AN ADVANCED SPACE SURVEILLANCE SYSTEM (UNCLASSIFIED) Space Surveillance Branch Applications Research Division. 8 February 1961

Transcription

1 «* NRL Memrandum Reprt llvf AN ADVANCED SPACE SURVEILLANCE SYSTEM (UNCLASSIFIED) Space Surveillance Branch Applicatins Research Divisin 8 February WMGHADED AT U.S. NAVAL RESEARCH LAB0RAT0^ 3VEAH Washingtn, D.C. ^DECLASSIFIED AFTER 12 YEARS INTERVALS APPROVED FOR PUBLIC RELEASE - DISTRIBUTION UNLIMITED P* <,/ / aan Ugl_ Cde. gt5c * ' Beard. IJl Jim<\

2 CONTENTS Abstract Prblem Status Authrizatin INTRODUCTION Backgrund Requirements System Cncept Early Detectin High-Altitude Cverage Size f Objects SYSTEM DESIGN Detectin f Distant Objects Calculatin f Range Angle Measurement Detectin at Shrter Ranges Orbit Predictin Increased Accuracy Lw-Height Cverage Site Selectin Frequency f Operatin Feedthrugh Saturatin COST AND TIME SCALE IV iv iv It, 16 n, H ^ 2S in

3 ABSTRACT Requirements fr a space surveillance system are nw being generated by the varius unified and specified cmmands. While details remain t be determined, in general it may be stated that the requirements express a need fr three categries f cverage: (a) early detectin, (b) cverage f lw inclinatins, and (c) cverage at extreme altitudes. A minimum-cst, space surveillance system which prvides early rbit determinatin f all satellites ut t nearly 30,000 nautical miles is described. The system cnsists f U.S. installatins, t prvide refined rbital data n mst knwn satellites, and special installatins, t prvide data n rbital elements f new satellites and special rbits. The special installatins wuld be lcated n islands in the Pacific and Caribbean t prvide extended lngitude cverage and t prvide equatrial cverage fr bth lw-perid and 2^-hur-perid rbits. The detectin device described utilizes high-pwered, cntinuus-wave transmitters, fixed antennas, a nnambiguus ranging technique, and precise determinatin f angles t give a gd rbit (errr in perid apprx. 0.1$) fr satellites abve 350 nautical miles (seen by tw statins) and a less accurate rbit fr satellites belw that altitude seen by ne statin (errr in perid 0.1$ t 1$). The prpsed initial installatin at Truk-Pnape and in Flrida prvides 30,000-mile cverage fr the island installatin and lwer latitude cverage fr the U.S. experimental installatin. PROBLEM STATUS This is a prpsal fr extensin f the Space Surveillance System, cntinuing n the apprved prgram. Wrk is AUTHORIZATION R02-35 iv

4 AN ADVANCED SPACE SURVEILLANCE SYSTEM INTRODUCTION Backgrund The U.S. Naval Research Labratry initiated a system develpment in June 1958, designed t detect and predict rbits n nnradiating satellites. The prgram was spnsred by the Advanced Research Prjects Agency under Order 7 until Octber 18, i960, whereupn administrative and technical respnsibility was transferred t the Department f the Navy. NORAD is respnsible fr prviding requirements, representing the needs f all services, which are t be the basis fr a develpment plan. The present Naval Space Surveillance System measures angles, angular rates, and dppler. Satellite psitins are determined by simultaneus angular measurements. The use f angular rates and dppler prvides a crude rbital determinatin frm ne pass. The use f psitins determined n tw passes prvides a gd rbital determinatin. This system was adapted t a cntinental U.S. lcatin and is nt cmpatible with installatin n restricted areas such as an island. Fr early detectin f satellites, an island installatin will be necessary. The detectin system t be described in this prpsal is an extensin f the present system, is cmpatible with an island installatin, and is designed t satisfy the anticipated military requirements. Requirements The requirements, nw in preparatin, are knwn in brad terms and indicate a need fr: (a) cverage f all rbit inclinatins, (b) extending the range t detect satellites f military interest, (c) prviding detectin f lw satellites, (d) early detectin f unannunced satellites, (e) prvisin f timely utput infrmatin, and (f) suitability fr wartime peratins. This last item includes High reliability and freedm frm interference. The range f detectin has n Finite limit. There is always a tradeff between range, assumed reflecting area, md prbability f detectin. A special, and useful, satellite is ne apprximating a 2^-hur perid, statinary rbit. Such satellites wuld be lcated at abut 22,000 nautical miles frm the earth. A requirement fr cverage t include such satellites is assumed. lystetn Cncept The present Naval Space Surveillance System has emplyed techniques selected rimarily t slve the prblem f detecting rbiting bjects. Simplicity, lw st, and high reliability have been imprtant cnsideratins. Radi illuminatin frm the grund prvides the means f "seeing" nnradiating atellites. Because the returned signal varies with the inverse furth pwer f he range, and because ranges are extremely great, it is imprtant t maintain a

