t\
`Nl
`
`- ,,i
`
`ilffiffi
`
`ffiB"
`
`SERIAL NUMAER
`
`DATE
`
`ct.Ass
`
`-3;1 :
`
`EXAMTNER I_ zs
`b4n"t
`'F, /i{J ( r t,
`
`l^
`
`E;+l+=:= r.;, ir:.t 1.-t*7, i,;.ztig
`<r/<)/
`Jo4
`
`T = Fil; Hi:1*=F:T 14= *O=,pEi";TtrF;, i-*F;T=i-+f:i-.!Ti+= ?-1"
`
`:t-:+!
`
`i i!i:
`
`|
`
`| tz.l,,
`
`z
`
`fl ri j" Pi + + + E: :* * :+ + t 4': + + .*'1. :+ :I- +'-"1: :-;' li. +
`rr.r op€
`
`+'+":7'*i i:*T : i:ji{:=:;: # * :t :ii * * :4': 'ir r': + *
`lf ur'O lE
`
`? :-'i-
`".tt1-4i:i''1zJ
`
`i
`
`i fr
`
`i-.
`
`f'
`
`?
`
`irEf,!-
`
`M
`
`::',:iiiF : +i-j F j ;-- :i i,ji; L: i:-Cili::;f
`!J yes ltr no
`Forelgn prlorlty clrlmed
`35 usc 119 condltlonr mct El vo E no.
`
`verlfl6d and Acknowlcdgsd
`E ::l-=;;:H ijf- r:*i-Ef+*ri-" ELij*"
`H l;+"+,ti- . #i'.ii:,f E:-:i;ii:iTrri EY=T=ti':#
`
`ATTOR
`DOCKET NO.
`
`iiii.l;tiri: :f,ij *:l*'**T 1-i: FF:';'JE=li il-iG
`-t"
`
`FilF: i;ilf,Elj:}TIt: ii4ir:;E tt #3*if tr-1#rTILlili
`
`US. OEFT. ot CottllMrPtt. & TM Ofilco- PTO435L (t'Y' ls78l
`
`PARTS OF APPLICATION
`FILED SEPARATELY
`
`NOTICE OF ALLOWANCE MAILED
`
`,>/s/rt
`
`Amounl Due
`
`ISSUE FEE
`Date Pald
`
`t6fl.r7)
`
`rh-r\qz-
`
`Label
`Area
`
`/
`-/ctatts ALLowED
`PREPAREDJQ R ISSUE/Z /4 /QI
`o"",.,r1ilIl", \ hht/?,,* Total ClalmsS
`!dar€)almI
`>o,*#*,T1ffi^r/
`
`,/
`prt5 orwg.
`,
`
`DRA\['!NG
`Figs. Drwg.
`I
`
`Prlnt Flg.
`
`I
`
`ISSUE
`BATCH
`
`NUMBER v23
`
`I
`
`[*Yf'ffifl$fr{t
`,f,t!il j.riiit d.ff
`,.t\.f,,ltirr'tqil
`-/
`prlmary eSr(in",
`ISSUE CLASSIFICATIqDT
`Subclass
`
`Class3c)
`
`ot?
`WARNING: Ths inlormation disclosed herein may be restricted. Unauthorlzed dleclosure may be
`8111'lIi:l.l*,'""il:'r:J:1,11"i:';L:'r""L';,i"31"?J;,l1,rl;1l,llo,"'l?;n"r,zed emp,oyees
`
`ffit.
`
`11 -H2-01-e-65-0031-1-08-04 0103243018
`I SKP:RF058933346 - 00007 G-UST:RF068e33346
`1 u,,,,r,
`ro'rrro
`w
`IFr
`ilflltflillt tIlIililt Iilillillt lil]illt lllillillllllll llllllllllllllllllll llllllrll llll
`
`tTfTtr
`
`NRM
`
`IPR2020-00038
`MM EX1050, Page 1
`
`

`

`,/
`
`"4?t{
`
`A,FPROVED FOR LICENSE N
`
`pers.
`
`Receiued
`or
`tulailed
`
`,t/4y'*t
`tz/?
`i4'' /f ,
`
`,/,
`
`/
`
`ffi ?836
`
`Entered
`or
`Counted
`
`CONTENTS
`
`PToGRANT Hliy o 5ffi2
`
`1.
`
`2.
`i.
`
`7.
`
`9.
`
`10.
`
`11.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`
`22.
`
`23.
`
`29.
`
`30.
`
`31.
`
`32.
`
`:l
`',ll
`
`lr
`\r
`
`IPR2020-00038
`MM EX1050, Page 2
`
`

