209 ROTRON zs [oe Bak mm Ses oe came ame aE | In Alas A Melee Dew Lo Pb oe igure G4 Types of gyrotron amplifier configurations (fom Ewell, 1981, Fig. 2-44, p. 79, reprinted by permission of MeGraw-Hil) le G Peak Power Levels from Cyclotron Masers Driven by Intense Relativistic Electron Beams Wavelength | iererave | “volte Diode fm) | pomercarwy | cryy | erent) 70 300 Bg T 20 30 Ey a0 oO ¥ oe oF z 08 iS H HAMMING WINDOW (see WEIGHTING). HEIGHT FINDER. Height finders are radars designed 10 ‘measure the elevation angle of targets in a surveillance sys- tem, permitting target altitude to be caleulated from measured range. The methods by which elevation angle and hence alti tude is determined include: (1) Assignment of a specialized radar that performs sec- tor scan in elevation for measurement in that coordinate, on targets designated by a 2D search radat. (Q) Seareh with a scanning-beam 3D radar, in which a narrow beam is scanned over a raster covering both azimuth and elevation and providing measurement of both angles, along with range, on detected targets. {G) Search with a stacked-beam 3D radar, in which multi- ple beams cover the elevation sector as the antenna scans in azimuth, providing monopulse measurement in elevation. (4) Measurement of multipath time delay on targets detected in a 2D search radar, such that target altitude may be calculated from known target range, radar antenna altitude, and multipath delay. (5) Measurement of the ranges at which a target passes height-finder, nodding-beam through multipath lobes of a 2D search radar antenna pattern, from which a constant target altitude may be calculated. (6) Measurement of relative amplitude or phase of target echoes in two antennas displaced in altitude, leading to a ‘monopulse estimate of elevation angle. Some types of height finders are described below. (See also RADAR, ‘3D.) DKB ‘The nodding-beam height finder is the specialized radar (1) in the basic height-finding methods. A 2D search radar main- tains surveillance over the volume of interest, detecting tar- ‘gets and measuring their ranges and azimuth angles. Upon detection and establishment of a track on a new target, a request is sent tothe height finder for elevation measurement. ‘The height finder slews its antenna to the azimuth designated by the search radar and performs a scan over an elevation sec- tor appropriate to the range designated by the 2D radar. Target echoes are displayed on a range-height indicator (RH), and elevation angle 6, is measured by an operator or an angle-gate circuit, Range R is also measured to an accuracy better than that provided by the search radar. The target height above the horizontal plane at the radar site is then calculated as, h, = Rsin 8, This height is corrected as necessary for site altitude, atmo- spheric refraction, and curvature of Earth to give target alti- tude above sea level, as needed for ground-controlled intercept of the target by a fighter aircraft. A typical nodding- ‘beam height finder is shown in Fig, H. The antenna is elon- gated in the vertical direction to provide a narrow elevation bbeam for accurate measurement, while the azimuth beam- ‘width is wide enough to accommodate errors in designation from the search radar. DKB Ref: Stolk (1990, p. 203. Figure HI A Russian S-band nodding beight-finder radar (rom Jane's Radar and Electronic Warfare Systems, 1993-94).height-finder, squinted-beam A squinted-beam height finder uses a modified monopulse beam pait, pointed far enough above the horizon to minimize ‘multipath reflection errors. The technique is similar to using off-axis monopulse tacking and offers reduction but not elimination of the twoublesome multipath errors on low-eleva- tion targets. Figure H2 compares conventional monopulse and squinted-beam antenna patterns. DKB oA wi Figure H2 Multipath effects on monopulse and squinted-beam height finders: (a) conventional ¥,A monopulse, (b) squinted- beam low-angle technique (from Skolnik, 1990, Fig. 20.12, 20.37, reprinted by permission of McGraw-Hill) ‘Time-difference height finding uses the multipath propaga- tion phenomenon to obtain additional target data needed for estimation of altitude. The time delay between the direct path and the specularly reflected multipath, over a flat earth, is a tien by Se where 2 is target range, @, is elevation angle, y is grazing angle at the surface, his radar antenna altitude, and his tar get altitude. Ifthe range, radar altitude, and one-way (receiv ing) multipath delay p/e are known, target altitude can be calculated. The calculation is more complicated over a spher= ical earth, but it can be performed easily with modem computers. "The multipath delay is very small for surface-based radars, so the technique is restricted in practice to airborne radars and those on high mountain sites, where the delay commonly exceeds the processed pulse width, When the sur= face is too rough to support specular reflection, a spread mul- tipath return is received, but the delay to its leading edge is sli useful for target height measurement. DKB Ret: Long (1992), p. 349 HOLOGRAM, HOLOGRAPHY, radar 210 A V-beam height finder, depicted in Fig. H3, uses two broad radar beams from a single radar scan in azimuth. One of the ‘beams is tilted in the elevation plane so as to produce an out- put whose delay, in the azimuth sean eyele, is proportional to elevation angle Figure H3 V-beam geometry (Grom Fink, 1989, p. 25.49, reprinted by permission of McGnaw-Hilb, ‘The elevation error is proportional to the difference between the two azimuth readings multiplied by the cotan- ‘gent of the tlt angle. (See also PATTERN, V-beam).PCH Ref: Fink (1989), p. 28.49 HETERODYNING is the process of the conversion of mod- ulated radio-frequency oscillations into modulated oscilla. tions of a lower intermediate frequency to make subsequent amplification more efficient. SZ Ref. Terman (1955), p. S68; Popov (1980), p. 8 A HIT is “a target echo from one