5 high average transmitter pwer. A pulse system may emply pulse cmpressin and signal prcessing t increase the average pwer. A cntinuus-wave (cw) system reaches the ptimum in this respect and, in additin, is simple, reliable, and permits the use f narrwband receivers fr high sensitivity. Therefre, cw illuminatin has been selected. A thin vertical-fence type f cverage has been selected. This is practical because f the inability f satellites t make radical and frequent changes in their rbit. The bservatins are made at the pint f clsest apprach t maximize the range capability. The thin beam reduces the pssibility f multiple satellites passing simultaneusly thrugh the system. Als, there is n scanning lss. High-gain antennas must be emplyed t make lng-range detectin practical. Rws f diples abve a hrizntal grund screen are used fr mst arrays. Narrw predetectin and pstdetectin bandwidths are used in the receivers. The pstdetectin bandwidths are chsen t match the signal duratin. These will be autmatically selected in systems cvering bth shrt and lng ranges. Narrw predetectin bandwidths are btained by use f a cmb filter, which finds the dppler-shifted frequency and tunes the measuring system t the prper frequency s that a narrw bandwidth may be emplyed. The bservatinal data are taken with a radi-interfermeter system. Such a system permits very high angular reslutin with simple antenna design. The reslutin is dependent upn the separatin f a pair f antennas. Separatins f ne mile are t be used. Fr installatins n limited prperty, such as an island, range as well as angular measurements wuld be made. Range is"* prvided by emplying what might be termed a frequency interfermeter. A side frequency, remved, say, a few hundred cycles frm the primary illuminatr frequency and lcked in phase with it, is transmitted. The phase difference between the received frequencies when cmpared t the directly received transmissin is a measure f the range. Because the side frequency is fixed in frequency relative t the primary detectin frequency, knwledge f where t lk can be prvided by the detectin channel, thereby minimizing the required side frequency pwer. Use f several side frequencies increases the accuracy in ranging and eliminates ambiguities. The principal prblem with a cw detectin system is the need fr eliminating feedthrugh frm transmitter t receiver. This prblem is reduced by separating the transmitter by a gd distance frm the receiver. A minimum distance f apprximately thirty miles is indicated. Further islatin is btained by the use f suitable antenna patterns and filtering. Early Detectin Figure 1 shws the apprximate, initial, subsatellite tracks fr Russian satellites launched t date. As shwn, the first pass that crsses the U.S. is #6. T detect pass #1 a statin n Jhnstn Island is indicated. Hwever, if the perid