`

`lNlT.
`,5le
`Zoi
`
`DITET
`// i*/q r
`//-/,{4/
`1l-rL
`
`POSITION
`CLASSIFIER
`EXAMINER
`TYPIST
`VERIFIER
`CORPS COHR.
`SPEC. HAND
`FILE NflAINT.
`
`INDEX OF CLAIMS
`
`t-v I
`
`iI
`
`Claim
`
`Date
`
`oEoo
`
`51
`
`52
`53
`
`54
`
`55
`56
`57
`58
`59
`60
`
`61
`62
`
`63
`
`64
`
`65
`66
`
`67
`68
`
`69
`70
`71
`
`72
`73
`74
`
`75
`76
`
`77
`78
`79
`80
`
`81
`82
`83
`84
`
`85
`
`86
`87
`
`88
`89
`90
`s1
`
`92
`93
`94
`95
`96
`
`97
`
`98
`99
`00
`
`oc
`E
`
`Date
`
`sYmB0rs
`/................hirctcd
`.. .. ....Allourd
`- (Through rumolal) Canolrd
`+ . . . . . . . . . . . . . . . . hstrictod
`l{ ................l|on-rhclod
`I ..... .. . ...... ..lnttrbrcncr
`A . . . . . . . . . . . . . . . .Applal
`0 ................0bi0c10d
`
`ll
`tl
`4t
`
`Claim
`
`oc
`
`oo
`
`(1
`
`oEr
`
`t
`
`2-
`5 J
`q
`4r 5
`
`6 7 8I
`
`10
`
`11
`
`12
`
`13
`
`14
`'15
`
`16
`17
`
`18
`
`19
`20
`
`21
`
`22
`23
`
`24
`
`25
`
`26
`27
`28
`
`29
`30
`31
`
`32
`33
`
`34
`35
`
`36
`37
`38
`
`39
`40
`
`41
`
`42
`43
`44
`45
`46
`
`47
`
`48
`49
`
`50
`
`IPR2020-00038
`MM EX1050, Page 3
`
`

`

`SEARCH NOTES
`
`Date
`
`Exmr.
`
`SEARCHED
`
`ll -?r -1r
`
`INTERFERENCE SEARCHED
`Class
`Sub.
`Date
`Exmr.
`
`X[r
`
`g)
`(st
`
`It -?'?- at t
`
`^P
`
`IPR2020-00038
`MM EX1050, Page 4
`
`

`

`ilrillfliltil ilil ilil il11 tfl iltil iltil ]il llll
`
`U.S. PATENT APPLICATION
`
`SERIAL NUMUEK
`
`o7 /783,665
`
`rILIN[,
`
`UAIE
`
`I-LAJI
`
`UKUUT AKI UNT. I
`
`10/28/91
`
`367
`
`2202
`
`JAilES G. KELLY, NEWPORT, Rl; R0BERT N. CARPENTER, P0RTSI'10UTH, Rl
`
`:'r :t C 0 N T I N U I N G [J I J [ :k rt :t :t :t,'t :t :t :t :t rt rt :t tt rt * :'c :t :t :t :t
`VERIFIED
`
`:t:tIQft[ I GN/PCT APPL I CAT lQfrl$rk:t:t:t:t:t:t:t:t:t;t:t
`VERIFIED
`
`Fe
`
`oHJGo
`
`.
`
`TILINU TEE
`RECEIVED
`
`$ 630.oo
`
`AI IUKNEY UUUKEI NU.
`
`7 2251
`
`FOREIGN FILING LICENSE GRANTED 11/15/91
`
`SIAIE UK
`COUNTRY
`
`>NEE I J
`DRAtTING
`
`IUIAL
`CLAIMS
`
`RI
`
`I
`
`5
`
`0FF I CE 0F CoUNSEL, BLDG. I I 2T
`NAVAL UNDERWATER SYSTEI'IS CENTER
`NEWPoRT, Rt 02841-50\7
`
`6@u
`
`ldoo
`
`UJ mrrHoD 0F Ac0usTtc pRocEsstNG F0R AC0usrrc lr4AGE cLAsSlFlcATl0N
`HF
`
`This is to certify that annexed heneto is a tnue copy from the reconds of the united States
`patent and Trademark Office of the application as oniginally filed which is identified above.
`By authority of the
`COMMISSIONER OF PATENTS AND TRAOEMARKS
`
`Date
`
`Certifying Of+icer
`
`IPR2020-00038
`MM EX1050, Page 5
`
`