single pulse.” Ref: IEEE (1990), p. 18, HOLE, radar. A radar hole is region with reduced coverage, ccaused by the extension of the radar range in other directions due to the effect of ducting. For example, air targets above a surface duct, within which the radar range against surface tar- gets is enhanced, might be missed, though they would be nor- ‘mally detected when there is no ducting, SAL Ref: Skolnik (1980, p 481 HOLOGRAM, HOLOGRAPHY, microwave. Radio-fre- quency (RF) holography is a method of restoration of the ‘wave front of a radio wave in which the indicator, the radio hologram, registers information both about the amplitude and about the phase of the field dispersed by the object. As a result of subsequent illumination of the radio hologram, the restoring wave forms an image of the object. The processes of | formation and restoration of images constitute vietual forward and reverse integral Fourier and Fresnel transforms depend- ing on the curvature of the front of the object and reference An RF hologram is a fixed picture of interference between the field scattered by the object (object wave) and a coherent reference wave. From the picture received, the image of the object is restored by the methods of RF hologea- phy. RE holograms are formed by two basic methods. The first is associated with the creation of a real aperture315 pulse, uncoded [simple] igure #25 Simple rectangular pulse (trom Bogush, 1989, Fig. 3.54, p. 193). inexpensive radars when signal generation and signal pro- cessing cost must be minimized, and when the requitements to have sufficient energy for detection and tolerable range accuracy can be met simultaneously. Sometimes simple pulses are called uncoded pulses or Class A waveforms. SAL Ref: Bogush (1989), p. 192. PULSE COMPRESSION is “the processing of a wideband, coded signal pulse, of initially long time duration and low- range resolution, to result in an output pulse of time duration corresponding to the reciprocal of the bandwidth and, hence, higher range resolution, and with approximately the same pulse energy.” In principle, the process of pulse compression is the by-effeet when the signal with intrapulse modulation is processed in the matched filter to maximize signal-to-noise ratio, Since the matched filter does not preserve the initial shape of the waveform at its output, but on the contrary, dis- torts it to obtain the benefit of superposition of the maxima of different harmonics (to get the highest possible signal-to- noise ratio), the resultant output waveform compresses in time in comparison with the input waveform by value of the pulse compression ratio (Fig. P24). This useful property of matched filter processing gives the benefit of radiating long pulses on transmit (and, hence obtaining efficient use of power capability), and simultaneously obtaining short pulses on receive (and, hence obtaining good range resolution) when employing pulse-compression waveforms. The side effect of pulse compression is appearance of range or time sidelobes that can mask nearby echoes, requiring the use of special ‘measures for their suppession, ‘There are two basic ways to implement intrapulse modu- lation: coding either the frequency or the phase of the trans- mitted pulses, resulting in fequency-coded or phase-coded waveforms, which are the basic types of waveforms used in modem radars. (See WAVEFORM, pulse-compression.) ‘The basic methods of implementation of pulse compression are analog and digital pulse-compression techniques. The main advantages of pulse compression, leading to wide usage of this technique in modern radar are the following: increased. detection capability inherent in long, high-energy pulses is combined with increased resolving capability inherent in shott pulses; generation of high peak power common in short- pulse systems ean be avoided and mote efficient use of aver= age power can be obtained without increasing pulse repetition pulse compression, analog frequency, which would result in decreasing the unambiguous range; since the long pulse is used on transmit, increased dop- pler resolution is possible; when coded waveforms ate used, 2 radar is less vulnerable to interference. The cost to be paid for these advantages is greater complexity relative to simple pulse transmissions, complexity of pulse-compression wave- form generation and processing, all increasing the cost of the radar system. SAL Ref: IEEE (1990, p28; Barton (1988), pp 220-230; Barton (199), pp. 72-731; Skolik (1980), pp. 420-434; Skolik (1990), pp 10.1- 10.39; Brookner (1988), pp. 143-148; Lewis (1986), pp. 7-116; Leonoy (1988), p62 bs és Figure P24 Pulse compression (a), and resolution (b) of pulse- ‘compression waveforms aftr processing in matched filter (after Leonov, 1988, Fig. 2.21, p. 62). Analog pulse compression involves the use of analog meth- ‘ods to generate and process pulse-compression waveforms, ‘The main techniques incorporate active devices, primarily oscillators (see OSCILLATOR, voltage controlled) and passive devices, primarily different delay lines for linear FM ‘waveforms. Passive devices can be divided into two general classes: ultrasonic devices (bulk-wave or surface-acoustic- wave (SAW) types) and electrical devices using the disper- sive characteristics of an electrical network. In ultrasonic devices the input electrical signal is tansformed into an acoustic wave propagating through the medium at sonic speed, and at the output the signal is converted back to an electrical waveform. Longer delays may be achieved than with purely electrical devices of comparable size, as the wave travels at sonic speed. The most popular technique for linear FM waveforms is the SAW delay line. In SAW technology the energy is concentrated in a surface wave, making it much ‘more efficient than bulk-wave devices where the wave propa- ‘gates through the crystal. The main limitation of bulk devices is the necessity to arrange the coupling between acoustic ‘medium and electrical signal, typically with transducers inserting high losses (currently interdigital transducers are considered to transform an electrical signal most efficiently to acoustic energy and vice versa). Electrical networks with dis- persive characteristics are typically electrical delay networks having a linear delay-versus-frequency characteristics. The ‘main characteristics of analog pulse compression devices for linear FM waveforms are given in Table P6.pulse compression, analog ‘Table P6 Characteristics of Passive Linear-FM Devices pulse compression, frequency-modulated [FM] 316 2 od Be | Tas ar |foMHe| ab Stel iy aja | mw | 6] » aes neu 3D jm | we) | Render taco |) |S | wm | om | 0 ey ne Wedge why ee | 2) | | moo | 200 2 ‘The main disadvantages of analog pulse compression are the instability of network parameters, resulting in poor repeat- ability of compressed waveforms, and larger size than digital circuits of compatable performance, SAL Ref: Skok (1990), pp. 1010-10415 Digital pulse compression involves use of digital technology for generating and processing pulse-compression waveforms. I has some distinct advantages over analog pulse-compres- sion technique. It is more flexible: it allows implementation of different types of pulse-compression waveforms (binary- and polyphase-code waveforms, linear and nonlinear FM waveforms or even other exotic waveforms at a designer's discretion); it can operate with signals of different durations; the same implementation can be used to handle multiple types of waveforms; and digital waveform generators are very sta- bile devices ‘A system for digital implementation of a binary-coded \waveform compression employing in-phase and quadrature channels (often called homodyne or zero IF) is shown in Fig. P25. In this system the phase of each transmitted binary clement is defined by the code generator and is equal to 0 ot 180° with respect to the local oscillator signal. The phase of the received signal is shifted in respect to the LO by an amount depending on the range and velocity of the target. Each correlator may consist of several correlators depending fon the number of quantization bits in the digitized signal and only one channel instead of two channels can be used that vill insert an average loss about 3 dB. A typical scheme that may be used to generate digital frequency-coded ot polyphase- coded waveforms is shown in the Fig. P26. ‘The development of the digital pulse-compression approach is closely tied with the general development of the digital technology and will develop with the further increase of memory units capability and computational speed. Nowa- days the main limitation is the upper limit of frequency avail- able for digital devices, so digital approach is restricted in bandwidth under about 100 MHz, One ofthe ways to increase the bandwidth limitation for linear FM waveforms is 10 use Figure P25 Digital pulse compression for binary-coded wave- form (ftom Skolnik, 1990, Fig. 10.12, p. 10.28, reprinted by permission of McGraw-Hill). pes fened fice Figure P26 Digital pulse compression for FM and polyphase- coded waveforms (afer Skolnik, 1990, Fig. 103, p. 10.7) frequency multiplication combined with stretch processing, ‘The popular digital technique for pulse compression is the fast Fourier transform. SAL Ref: Skolnik (1990), pp. 10.7-10.10, Weer (1987p. 138; Leonov (1988), pss Frequency-modulated [FM] pulse compression is one of the basic techniques for improving the range resolution of a radar while maintaining adequate duty factor and average power. The simplest approach in this technique is to use linear FM waveforms. As shown in Fig. P27, the cartier frequency is varied linearly during the pulse, broadening the transmitted spectrum. Upon reception, a matched filter produces the out put of Fig. P27(e), with a narrow pulse surrounded by time sidelobes. ‘When using a complex signal with linear frequency mod- ulation (chirp), the carrier frequency change during the pulse ty is linear at rate Aft, for a frequency deviation Af (Fig. P27e). The instantaneous value of signal frequency is determined by fo) = est where fy is the central frequency of the signal. Ifthe product 7)A/is large, the increase of the cartier fre- quency after the linear sweep (Fig. P27¢) will yield an almost rectangular spectrum (Fig. P27b). The dependency of delay f,317 pulse compression, frequency-modulated [FM] in time on filter frequency is linearly decreasing function of frequency (Fig. P27d). The linear characteristic of delay in time provides large delay of low-frequency components at the beginning of the pulse in comparison with high-frequency components at its end. ‘All spectral components are delayed in an optimum filter as well as during digital processing by the amount of time necessary to arrive simultaneously at its output. Having the same zero phase, they add to form a peak signal blip (Fig. P27e). This also explains the increase of signal ampli- tude after passing through the filter and the compression in \width. The output pulse at a level ~4 dB below the maximum, hhas width +, = /A/. The ratio of pulse width at the input z) to \width at the output ¢, for an optimum filter is called the pulse compression ratio, ke t/t, = TA. ry © o © Figure P27 Compression of a linear-FM (chigp) signal (after Fal kovich, 1989, Fig. 6.13 and 6.14, page 245). A disadvantage of pulse compression is the appearance of sidelobes in the compressed signal that can lower the radar resolution. To control this, special measures are taken, To reduce sidelobes one can use a filter with smooth but sharply descending amplitude-frequency characteristic (for eg. a Gaussian response) in place of an ideal filter. This allows one to reduce the sidelobe level considerably with only slight \widening of the mainlobe. “Time sidelobes may be reduced by applying weighting in the compression filter, causing a weighting loss but greatly reducing the sidelobe levels, The