6 tn X y en ^J O =5 ^ w en - CD. I!-> a) «J en fl en.5 2* a, u a, s i DJC rh

7 f the rbit is lng and the launching site is west f the spt shwn this pass might be missed. In the case f a miss, the first detectin will still ccur ver the U.S. A statin in the area f Pnape wuld, at wrst, detect the secnd rbit and might detect the first rbit. The lngitude cverage that can be btained frm a single site is limited by the minimum height f the satellite t be detected. Fr detectin t 85 nautical miles, the earth's curvature limits the detectin range t 750 nautical miles. This cverage frm Truk-Pnape is shwn in Fig. 2. The lngitude f Pnape is apprximately 90 frm the center f the present line and is thereby in the ptimum psitin t minimize the detectin time (fr a single installatin) fr the general case f any satellite having an inclinatin abve 33 Thus, Pnape is an attractive lcatin fr an initial extensin f the present system. The additin f statins in the regin f the Philippines, Hawaii, the U.S., and Puert Ric wuld prvide nearly 100 percent first-rbit cverage fr satellites having inclinatins greater than abut 10. Fr lwer inclinatins, an installatin lking suth frm the Philippine will detect every rbit (Fig. 3)- The Pacific Missile Range has cmmunicatins t Hawaii, Kwajalein, Guam, and the Philippines and prpsed cmmunicatins t Tarawa (Fig. k). The Army has the additinal cmmunicatins shwn. High-Altitude Cverage The cverage f satellites at extreme altitudes is cmplicated by their extrem< range and by the characteristic (fr 2^-hur satellites) f nt necessarily crssinf many degrees f lngitude. The subsatellite tracks f 2l+-hur satellites may be a pint (fr a circular, zer-inclinatin rbit). Fr a zer-inclinatin, nn-circuls rbit, the track will cver a path alng the equatr. Fr different inclinatins and eccentricities the rbits will describe tracks which resemble figure 8's f varius distrtins. Since such satellites can cntinuusly survey any particular sectin f the earth's surface (except near the ples), they are f particular interest fr recnnaissance devices. Fr detecting all such devices, a minimum f three statins wuld be required t cver the entire equatrial plane thrughut the range f interest. Size f Objects In designing a detectin system ne f the first parameters that must be selected is that f effective target size. In the Navy Space Surveillance System the measured variatins in size f several satellites have been determined and are shwn in Fig. 5- These curves shw that the plt f size (<r) versus the prbability f a given size being exceeded is a lg nrmal distributin. The distributin shwn in Fig. 6 has been chsen fr calculatin purpses. Here, a ne-square-meter bject is defined as ne which appears t be ne square

8 n a r-l V ft m I H a; bo n) M > 0 c H M u v Q i 60

9 h d (M 0 O <) f) 0 O <\l f) 0 O <* lo A V 4-» rh 0 O > -*-> () m >^ 1 1 rrt w rh * M 01 n u -*-> 0 Cj 01. i r 1 fh 4) 0 g > CO M =5 3 cr (T) en <1> bfi 8 t; >-< 0, OJ 0 m > ~ V u -4-> O 5 0 IC nj. 1 c -.H X) S U r-( 01 i i n i i CO 0 00 IT" e e 0 0 p O O O O O 0 b O O O O O r^ e in <J- fo <\J ~ CM ro <T tf-> t

10

11 , [y 133d OS Nl - / '> 1 / / \f \ / > / f 7 /} X / UJ k <t fc CD 1 -MIDAS L 1 1 CM M O CO ^ ^ P => CL t i O CD O CD cvi U313W s r 3Aav aa NI - 1 t O H < > Q Q H t 5.35 DB DB DB 6.08 DB t 4! 1 < * ; 2<P : CD2CO ; -. CM grsi VWN \ OOOOC UCDCDCOC it i ii >( 5<» D 1 O IO N u O _i CD 0 i 0 3 ^1 1 8? CD 0 CD H w CD CD <D CD ^^ LU CD h- CD < a* X! 3 DC-J 3 s z 2 rt 4) -H Q r H r LU rt fj CD X xi -a 1- c -M a, H tn I-I Q LU LU X> O (0 O O U X 3,«LU O; +3 b 5 "2 T3 0) r H d a> < rt u X 1- SI " O > t R h- S 6 _J i i ^ CD 3 < tn V a ~ Q: 5 D Q. Fig. 5 - Ac th

12 $ <r> (T> 9) N H i M - Li L &l 1 t >/ a / m ^ ^ UJ < 4-1 C 3 ^ r 3 > Q r 3 «? m a ^ UJ <T> X._, 1-5 be (f) Q 4-> C UJ r4 O UJ 3 S, u, > 05 (0 X Ul b r < P O I- _i 00 rba rre OH U d the s the r <D >- 0) H en 4j < - 2 DQ a fa a: Q_ satel sure r Xl 0) 4J c w ^ t ^ U313W OS ) laav aa NI * Mi rh