`

`0i 783665
`
`PATENT APPLICATION SERIAL NO.
`
`U.S. DEPARTMENT OF COMMER.CE
`PATENT AND TRA-DEMARK OFFICE
`FEE RECOR-D SHEET
`
`n$:rff0li,r 1 l. /CI:l/!:t {}l??81{,*s
`
`0ri*c,[]1.4, 0?0 101
`
`6lft ' {iCIL:l I 7t}IS1
`
`PTO-1 5 5 6
`(s/87)
`
`IPR2020-00038
`MM EX1050, Page 6
`
`

`

`DEPARTMENT OF THE NAVY
`OFFICE OF COUNSEL
`NAVAL UNDERWATER SYSTEMS CENTER
`NEWPORT, Rt 02441-!5047
`
`0i ?83SS5
`
`IN REP(Y REFEF IO:
`
`587 0
`Ser 100OC/4O3L
`N.C. 7225L
`24 Ocl- 1991
`
`Commissioner of Patents and Tradernarlcs
`Washington, Dc 2o23L
`
`Applicant: JAMES G. KELLY ET AL
`Fors METHOD OF ACOUSTIC PROCESSTNG
`FOR ACOUSTIC IMAGE CLASSIFICATION
`
`$630.00
`
`....$
`
`Sir:
`Transmitted herewith are the above-identified papers constituting
`a Patent Application filed by the Department of the Navy on
`behalf of the above-named applicant.
`The total filing fee has been computed in accordance with the
`following formula:
`Basic Application
`Basic SIR fee (See
`Additional Fees:
`1-. Total number of claims in excess of
`.....$
`20 tirnes $10.00
`2. Number of independent claims minus
`......$
`3 times $so.oo
`3. Filing multiple dependent clairns
`.....$
`$L00.00 per application..
`Total filing
`fee..
`.$530.00
`Kindly charge the aforementioned total filing
`fee and a'ny
`additional fees to Deposit Account No. 04-081-4.
`^WJL,7'elnll
`Re;qectfuIIy,
`PRITHVI C. LALL
`Attorney of Record
`Reg. No. 26L92
`
`fee. .
`attached Request for SfR)
`
`Encl:
`(1) Application papers
`
`IPR2020-00038
`MM EX1050, Page 7
`
`

`

`ot 783665 k
`
`N.C. 7225L
`
`5"0 /
`
`TO ALL WHOM IT MAY CONCERN
`BE IT KNOWN THAT (1) JAMES G. KELLY and (2) ROBERT N.
`CARPENTER, citizens of the United States of America, employees
`of the United States Government and residents of (1) Newport,
`County of Newport, State of Rhode Island and (2) Pprtsmouth,
`County of Newport, State of Rhode Island have invented certain
`new and useful improvements entitled as set forth above of
`which the following is a specification:
`
`}TICHAEL F. OGLO
`Reg. No. 20464
`Naval Underwater Systems Center
`Newport, Rhode Island
`o284L-5O47
`Telephone No. (401) 84L-4736
`FAX (401) 84L-L23L
`
`IPR2020-00038
`MM EX1050, Page 8
`
`

`

`ffi
`
`,i;
`
`II :::":":: ::",::::-"-"CESSING FOR ACOUSTIC IMAGE CLASSTFTCATION
`
`il
`i t ltil
`srArEMENr oF G.'ERNMENT rNrEREsr
`the inven.*
`l' ll
`actured and
`by or for the covernment of the United States of America
`ll "="U
`t"" covernmental purposes without the paYment of any royalties
`ll
`ll tn"r.on or therefor.
`il
`
`il
`
`'L ll
`BACKGR.,ND oF rHE rNVENrroN
`ll frl Field of the -""."at.-
`/il
`,he present invention relates generally to acousti-ca1
`ll
`l'
`n""""ssors and more particularly to a method of acoustic
`ll
`ll nrocessing for acoustic irnage classification.
`, :, ll Q) Description of the Prior Art
`Underwater systems capable of performing the function of
`I' ll
`ll target classification typically reJ-y on acoustically derived
`ll *."=-rements of the physical extent of the targets or false
`ll'
`tarSets as part of the classification process. The
`ll
`ll mea=urements, gathered from spatially separated sensors
`ll ,"r,.r.ting sprit beams, are processed to form an acoustic image
`ll "t the object in question. characteristic parameters of the.
`
`1 2 3 4 5 5 7 I 9
`
`10
`
`11
`
`L2
`
`l_3
`
`L4
`
`15
`
`15
`
`L7
`
`18
`
`19
`
`20
`
`2L
`
`22
`
`il
`
`il
`
`il
`
`ll
`
`"il'll
`
`IPR2020-00038
`MM EX1050, Page 9
`
`