characteristics of various amplitude weighting functions are shown in Table P7 (see also WEIGHTING’). Greater attenuation of the sidelobes can be obtained by using special weighting processing. Its essence consists of selecting the pattem of amplitude variation (weighting func tion) of the amplitude-frequency characteristic of the filter, through which the signal passes. As a result one can distort the signal spectrum in such a manner it approaches a signal spectrum that has low sidelobe level. Another approach is to pulse compression, phase-coded use a nonlinear FM waveform, in which the transmitted spec trum is weighted, providing a low-sidelobe response with a matched filter AIL Ref: Cook (1967) Chs. 6,7: Ric zek (1965), C. 7 Phase-coded pulse compression uses phase-coded wave- forms. With pseudorandom phase-coding having m subpulses, the ratio of the mainlobe power over the average sidelobes is approximately 1». Therefore, for large n, on the order of a thousand, sidelobes are small and measures to correct them are not called for, Frequency signal compression is carried out in a correla- tor. To compress a complex signal in frequency itis necessary to compensate for signal phase. To do this one must supply a copy of the signal to the correlator synchronously and in phase. One thereby removes intrapulse modulation and the spectral width of a signal with duration t, will equal f, = 1/4 following compression. Thus, the correlator performs spec~ ‘rum compression ‘To the same degree that time compression improves range resolution, frequency compression improves velocity resolution. By using the combined correlation-filter method ‘of complex signal processing, one can accomplish compres- sion in both time and frequency. ‘The compression ratio of a phase-keyed signal Kye fly is equal to the number of increments 1 (in the example of Fig. P28, » = 7). Thus, compression of phase-keyed signals in time sub- stantially improves radar range resolution for a given average transmitted power. Range resolution is determined by dura- tion of the compressed pulse t, and duration of the com- pressed pulse will depend on specteal width of the complex signal, For a phase-coded signal the spectral width is deter ‘mined by the equation f,~ 1/to Radar sets using signals that are phase keyed radiate a long pulse t,, which consists of several short pulses £, with identical duration and frequency, but which differ in phase, by values 0 or 180° (Fig. P28a). In this figure we show the typi cal patter of a phase-keyed signal, the phase-keying of which is accomplished in accordance with a 7-clement (1 = 7) Barker code. ‘Table P7 Characteristics of Pulse Compression Weighting Functions Weighting Max. | Relative | Change of function sidelobe | width of | SiN ratio, level, dB | mainlobe | dB Without correction [132 10 0 Hamming 428 1S 4 Cosine-square 314 159 1.76 Cosine-cube 39.1 Lal 28 Daliph-Chebyshev 40 Lalpulse compression, phase-coded Figure P28 Compression ofa phase-coded signal Biphase waveforms (phase-coded by 180°) using binary pseudorandom sequences have been most popular. As with a chirp pulse, a phase-coded signal is compressed by using a matched filter (Fig. P28c). Itconsists ofa delay line with taps, phase inverters (r), adder (2), and filter (F). In Fig. P28d we have shown pulses which come from taps of the delay line to the adder. The result of summation is shown in Fig. P28e, and in Fig. P28f we obtain an envelope of a signal s, atthe filter ‘output F, that is compressed relative to 7 by the factor m, AIL Ret: Cook (1967), Ch 8; Popov (1980), p. 10S; Sosuli (1992), p40. ‘The pulse-compression ratio is the ratio of transmitted pulsewidth to the pulsewidth after the process of pulse com- pression (at the output of the matched filter). An altemative term used is time-bandhvidth product, although this term is, also applicable to uncompressed waveforms (e.g., pulse trains). SAL. Ref: Barton (1988), p. 221, PULSE REPETITION FREQUENCY (PRE). The pulse repetition frequency “is the number of pulses per unit of time, usually per second.” The choice of proper PRF is very impor tant in radar design, There are three classic cases of PRES are distinguished in radar: Low PRE (LPRE) gives unambiguous range measurements forall the targets of interest forthe given radar; medium PRF (MPRF) is too large for unambiguous ‘measurement of all targets of interest and too small for unam- biguous doppler measurement of these targets, so the expected targets are ambiguous both in range and in doppler. High PRF (HPRF) is high enough to obtain unambiguous doppler measurement (and, in airborne radar, to obtain a clear doppler region for detection of approaching targets). LPRF gives unambiguous range measurement with a single PRF and permits use sensitivity time control (STC) to reduce the dynamic range requirements for the receiver. However, the average power may be low unless pulse compression is used to meet the combined needs of good range resolution and high average power (with long pulses). For HPRF the mea- surements are ambiguous in range, so typically several differ- pulse repetition frequency, high (HPRF) 318 cent PRFs are required to eliminate range ambiguity. Eclipsing losses are appreciable, and sensitivity time control cannot be used, resulting in requirements for the receiver to handle a ‘wide range of amplitudes. The advantages are that doppler ‘measurement is unambiguous (so blind speeds do not occur from ambiguous target doppler within the main-beam clutter notch), and long transmitter pulses are not required to obtain high average power. The MPRE to large extent is a compro ‘mise mode between LPRF and HPRE, so it tends to share both advantages and disadvantages of these modes. Since the requirements to the proper choice of optimal PRF are contradictory, a useful mode of radar operation is staggered PRF (ie., the mode when several different PRES fare used in a definite sequence). Sometimes this mode is called multiple PRFs and when the interpulse interval varies in a random manner it is termed PRF jitter. Pulse tains employing staggered PRF are termed staggered pulse trains ‘The PRF may be switched on a pulse-to-pulse basis, every other sean, or every time the antenna is scanned a half beam- ‘width, Such a mode may be used to eliminate range ambigu- ity (Fig. P29), to improve the characteristics of blind speed cancellation in MTI radars, or to provide enhanced ECCM capability against jamming. SAL Ref: IEEE (1990), pp. 23, 20,2118; Johaston (1979), pp 64,67; Baton (1991) p44 Sholnk (1980, p 114; Long (1992), p. 289; Chez ‘owski (1990) p. 6; Nei (1991), . 