13 meter r larger fr 50 percent f the bservatins, and frm the figure it is seen that fr 95 percent f the bservatins it will appear t be larger than 0.1 square meter. T cmpute the range f a system having a 95 percent prbability f detectin n a ne-square-meter target, a size f 0.1 m 2 is used in the radar equatin. In the detectin system t be described, a 0.1 m target size is used in the calculatin fr clse satellites (under 4000 nautical miles) and 1 m 2 fr targets at great distances (4000 t 30,000 nautical miles) giving, in each case, a 95 percent prbability f detectin fr satellites ten times as large. SYSTEM DESIGN Detectin f Distant Objects (1+000 t 30,000 naut mi) The first system described is designed fr detectin at extreme ranges. It may be easily degraded fr use at any lesser range. It is assumed that detectin f this bject will be accmplished by means f a radi reflecting system, which is limited in range accrding t the well knwn equatin P t Gi G 2 \ 2 <r R*- - P r (Utr)3 * Frm this equatin it is bvius that fr detectin at great ranges the transmitted pwer must be large, the antennas must have high gains, and the receivers must be sensitive. The transmitted-pwer parameter that determines range in an ptimized system is the energy, r average pwer, multiplied by bservatin time. The mst efficient means f generating such energy (high average pwer) is by the use f a cntinuus-wave transmitter. High-gain antennas can be prduced ecnmically nly if they are fixed devices. Since the functin f the antennas is t illuminate the equatrial plane, fan-type beams are indicated. Such beams can be prduced by antennas which are bth ecnmical and rugged. The requirement fr a sensitive receiver affects the chice f frequency f peratin. Lw-nise receivers are easily built at the lwer frequencies, and at these frequencies (apprx. 100 Mc/s) the system sensitivity is determined by sky nise. The dppler band als is less, s fewer filters need be used in the cmb preselectr. The ptimum detectin frequency fr a system using fan-type antennas is in the regin f rughly 150 t 500 Mc/s. Calculatin f Range The range is calculated n the basis f a transmitter f 2 raw and an antenna 2 miles lng. The receiving antennas are 1 mile lng. Calculatins are made fr bth 150 and 1+50 Mc/s. 10

14 The sensitivity f the system is determined by ne f tw pssible cntrlling factrs. One factr is the detectin sensitivity f the system, and the ther is the sensitivity f the phase-measurement equipment. Fr detectin sensitivity the signal required t attain a signal-t-nise rati f 13 db in a given filter band is first determined. Cnsider Fig. 7 fr the case f a pstdetectin bandwidth (B) f 0.2 cycle and a predetectin bandwidth (b) f 100 cycles. The sensitivity is 153 dbm. Nw cnsider Fig. 8 fr the same pstdetectin bandwidth and a predetectin bandwidth f 25 cps and a phase nise ((j^) f 25. At a pst detectin bandwidth f 1 cycle the sensitivity is 156 dbm. At B = 0.2 it wuld be 163 dbm. By making the detectin antenna at least 7 db mre sensitivity then the phase-measuring system, the phase-measuring device becmes the cntrlling factr. At U50 Mc/s the prcedure is the same, except*the bandwidth is greater, and since the temperature is less the sensitivity is 6 db greater than fr the 1000 K :urves shwn. T keep the number f filters cnstant, the bandwidth at 450 Mc/s Ls assumed t be 300 cycles, as cntrasted t 100 cycles at 150 Mc/s. Under ;hese cnditins, as shwn in Table 1, the range btained at the tw frequencies Ls identical. A schematic f the range cverage is shwn in Fig

15 CO a- _ X 9 Q < I- I CD a + > OH > XI r* TO CO,jQ t CO v a, em H O O O - 00 i^- CD in <fr r <NJ waa-u3m0d indni YOISAS ±u3~iv-dd 12

16 CM 1 g CVl 1 OJ CM 1 en CO i O CD I 0 OJ in rt O a 0) in rd A tx in > O *_» "5 en C m U V in n) i CO tx O CD 00 h- t lo ^J- r CO ^~ ^~ v\iaa 13