`

`object (length, width, etc. ) are then extracted from this
`acoustic image and used to determine the classification of the
`object
`split beam processing is one conventional approach to the
`image formation process. Offset phase center beams
`(left/right, up/down) are used to generate angle estimates of
`threshold crossings produced from a sum beam output. The
`underrying assumption behind the sprit beam processor is that
`onry one scatterer exists in each range bin of the acoustic
`image. Detecting the return from that scatter, the processor
`estj-mates the off-boresight angle to that scatterer. The
`estirnated scatterer locations are then processed further to
`generate the irnage. The assumption of a single scatterer per
`range bin places a linitation on the usefulness of this
`approach. Many objects of interest in the classification
`problem can have significant cross range extent. The singre
`scatterer model does not hold for these objects and the
`resulting inage may be difficult
`to classify.
`
`SUMMARY OF THE TNVENTION
`Accordingly, an object of the present invention is to
`provide a method of acoustic processing for acoustic image
`classification of an underwater target.
`Another object of the present invention is to provide a
`method of acoustic processing for acoustic image classification
`that, processes the scatterer returns to achieve a cross-range
`
`L 2 3 4 5 6 7 I 9
`
`10
`
`L1
`
`L2
`
`13
`
`L4
`
`15
`
`t'6
`
`L7
`
`L8
`
`19
`
`20
`
`2L
`
`22
`
`23
`
`24
`
`25
`
`26
`
`IPR2020-00038
`MM EX1050, Page 10
`
`

`

`that processes the scatterer returns to achieve a cross-range
`extent.
`Other objects and advantages of the present invention will
`become more obvious hereinafter in the specification and
`drawing.
`In accordance with the present invention, an acoustic
`waveform of known energy is transnitted into a test volume of
`water. The test volume of water is divided into a number of
`test ceLl-s. Each test cell contai-ns one scatterer that
`reflects the transmitted waveform. A portion of each of the
`reflections is then received by an acoustj.c receive array.
`There exists a scattering coefficient for each test ceII to
`describe the reflection in terms of the transmitted waveform.
`A scattering coefficient for a particular test cell will vary
`depending on the type of target. The scattering coefficient is
`regarded as a (zero-mean and cornplex) random variable with a
`covariance matrix presumed known a priori. Accordingly, an
`estimate of the set of scattering coefficients may be used to
`classify a target. A database is provided with known
`covariance matrices relating the test cells for the volume of
`water for a number of known possible targets. Using the
`covariance matrices, the transmitted waveform, and the
`reflected waveform, estimates of the scattering coefficients
`are calculated for the possible targets. These estimated
`scattering coefficients for all the possible targets along with
`the transmitted waveform are then used to generate estimated
`
`1 2 3 4 5 6 7 I 9
`
`10
`
`11
`
`L2
`
`l-3
`
`L4
`
`15
`
`16
`
`L7
`
`18
`
`19
`
`20
`
`2L
`
`22
`
`23
`
`24
`
`25
`
`25
`
`,-l
`
`IPR2020-00038
`MM EX1050, Page 11
`
`