421; Nathanson (1990), . 33, | SS Figure P29 Removal of range ambiguity with staggered PRP (Grom Long, 1992, Fig. 6.15, p. 284). High PRF (HPRF) has a special meaning for pulsed doppler radar: a pulsed doppler radar with a HPRF waveform is ‘unambiguous in doppler but totally ambiguous in range. This is highly desirable for those cases in which targets can be dis- tinguished from clutter on the basis of radial velocity. For this reason HPRF waveforms have found extensive application in airborne radar, where the strongest clutter return lies at a range no closer than the aircraft altitude, and all mainlobe and sidelobe clutter is located in the doppler frequency region £2y, (2, where ¥, is the velocity of the radar-equipped air- craft. The PRF is selected so as to create a large clutter-free doppler region that encompasses the range of expected clos- ing velocities. Range ambiguities occur every cf, meters, where c is the speed of light and f, is the waveform PRE, but target range, if required, can be resolved through the use of20 Apr 45 T™ E 11-219 AIRBORNE RADAR FuG 25 AIRBORNE IFF TRANSMITTER-RECEIVER The FuG 25 equipment is carried iy an cruft to notify short-range ground radar sta tions (Wiraburgs) by coded pulses that the plane is friendly. Manufactured by Lorenz, the FuG 25 has a range of approximately 40 miles, within the range of Warzburgs. It comprises two units: a transmitter-re- ceiver unit (SE 25) and a tone generator, which includes a motor generator (no tone generator has yet been euptured). The receiver- detector is tuned to the Wiirzburg frequency of 500 me (approximately) and the transmitter to the 150- to 160-me band covered by a svecial receiver in the Warzburg. Identification is by phone with an auxiliary signal strength meter which allows D/F on minimum signal, The receiver uses a simple diode circuit fol lowed by three amplifier tubes, the output. controlling @ three-tube transmitter (oscillator and two umplifiers), A signal Input to the receiver of at least 2 millivolts at 30 percent modulation will cause the transmitter to operate ata fixed earrior frequency of 167 me with a power output of 200 milliwatts. It is believed thaz the equipment is coded by keying both the carrier and the modulation frequency with a choice of two codes. Operation of « reluy lights a neon lamp lo inform the pilot that he is heing interrogated. Jt is believed that the FuG 25 equipment was designed originally for bombers, A cap tured document, states, however, that it was to he uscd only in day and night fighters of the types Me 109, FW 190, Me 110 F-4, Ju 88 C-6, and Do 217 J/N and that other aircraft so fitted should have antennas and units removed, leaving mountings and permanent wiring un- touched. It is believed thal this policy has nol been adhered to, Actually, FuG 25 fittings have heen found only in two-engined fighters, It 1 thought that this equipment is not in use at the present time, The characteristic follows: RANGE (miles) : 40 (approximately). FREQUENCY RANGE (me): Interrogates 550 to 580; responds 150 to 160. PULSE RECURRENCE FREQUENCY (cps): 5,000. PULSE LENGTH: 2 mieroscconds. ANTENNA: Vertical rod 14 inches long, di- rectly connected with receiver feeder and matched to transmitter feeder, Antenna is mounted in vulnerable position; damage to it would open the high-voltage circuit and render the transmiller inoperative. POWER SOURCE: “Tongeber” (tune genera- tor) and dynamotor in single unit operating from ac battery. SIMILAR SETS: None. FuG 25A is a later development. POWER INPUT REQUIRED: A signal input to the receiver of at least 2 millivolts at 30 percent modulation is required to cause the transmitter to operate at a fixed carrier frequeney of 157 me. POWER OUTPUT: 200 milliwatts (peak). TYPE MODULATION: CW transmission is modulated by tone generator and coded in Morse. The receiver is sensitive only to a modulation of 5,000 cps. TUBES (type and number) : Transmitter: one RV 12P 2000 MO, two RV P 2000 in push- pull frequency doubler stage, Receiver; one diode (LG 1) in r-f stage and three RV 12 P 2,000 in 1-2 stages. PRINCIPAL COMPONENT of the FuG 2h are as DIMENSIONS WEIGHT Heinht Width Depth ‘T/R (SE 25) unit..8in, 5% in. Gin. 7b. Mon ‘Total weight. 16% Ib.TM E 11-219 20 Apr 45 FuG 25A AIRBORNE IFF TRANSMITTER-RECEIVER ‘The identification equipment FuG 26A, a modified version of FuG 25, is designed to operate with medium- and long-range ground radar stutlons (freyas). Consequenuly Ios. range, 100 to 150 miles, is much greater than that of the FuG 25. White its primary use is for identification purposes, the FuG 254 is also em- ployed as a bombing-release controller and as a navigational aid. Manufactured by Gema, the equipment eon sists of one unit in three sections: transmitter- receiver, cod keyer, and power supply. Two preset codes ean be selected for the trans- miter by inserting two keys in the code keyer. Turning the keys 90° lifts a number of the switehes clear of the xotating cams, leaving the vemaining switches te form the coding sequenee, making and breaking the high-voltage cireuit to the limiter tube and thus keying the transmitier. Only one of the switeh bunks, cun- sisting of 10 parallel contacts, can be selected ul a time, Over 1,000 codes in each bank are possible, The reeeiver is an