17 \ 8* V 60 flj U V > O O in cr- rh 14

18 Table 1 Evaluatin f Detectin System Parameters at 150 and 450 Mc/s System Parameters 150 Mc/s 450 Mc/s 150 Mc/s Bearawidth (interfermeter antenna) x x x 120 Gain , Near Field 1600 stat mi 4800 mi 1600 stat mi Bearawidth (Detectin Antenna) x x x 15 Gain 37, ,000 37,700 Bearawidth at 20,000 miles 25.2 mi 8.4 mi 25.2 Time in beam (at 4 mps) 6.2 sec 2.1 sec 6.2 Bandwidth needed 2 cycles 5 cycles.2 cycles Nise temperature 1000 K 250 K 1000 K Predetectin bandwidth (Detectin) 100 cycles 300 cycles 100 cycles Sensitivity (13 db = S/N) 153 dbm 154 dbm 153 dbm Sensitivity f phase meter 163 dbm 163 dbm 163 dbm Gain f transmitting antenna 9,44 28, Pwer 2 x 10 6 watts 2 x ^ watts 2? Range fr <T = 10 ft = 1 vr 1.4 x 10 8 feet 1.4 x 10 8 feet 10 8 feet (Size used in calculatin) 26,700 stat mi 26,700 stat mi (23,200 naut mi) (23,200 naut mi) 19,000 stat mi (16,400 naut mi) Range fr cr = 100 ft^ = 10 ra c (Size used in calculatin) 47,000 stat mi 47,000 stat mi (41,000 naut mi) (41,000 naut mi) 33,700 stat mi (29,000 naut mi) *ange fr cr = 1 ft = 0.1 m 2 (Size used in calculatin) Bandwidth 0.4 cycle 1.0 cycle 0.4 cycle Sensitivity 160 dbm 160 dbm 160 dbm Range 12,600 stat mi 12,600 stat mi (10,900 naut mi) (10,900 naut mi) 9000 stat mi (7800 naut mi) 'Assumed bandwidth t keep the number f cmb-preselectr filters the same at 450 Mc/s as at 150 Mc/s. 15

19 Angle Measurement The antenna patterns determine the angle f arrival f the return signal t within 0.02 t 0.07 in ne directin and t +7.5 in the ther. This angular reslutin is increased by using several interfermeter antennas. The sensitivity f the receivers in this angle measuring system, nce the dppler slt is knwn, is 162 dbm. Thus,antennas may be used having gains 9 db less than were needed fr the detectin channel. Antennas emplying nly ne lbe will be adequate fr the angle measuring system. At 460 Mc measurements can be made t 0.1 mil accuracy with reslutin t 0.02 mil; at 150 Mc the accuracy is 1 mil and the reslutin is.06 mils. Range Measurement With the angle f arrival measured quite accurately by the interfermeter, the range t the target is all that is needed t lcate it. A ranging system using as much f the detectin equipment as pssible, but in n way limiting the detectin range, is emplyed. It may be as nnambiguus as needed and as accurate as required Cnsider a secnd transmitter generating a frequency remved 1 cps frm the detectin transmitter. The phase f the tw transmitted signals is cmpared t the phase f the reflected signals. The phase f the reflected signals, as cmpared t the transmitted signals, will be a direct measure f the range t the target and will be nnambiguus t a distance f 83,000 nautical miles. While this measurement will be nmambigupus, it als will nt be very accurate. T imprve the accuracy, anther frequency separated by perhaps 5 c Ps frm the detectin signal will imprve the accuracy by a factr f five. Other frequencies, say 25 cps, 125 cps, 625 cps, etc., may be used t further imprve the accuracy. When the phases f all f these side tnes are cmpared t the detectin frequency, a range nnambiguus t 80,000 nautical miles and accurate t perhaps 0.10 miles is btained. Since the apprximate frequencies f the side tnes are knwn when the dppler f the detectin device is determined, the pwer f each side tne need be nly apprximately 0.1 that f the detectin device. The ttal pwer in the side tne will be apprximately the same as in the detectin transmitter. Detectin at Shrter Ranges Detectin at clser ranges is, f curse, much easier than at 30,000 miles. Prtins f the lng-range receiving antennas can be used fr measurements dwn t ranges f 85 nautical miles. Hwever, at shrter ranges additinal requirements are impsed in rder t prvide adequate rbit predictin. Orbit Predictin In general, rbit predictins can be btained in tw ways. One methd invlves measurement f psitin and time at tw widely separated pints in the rbit. The 16

20 fan-type system is especially adaptable fr this measurement. Lwer satellites, n which nly ne bservatin can be made, require a different technique, that f measuring psitin and velcity. Cnsider nw that the rbit f a satellite is t be predicted using a velcity measurement. The velcity V may be expressed as fllws: - v 2 j. v 2 - ( 2 1 > - V r + V t - f, (- - - ), where V r and V-t are the radial and tangential cmpnents f velcity, r is the distance frm the center f the earth, a is the semimajr axis f the rbit, and /l = g R 2. One btains 2, dr V r dv r + V t dv t da = 2a (-«+ \. ' r 2 yu ) Fr simplicity, cnsider the perid as the desired parameter. Since T 2 - (^) 2 a3 M-H da= 3 at (* + Vr dv r + V t dv t) _ p ^r a circular rbit, a = r, M = av, dr = and V r =. Therefre, dt dv t T vt ut dv-t/v-t = do/9 where do is the reslutin f the receiving system and 9 the eamwidth. Fr ptimized detectin, antenna beamwidths f apprximately 0.1 (0.002 r ) ppear desirable. At 150 Mc/s an angular accuracy f r and a reslutin f.0001 r appear reasnable with interfermeter baselines f abut 1 mile. Under hese cnditins, dt _, ( T.002 ) ie perid can thereby be measured t 15 percent. 17