`

`refLectj-ons for all the possible targets. The estimated
`refrections are correlated with the actual refrected waveform
`at the acoustic receive array where a high correLation is
`indicative of the target to be classified.
`
`BRIEF DESCRIPTION OF THE DRAWING
`,/
`y'rhe figure is a schematic representation of an acoustic
`system transmitting and receiving acoustic information with
`respect to a test vorume of water used by the method of the
`present invention.
`
`DESCRIPTTON OF THE PREFERRED EMBODIMENT(sI
`The active sonar detection probrern is usuaLly formurated
`as a binary statistical hypothesis test. Under the nuII
`hypothesis Ho, the sensors, outputs consist of noise alone,
`while arternative hypothesis H, consists of noise prus the
`sensors, response to the presence of a,target. The active
`sonar crassification problem can be similarry formulated, but
`ps a multiple hypotheses test. There are at least two
`al-ternative hypotheses because, in addition to the, signal
`produced by the target, another signar can appear in the sensor
`outputs. This other signal results from the sensors, response
`to the pressure fierd scattered from an object, distinct from
`the target assumed under Hr; however, it resembles in some
`significant way the signal produced by that of Hr. rt may be a
`false target or the rear target at some orientation other than
`
`I
`
`1 2 3 4 5 5
`
`I
`
`8 9
`
`10
`
`l)i,l ('"
`i')
`7t
`
`11
`L2 l,; i t,
`13t
`
`'i
`
`I
`
`.
`
`I
`
`L4
`15
`16!
`
`il J
`
`1
`
`\..
`
`L7
`
`18
`
`19
`
`20
`
`2L
`
`22
`
`23
`24i
`25i
`26
`
`I
`
`l
`
`IPR2020-00038
`MM EX1050, Page 12
`
`

`

`that under Hr. In general, there may be several of these
`alternative tttarget-liketr signals, each corresponding to a
`distinct class of scattering objects (either natural or man-
`nade). Thus, the classification problem must be fornulated as
`a multiple hlpothesis test with a nuII hypothesi-s Ho and
`alternative (exhaustive and mutually exclusive) hypotheses H,,
`where i = L, 2, ...,
`f and f > 1. Mathematically, Ho and Hi can
`be expressed as follows:
`
`Ho:x(t) =a(t) ;
`
`ur:x( t) =.i, ( e) +a ( r) ,
`
`(1al
`
`(1b)
`
`where t is the time interval of observation where T, and T, are
`the endpoints of the observation interval;
`x(t) = [xr(t), xz(t), ..., xu(t)]t is the N x l complex vector
`of the N sensor outputs;
`Sr(t) = [sil(t), srz(t) , ..., sn(t) ]t is the N x J- complex vector
`of the signal components for the N sensor outputs due to the
`presence of the it object; and
`n(t) = tnr(t), nz(t)
`, nr.r(t)lt is the N x 1 complex vector
`of the noi.se components of the N sensor outputs, and [ ],
`denotes the transpose of any vector.
`Referring now to the drawing, a schematj-c representation
`is shown of an acoustic system for transmitting and receiving
`acoustic information with respect to a test volume of water 10
`
`,il
`
`2 3 4s
`
`11
`
`
`6l
`11 z :.t J t"'
`I
`tlP -'f ;1, -r .il,irl
`.,,, --._--,. ,
`
`iitl i
`
`I
`
`L
`
`-,._",*-
`
`ei
`I
`
`IIt
`
`_-
`
`10
`r ),
`
`/ rrf ), ,'
`
`L
`
`L2
`
`r '.1; L3 ' ). ,)
`
`J,
`
`,1,
`
`L4
`/ 15r
`16
`
`L7
`f'i LB tl
`Lg ,{i
`
`.,u
`
`,/
`r'i
`
`,.,
`
`../r
`
`)
`
`20
`2L ;"'
`22
`
`23
`
`IPR2020-00038
`MM EX1050, Page 13
`
`

`

`used to describe the method of the present invention. Test
`vorume 10 is divided into M test cerls where it is assumed
`that an i-th target from the set of r hlpotheses resides. The
`method of the present invention is a novel approach to crassify
`the i-th target. For ease of description, only one test cell
`at a rocation m wirr be discussed. Hereinafter, this may arso
`be referred to as the m-th test cerl. However, the method of
`the present invention is applicabre to all of the test cerrs in
`the test volume 10.
`A transmitted acoustj-c waveform f (t) is transrnitted from a
`sonar platform 20 beginning at t=.9 for a duration of T.
`Transmitted waveforn f(t) is designated by the lead line
`indicated by reference numeral 11. Transmitted waveform 11
`strikes the m-th test ceII and is scattered. A portion of this
`scattering designated by the lead line indicated by the
`reference numeral 15, is detected by an acoustic receive array
`l-7 comprised of N sensors. For ease of description, only one
`sensor at a location n witr be discussed. Hereinafter this may
`arso be referred to as the n-th erement. However, the method
`of the present invention is applicable to alr of the sensors in
`the receive array L7. Typicarry, each sensor in array 17 is an
`omni-directional acoustic sensor. The n-th element of receive
`array 17 receives a portion of the scattering from the m-th
`test cell which is observed beginning at a time T, where Tr > T.
`
`1 2 3 4 5 6 7 I 9
`
`J
`
`10
`11 r'r
`L2
`
`13
`
`t4
`
`15
`
`16
`
`L7
`
`18
`
`19
`
`20
`
`2L
`
`22
`
`23
`24''t r '
`
`25
`
`/
`
`IPR2020-00038
`MM EX1050, Page 14
`
`