eight-tube superhetero- dyne. The i-f amplifier is tuned to a fixed freguency of 7 me; the r-f oseillator thus oper- ates 7 mc above the incoming signal frequency band and is swept through the 120- to 185-me band by a split-stator capacitor vonnected across the tuned circuit of Uhe oscillator at a speed of about 400 eps. This short-fast sweep of the receiver frequency insures a wide band with sensitive spot-trequency reception. ‘An h-f tube (type LD 1), capable of operat- ing up to 6,000 me, is used in the oscillator cirenit. Resistors are placed across the i-f transformer windings to insure a bandwidth of 700 ke at 6 db down. When the second detector has a peak signal voltage of at least 2.6 volts at the plate, the trigger tube causes the transmitter to operate and to deliver a narrow pulse of high amplitude to the antenna. High output at iow voltage makes the equip- a ment idea! for high altitudes. The FuG 25A, though airborne, is presct on the ground before the take-off and the Morse recognition elpnat cannot ve changed tn Ube air. When used as a navigational aid, the air- craft is plotted from the ground by responses from the FuG 25A and is given course direc tions hy voice radio, using FuG 16 or the transmitting radiotelephone attachment to FuG 10 which has a longer range than FuG 216. Lutest use for the FuG 25A is for remote bombing control, using two Freya stations for atfix.” Since FuG 25A functions with the German long-range Freya radars, it is to be expected that it would be fitted in all aircraft likely to be employed in attacks on Allied territory in order that they miyht be identified as friendly on their return journey. Crash examination, however, has shown that very few aircraft operating over Great Britain and the Mediter- ranean area have ever carried the equipment, although the mounting frame may be fitted, It could easily be installed in many types of aircraft should tactieal requirements demand it. ‘The German policy with regard to this equip- ment is nol definitely known, The charaeteristies of the FuG 25A are as follows: RANGE (miles) ; 100 to 150, FREQUENCY RANGE (me): Interrogates 193 to 128; responds 152.2 to 161.0. Trans- mitter spot frequency 156.0. A received pulsed signal operates the transmitter on a preset frequency within the preseribed range. PULSE RECURRENCE FREQUENCY (cps): 500 (minimum). PULSE LENGTH: 0.5 microsecond. ANTENNA: Vertical rod 14 inches long for both transmitting and receiving through the special antenna matehing unit. TYPE OF PRESENTATION: Visual. POWER SOURCE: A self-contained rotary274 11.6 Airborne and Space-Based Radar Issues ridge in Figure 11.6. A target, slightly offset from this clutter ridge, is also shown 10, pp. 51-64]. The plot at the top of Figure 11.6 shows the angular response of the antenna. The clutter fills the beam and screens the target. Similarly, the plot at the right of the figure shows the radial-velocity response of a pulse-Doppler processor. Here, the clutter fills the region 4MDV, screening the target. Neither form of independent processing will allow detection of the target. With STAP, the angle and radial- velocity data are jointly processed, allowing detection of the target. STAP involves processing of the M pulses in the pulse train with each of the N elements in the line: array that forms the azimuth beam, Thus, the processing has a dimensionality of NM, which may be orders of magnitude greater than the process- ing for either array beam formation (N elements}, or pulse-Doppler processing (M pulses). For example, with 500 linear array elements and 200 pulses processed, NM = 105, STAP processing involves calculation of NM complex weights that, when applied to the signal data, will minimize the interference. The signal data in each range-resolution interval is then processed to yield the desired target signal. Both of these steps involve NM x NM matrix inversions [10, p. 65-71]. This pro dom are rarely needed to cancel the interference, and STAP processing usually employs reduced-rank processing, reducing the computational requirements [10, pp. 114162]. Much current STAP research is addressing techniques for reducing the computational requirements [11]. Application of STAP cancels jammers as well as clutter. However, the jamming, interference occurs at fixed angles independent of Doppler frequency. The sidelobe canceling technique described in Section 10.3, or its equivalent using an N-clement matrix, may be used. Thus, STAP is not required just for jamming cancellation, but if itis applied for mainlobe clutter cancellation, it cancels jammers as well. Factors that may limit the performance of STAP include: ing may be computationally intense. However, NM degrees of free * Clutter thar is nonstationary, and clutter that is discrete; + Clutter internal motion and scintillation; * Uncorrected antenna motion perturbations; + Radar transmitter and receiver instabilities. Synthetic-Aperture Radar (SAR) SAR is used to generate two-dimensional images of terrain and associated ground targets. Conventional radar may provide good range resolution by using wideband waveforms (see Chapter 4}. However, the cross-range resolution is limited by the radar azimuth beamwidth, 0,, to RO, For example, with a 50-MHz signal bandwidth, the range resolution is 3m. With a grazing angle of 20°, the ground resolution in the range dimension is 3.2m (see Section 9.1). An azimuth beamwidth as small as 3 mrad, produced by an 11-m aperture at X-band, provides a cross-range resolution of 300m at a range of 100 km, two orders of magnitude greater than the range resolution (see Section 8.2).11.6 _Synthetic-Aperture Radar (SAR) 275 As the name suggests, SAR produces the effect of very large airborne or space- based apertures, creating very narrow effective beams that can give cross-range resolution comparable to the range resolution. In an SAR mode, the moving radar views the target from successive locations of the antenna. The signal