21 Increased Accuracy T imprve the accuracy, a higher frequency culd be transmitted simultaneusly with the detectin frequency. The dppler experienced by this higher frequency is related t the dppler at the detectin frequency by the rati f the frequencies. Once the detectin frequency is knwn, the dppler f the higher frequency is knwn and phase-measuring receivers can be tuned t receive it. Cnsider nw that an angular rate frequency f 1500 Mc/s is selected. T have a 0.1 beam, this antenna need nly be 0.1 mile lng. The far field will start at 160 miles. In the 150 Mc/s antenna the far field starts at 1600 miles. The beamwidth f this antenna is never less than 1 mile, s an additinal measurement time fr the high-frequency signal can be btained by using tw 0.1-mile antennas lcated at either end f the 1-mile antenna t give an effective bservatin angle f l/loo r 0.01 r at 100 miles and at 200 miles. Assuming a ne-mile baseline, the reslutin f the high-frequency antenna will be 10"5 radians, giving the fllwing values fr. T Height: 100 miles 200 miles 300 miles _ d9 dt 3 -Q = T~ Accuracy f 100-minute rbit 18 sec 36 sec 48 sec The determinatin f rbits by making bservatins f psitin at tw separated pints in the rbit can be cnsidered as a special case f the d /0 criterin in the fllwing manner. In Fig. 10 cnsider that bservatins are made frm A and B with an accuracy f dx (A and B are separated by a central angle 0). Let r A and rg be the radius vectrs f the bservatins. Then dxi/r A is the angular errr in the bservatin at A, and dx^/rg is the angular errr in the bservatin at B. If we assume dxi/r A = dx2/rg, then doj. = dxi/r^ = dx2/rb«doj. is related t the accuracy f bservatin at A and B by the relatinship d A = &X±/E&. Frm the figure, d0 A d0 r (H A + R) H A d0 r = H A d0 HA + R Therefre, t d6j ^r _ 3 dq A, H A 0 HA + R ). 18

22 Fig Gemetry f the tw-fence velcity determinatin 19.

23 This equatin shws that bservatins made at lw altitudes prvide better rbit determinatins than at high altitudes by the rati f the altitude t the radius vectr. At great distances (HA>R) this equatin reduces t dt/t = 3 dqa/. In tw statin installatins f this type it might be expected that A and B wuld differ in central angle by perhaps O.lr. At 150 Mc/s d9a 210"3r; at k30 Mc/s doa = lo-'+r. Fr 0 = 0.1 r : HA 1000 miles 30,000 s miles 150 Mc/s 1*50 Mc/s 150 Mc/s 1*50 Mc/s do " * r r dt/t Imprved accuracy will be btained by either having statins spaced at greater central angles r by using mre accurate bservatins. All calculatins assume dr / Since dr can easily be made f the rder f 0.1 mile, dr/r = 2 x 10*5. r *» 0 This same technique can be used t cmpare tracking devices with fixed statins. Cnsider a device that tracks a satellite ver 90 f angle with an accuracy f 0.1. Then d9/ = 0.l/90 = fr dt/t = Lw-Height Cverage T cmplete an installatin, giving cverage dwn t the hrizn, tw lwangle installatins are necessary. The beamwidths n these installatins are 30 x The reduced fan width permits a pwer reductin by a factr f fur (t 250 kw). The range required at these angles is less, s the pwer can be further reduced. 100 kw has been selected fr the detectin transmitter and 100 kw fr the ttal ranging pwer. Due t the necessity fr measuring at lw elevatin angles a frequency f U50 Mc/s has been selected. Phased arrays will be used t btain the lw-angle cverage. A range f 2000 miles n a 0.1 m^ target (50 percent prbability), r 95 percent prbability n a 1 n? target at the same range, will be btained. Site Selectin The prblem f site selectin is that f finding islands separated prperly t minimize feedthrugh and large enugh fr prper antenna installatins. The first installatin shuld be such as t (a) minimize time between launch and detectin, (b) prvide cverage n lw-inclinatin satellites, and (c) cver satellites t extreme ranges. 20

