as claimed in any preceding Claim, in which the processor is arranged to receive a signal from the radar receiver and to calculate therefrom a range and a range rate between the radar transceiver unit and said object.

10. A motor vehicle with a monopulse radar system for

detecting at least an azimuth angle between the motor vehicle and an external object, wherein the monopulse radar system is as claimed in any preceding claim.

11. A monopulse radar system substantially as herein described, with reference to or as shown in Figures 4 to 11 of the accompanying drawings.

12. A motor vehicle with a monopulse radar system substantially as herein described, with reference to or as shown in Figures 4 to 11 of the accompanying drawings.