returns are combined coherently to give the effect of a large aperture that is equal in length to the space occupied by the antenna during the processing interval. This is illustrated in Figure 11.7. The real-aperture antenna having a dimension of Wey produces a beamwidth, Og,, of A/w,,, as shown in Figure 11.7(a), giving a cross-range resolution of Ri/vg,. Figure 11.7(b) shows the same real aperture as it moves through several successive positions to create a synthetic aperture having a dimension w,. The angular resolution, or beamwidth, 8,,, produced by this syn- thetic aperture is given by [12, pp. 21-4-21-7): 8x4 =H Qs) (11.8) and the cross-range resolution, ACR, is given by [1, p. 616]. ACR = R42.) (11.9) The factor of two in the denominators of (11.8) and (11.9) is a consequence of radar transmission and reception being done independently at each successive antenna location in the synthetic array, while a real aperture illuminates the target with a single field and receives the composite target return from that illumination [12, p. 159} For example, if the synthetic-aperture dimension is $00m, the SAR beamwidth at X-band is about 0.032 mrad, and the cross-range resolution at a range of 100 km. LL | r ® % Aim) © Figure 11.7 Ilustration of real-aperture and synthetic-aperture beam formation. (a) Real aper- ture. (b) Synthetic aperture.276 Airborne and Space-Based Radar Issues is 3.2m, the same as the range resolution in the previous example. For a platform velocity of 300 mvs, the time to generate this synthetic aperture is 1.67 sec. Such an SAR could generate images of the terrain and targets with resolution of 3.2m in the two orthogonal dimensions. ‘To avoid grating lobes in the SAR beam pattern, the synthetic aperture must be sampled often enough to produce the effect of a filled array (see Section 3.3). The required radar PRF is given by [13, pp. 21-18-21-21): AV PRE> (11.10) Since SAR relies on platform motion to generate the synthetic aperture, it may not be used by stationary platforms such as airships or synchronous satellites. A similac technique, called inverse SAR (ISAR), is sometimes used to image targets that are moving past a radar, or are rotating [1, pp. 617-618 and 640-651). SAR processing uses the phase changes during the processing time to determine the angular position of stationary ground objects. Moving targets that have radial velocity have additional linear phase change from pulse to pulse, which makes them appear offset in angle from their true positions. For example, vehicles may appear to be off the roads on which they are actually traveling. When this apparent position change is a problem, special processing techniques may be used to correct for it. SAR techniques may be classified by the complexity of the processing employed and the resulting radar capabilities. Three such classes of SAR (Doppler beam sharp- ening, side-looking SAR, and spotlight SAR) are addressed below. The processing of radar returns in Doppler beam sharpening assumes linear phase progression from successive returns. This is referred to as unfocused Doppler processing. The phase changes, due to the changing target range as the radar passes the target, cause the synthetic-aperture length to be limited to [13, pp. 21-4-21-7] ws, <(RA)"* (Doppler beam sharpening) (11.11) and the resulting cross-range resolution is limited to: ACR S05(RA)"? (Doppler beam sharpening) (11.12) For example, an X-band (9.5 GHz) radar, observing targets at 100-km range, would be limited to a synthetic-aperture length of 56m, and could produce cross- range resolution no less than 28m, For platform velocity of 300 m/s, the maximum processing time is 0.19 sec. These results are summarized in column 2 of the lower part of Table 11.2, which also gives the radar characteristics in the upper part of the table. Such cross-range resolution is usually not useful for terrain imaging. However, it may be used for resolving, detecting, tracking, and measuring the characteristics of closely-spaced targets real-aperture size, as discussed below for side-looking radar, but this rarely a limita- tion for Doppler beam sharpening. The cross-range resolution is also limited to one-half of the11.6 _Synthetic-Aperture Radar (SAR) 277 Table 11.2 _SAR Examples Parameter ue Frequency 9.5 GH Antenna length 2m (real aperture) Antenna beamwidth 0.905" (real aperture) Platform velocity 300 mis PRE 4,000 Hz Target range 100 km SAR Mode Doppler Beam Sharpening Side-Looking SAR Spotlight SAR Synthetic-aperture length 56m (maximum) 1,580m (maximum) 3,160m Cross-range resolution 28m (minimum) Am (minimum) 0.5m Processing time 0.19 see (maximum) 5.27 sec (maximum) 10.53 see In side-looking SAR, the radar antenna is usually oriented to look normal to the platform flight path. The elevation beam is broad enough to cover the ground area of interest, and may be shaped to equalize the sensitivity at the various ranges, as illustrated in Figure 11.8. The SAR processing generates a continuous map of the range swath observed as the platform passes the terrain (see Figure 11.8). This is often called a strip map. Side-looking SAR usually employs focused processing, which corrects the signal phase for range changes as the radar passes the targets. The synthetic-aperture length is limited by the viewing angles for which the target remains within the real- aperture beamwidth, (Og4, to ([13, pp. 21-4 to 21-7)]: Wy, SRO,, (sidelooking SAR) (11.13) The resulting cross-range resolution is limited to no less than one-half of the real-aperture linear dimension, 10,4: ACR 2w,,/2 (sidelooking SAR) (11.14) max Coverage swath ——>| Figure 11.8 Geometry for side-looking SAR.278 Airborne and Space-Based Radar Issues These performance measures are independent of range, since the time that a ground target remains in the real-aperture beam is proportional to range. For example, for an X-band radar having a 2m real apercure, the beamwidth is 0.905° (15.8 mrad), as