24 G T meet cnditin (a), the installatin shuld be apprximately 90 frm the present installatin in the U.S. In the Pacific, this places the statin in the regin f 150 t l80 E. Of the U.S.-cntrlled islands in the regin nly Pnape is large enugh fr a receiver installatin (but is nt large enugh t prvide the required separatin between transmitter and receiver). Truk-Pnape cvers a line within 7 f the equatr and will intercept early standard Russian-satellite passes. The criterin used fr site selectin f a new statin is that the site shuld fit int a plan t prvide eventual first-pass detectin f all lw-perid and 2k-hur-perid satellites. An installatin n r near the equatr apprximately 90 in lngitude frm the U.S., having a range capability f apprximately 30,000 miles, will prvide (a) early detectin, (b) equatrial cverage, and (c) cverage n 24-hur satellites. frequency f Operatin While the general requirements f a space surveillance system are relatively lncntrversial, sme f the details are subject t questin. One such prblem.s that f frequency f peratin. The prblem is cmplicated by several factrs, it the higher frequencies, bservatins can be made with much greater accuracy.nd less truble frm electrn cluds. At the lwer frequencies the accuracy is.ess. The cst and time required t perfrm a given task are als significantly wer. At the lwer frequencies (belw 200 Mc/s) the system nise figure is limited y sky nise. At the higher frequencies (450 Mc/s) it is determined by receiver ise. Fr calculatins f receiver sensitivity, a system temperature f 1000 K as been assumed fr the 150 Mc/s case and 250 K has been assumed fr the 450 Mc/s ase. Due t the difficulties in btaining better nise figures ecnmically, id due t antenna cabling lsses at higher frequencies, the system temperature Duld remain at abut 250 K. Absrptive catings are effective in reducing the detectin range f devices 'specially mn-static) perating at high frequencies. At lwer frequencies lere the wavelength is large cmpared t the size f the reflecting bject abjrptive catings are ineffective. On the ther hand electrn cluds are mre easily seen at the lwer frelencies. Hwever, blackuts due t bursts wuld nt appear t be serius at equencies f 150 Mc r abve. A burst, t be effective, wuld have t be curately psitined clse t the detectin device and wuld be effective fr time duratin f minutes. If an enemy were t fire a device fr the purpse f acking ut receptin temprarily he might better attempt t hit the detectin device d thereby black it ut altgether. Zl

25 Feed-thrugh The principal prblem with a high-perfrmance cw system is that f feedthrugh. The present Naval Space Surveillance System is a tristatic system having statins spaced far enugh apart t utilize the attenuatin due t the scatter link. In the scatter regin the radiatin reductin fr every dubling f the distance is apprximately 26 db, 20 db mre than in the square-law regin One f the reasns that the present system perates s well is that it is being used t detect high bjects. Fr this requirement the receivers must be placed a lng distance apart t give a reasnable triangulatin cut. At this distance the feedthrugh is small. T use such a system n islands, three islands wuld have t be lcated n a great circle. By using a receiving statin having a ranging capability, nly tw island installatins need be made. If a single island is large enugh, then bth installatins can be made n it. This discussin is aimed at finding the minimum size f an island large enugh fr bth installatins. The pwer rati between transmitter and receiver level will be calculated first. The transmitter perates at +60 dbw. The receiver at abut -I85 dbw fr a ttal excursin f 245 db. Antenna gains at 150 Mc/s are apprximately k-0 and 50 db. With care, the lw-elevatin-angle radiatin can be kept dwn perhaps 45 db n each antenna. By using an antenna lking fr signals at lw-elevatin angles, a signal can be intrduced t the receivers t cancel the lw-elevatin signal received by the vertically lking antennas. This technique will reduce the feedthrugh by 30 db. Filters in the receiver will reduce the feedthrugh by 50 t 80 db. Frm these three effects the cupling can be reduced by 2*4-5 db t 165 t 135 db. The difference between the tw figures is due t the care required t btain the 80-db figure. Much better filters must be used t btain 80 db. Figure 11 shws the attenuatin due t scatter and due t separatin. Figure 12 shws a graph f the cmbined effects. Frm Fig. 12 it is apparent that fr the 135-db islatin figure a minimum distance f 35 miles is required, while fr the 165-db figure a separatin f 90 miles is indicated. It shuld be recgnized that a cnsiderable separatin between transmitter and receiver has an advantage with respect t detecting an bject cated t minimize reflectin. While the reflectin at zer angle f incidence can be made small, the reflectin at large angles f incidence remains cnsiderably larger. 22

26 rh rh cr tn rh JH nl a, cu </> C <J 4. id c rh J si3ai03a Z5

27 in i i Pi 2 [0 GO LU _J S 0) u 0) T3 CD u in 1 1 N 1 M r-< O OJ ^ CD CO sn3ai33a O O O O ' CM <d- CO 00 CM CM CM CM CM 24

28 .--. If the effective size f an bject is decreased by a factr f 4000 bymeans f absrptive material a system designed fr a detectin range f 30,000 miles will have its range reduced by a factr f abut 10. Saturatin One f the advantages f a detectin system that uses a thin-fan beam is that the signal stays in the beam fr nly a shrt time. The system then has much time during which it can detect ther bjects. Fr instance, in the present Naval Space Surveillance System the average time used in a satellite bservatin is apprximately ne secnd. At each receiving site an bservatin n a single satellite will ccur perhaps twice per day. The time used per satellite then is tw secnds. A maximum f apprximately 40,000 bjects culd be tracked. Under these cnditins it is very unlikely that tw bjects will appear in the system simultaneusly. In the event they d, the system will chse ne and will nt be cnfused unless the tw bjects appear in the same dppler slt. Since 160 slts exist, this pssibility is unlikely. By having a phase meter assciated with each dppler slt a maximum f 6,400,000 bjects culd be bserved. Since this is a much larger number than is expected, a much mre mdest number f phase meters can be used. Since it may be expected that the mn will be detected fr a cnsiderable duratin f time, it is desirable that a minimum f tw satellites be measured simultaneusly. The use f three sets f phase meters allws a maximum f 120,000 bjects befre saturatin ccurs. Ten thusand bjects can be handled with a safety factr f 12. COST AND TIME SCALE The csts f varius systems are given in the table belw. Transmitter csts are based n the cst f the Lake Kickap site, with apprpriate adjustments. Receiver csts are based n the csts f present statins, again with apprpriate adjustments. The csts are based n a system capable f detecting bjects t 30,000 miles. This cst is suitable fr detecting 24-hur satellites. Fr satellites at lwer heights the small 1-mw statins can be used, at reduced cst. The prrated cst f perating a U.S.-based system is estimated as 1.1 millin; the cst f perating verseas is estimated as 3-0 millin, annually. The time scale is based n the experience gained in building the present system. A transmitter site takes apprximately ne year t build. Cmb filters take abut the same amunt f time. Antennas and ther items can be cmpleted mre rapidly. Since the rutine cntracting prcess requires 6 mnths, a 150- Mc/s installatin culd be cmpleted in the U.S. in 18 mnths. Fr verseas, six mnths f additinal time shuld be prgrammed. Fr the 450-Mc/s system, an additinal fur mnths shuld be prgrammed. 25

29 Table 2 Cst f System Items 500 kw Lake Kickap 1 mi antenna 2 raw U.S. 2 mi 150 Mc/s Island antenna 2 mw 450 Mc/ U.S. 2 mi antenna Detectin Transmitter Detectin Antenna Ranging Transmitter Ranging Antenna Public Wrks Pwer Statin Supprt Engineering nne nne Subttal Detectin Array (1 mi antenna) Interfermeter Cmb Filter Phase Electrnics Ranging Electrnics Public Wrks Supprt Engineering Subttal TOTAL Lw Angle Cverage Cmplete 100 kw at 450 Mc (2 lks) GRAND TOTAL

30 Time Scale: U.S.: 150 Mc/s - 18 mnths Island: 150 Mc/s - 2k mnths U.S.: 450 Mc/s -22 mnths Island: ^50 Mc/s - 28 mnths. The ttal cst f R and D, prcurement and installatin t prvide a cmplete system with a range t 30>CXX) miles and lw height cverage t 85 nautical miles ver a span f 1500 miles is 32.9 millin if cnstructed in the U.S. Yearly perating csts f 1.1 millin assume integratin with the present system. The increased cst fr peratins verseas assume independent data prcessing and cmputatinal capability. This initial unit wuld be cmpatible with pssible future extensins and grwth requirements in that high precisin angle measurements culd be added and traffic capacity increased as might be required. 27