given in the upper portion of Table 11.2. At a range of 100 km, the maximum syntheric-aperture length is about 1,580m, and the minimum cross-range resolution cell is Im. The maximum processing time is 5.27 sec. These results are summarized in column 3 of the bottom portion of Table 11.2. In side-looking SAR, the PRF determines the unambiguous range extent of the strip map that may be generated, as well as being constrained by (11.10). For exam- ple, with a platform velocity of 300 nvs and a 2m real aperture, the PRF must be at least 600 Hz. With this PRF, the unambiguous range is 250 km, and this is the maxi- ‘mum range extent of the strip map that may be generated. Side-looking radar is widely used on aircraft and satellites for mapping terrain and sea surface. While most dedicated side-looking radar operates with the beam normal to the platform velocity, other look angles may be used, with some degrada- tion in the resolution. Spotlight SAR is used when cross-range resolution is required to be better than that provided by side-looking SAR. This may be the case when extremely small reso- lution is required, for example to identify the type of ground vehicle, or when a lage antenna is used to provide the required radar sensitivity. The latter situation often occurs for SBR, where large antennas are required, due to the long ranges [14]. In spotlight SAR, the radar beam remains on the target area long enough to gen- erate the desired synthetic aperture. The processing corrects for range changes as the radar passes the target. It may also correct for motion of target pixels between range and cross-range cells, using techniques that are sometimes called extended coherent processing, This allows large synthetic apertures to be generated, producing small cross-range resolution [15]. To achieve long observation times, either the flight path must curve, or the beam must steer to follow the region being imaged. The latter mode is usually employed, using phased-array radar. The length of the synthetic aperture in a spotlight mode is given by: Wg, =Vpty (spotlight SAR) (11.15) where #, is the processing time. The cross-range resolution is then [14, p. 503]: Ri RA ACR 5 Waly (spotlight SAR) (11.16) An example is shown in column 4 of the lower portion of Table 11.2. In this example, a cross-range resolution of 0.5m is desired at a range of 100 km. This is one-half of the cross-range resolution that is provided by the side-looking SAR in the previous example shown in column 3. The required synthetic-aperture length is 3,160m, which is produced with a processing time of 10.53 sec, twice the values for the side-looking SAR. Note that the cross-range resolution for spotlight SAR is pro- portional to range. Spotlight SAR may be used to generate a high-resolution image of a limited sur- face region, such as a factory complex or an airfield. In this mode, a pencil-beam antenna pattern may be used. As with strip mapping, spotlight SAR may be done at11.7. Problems 279 11.7 look angles other than normal to the flight path, but with degraded resolution. The pencil-beam phased-array radar in fighter aircraft may be used in this mode, given suitable processing. Spotlight SAR may also be used by phased-array radar to produce images of large areas, or strip maps that have smaller cross-range resolution than indicated by (11.14). This is achieved by revisiting each portion of the terrain area periodically using electronic scan, so that the limitation imposed by the array beamwidth does not apply. Such techniques are used by large airborne surveillance radars, and by SBRs with their large antennas [14, pp. 503-505]. In spotlight SAR, the PRF also determines the unambiguous range extent of the area that can be observed. SAR signal processing employs Fourier-transform-like algorithms. It must cor- rect for perturbations in the flight path. Early SAR used optical elements to perform the processing, with the resulting images recorded on film [16, pp. 23-17-21-23]. ‘Most modern SAR use digital processors, which gives flexibility to handle a variety ‘of modes and parameters, The complexity of the required processing increases rap- idly with decreasing cross-range resolution, both due to the increased data and to the complexity of the algorithms needed [1, pp. 632-636]. When strip maps or images of limited regions are produced at different times, it is often of interest to find changes (e.g., the deployment of new weapons) that have occurred between observations. This is accomplished by a technique called change detection, which maps one image onto another image of the same area, and identi fies features that have changed. SAR may be used to identify objects in the images it produces. Algorithms for doing this are beyond the scope of this book, but typically from 10 to 50 resolved pixels on the target are needed to reliably identify the target [17]. This discussion of SAR has addressed two-dimensional imaging of terrain and surface targets. It is possible to measure the altitude of terrain and targets in SAR images by using interferometry between two SAR antennas having vertical separation, Problems The following problems are provided to assist in reviewing this chapter and to ensure a basic understanding of the material. For maximum benefit, the problems should first be solved without using the VBA custom radar functions. Solutions to these problems are given in Appendix E, Section E.11. 1. What physical factors limit the antennas that may be used for airborne radar? 2. Fora radar on an aircraft with a velocity of 350 mvs and having a depression angle when viewing the radar horizon of 10°, what is the velocity spread of the sidelobe clutter? If the radar frequency is 3.3 GHz (S band), what is the spectral spread of the sidelobe clutter? If the mainbeam has an azimuth of 60° relative to the platform velocity, and a depression angle of 25°, what is the mainbeam clutter velocity? What is the corresponding Doppler-