`

`observation of the reception will last untir time Tr. The time
`deray between transmission to the m-th element and reception by
`the n-th element is defined as ro..
`Each test ceII also has a scattering coefficient a_
`associated therewith which is an indication of how the m-th
`test eeII will scatter an impinging acoustic waveform. The
`value of the scattering coefficient a. is a function of the
`type of target residing within test volume 10. Accordingly,
`for every i-th target at an m-th test ce1l, there exists an
`individual a, which is true for one and only one of the I
`hypotheses. Since each test ce1I contains, at most, one
`scatterer, the set of scattering coefficient ar, m = 1 to M,
`is indicative of the target to be classified. The n-th test
`ceIl also has a velocity component, Vi-, in a direction along
`the line designated by the lead line indicated by reference
`numeral L1. This velocity component is due to the assumed
`rerative uniform velocity between the sonar platforn 20 and the
`test volume for the i-th hypothesis. rf no retative motion
`exists between the sonar pratform 20 and the test volume for
`the i-th hypothesis, the velocity component vi- is pero for ar1
`il=l , 2, ...rM.
`For the i-th target to be crassified, a signar received by
`the n-th element of receive array 17 during the tirne T, to T,
`from all of the M test ce1ls is defined as si,(t). The signal
`received by receive array 17 is a function of the transmitted
`waveform 11 appropriatery time-delayed, dirated (or contracted)
`
`il
`
`,i{, tJ
`
`rl
`
`JJ
`
`!i
`
`I
`
`I I Il
`
`-
`
`,.i
`
`/",
`
`1
`
`.2
`
`3 4 5 6 7 I 9
`
`10
`
`11
`
`L2
`
`13
`
`L4
`
`15
`
`15
`
`L7
`
`18
`
`19
`
`20
`
`2L
`
`22
`
`23
`
`24
`
`25
`
`26
`
`IPR2020-00038
`MM EX1050, Page 15
`
`

`

`and weighted by the individual- scattering coefficients ai..
`The dilation or contraction is due to the relative motion
`between the sonar and the test volume, and may sometimes be
`referred to as relative motion temporal adjustment in this
`specification and appended claims.
`The signal received by the n-th elenent si,(t) riray be
`modeled as:
`
`1 2 3 4 5 6 7
`
`\il> I
`I"i I
`q-
`gl
`| ',i I.-i,'l
`
`)
`
`L0
`11 13r
`y''ii, ; /'s, ,/t
`
`L2
`
`'a,,
`t
`,,t,.1 1,/
`
`,/
`
`13
`L4 l' '",
`
`15
`15 , i.'i t-/
`
`L7
`
`18
`
`M
`
`m.1
`
`I ar^f (pr, ( t-c-,) )
`
`or in vector form as:
`
`E t=E i(t'l a,
`
`where lt(t) =
`
`, t"i
`
`tl -j
`
`lEi ( c) ) *=f (F ,, ( t-t.r) )
`
`where p r^=*
`
`(21
`
`(3)
`
`(1)
`
`(s)
`
`is the farniliar expression for the Doppler effect
`dilation or contraction of the received waveform
`speed of sound propagation in the medium, and T
`delay for the transmitted acoustic waveform ao .I
`cell location and back to the m-th sensor.
`
`of temporal
`and C is the
`is the tirne
`m-th test
`
`IPR2020-00038
`MM EX1050, Page 16
`
`

`

`The followj-ng assumptions pertain to equation (3):
`a) The scattering from a given test cell is frequency
`independent, at reast in the band of transnitted waveform l-1.
`b) Each of the M disjoint test cerrs contains at most one
`scatterer.
`c) The transmit waveform 11 is known and has finite
`energy:
`
`r
`
`(f ( r) )2dt=E.
`
`17l
`
`ft can be assumed that, for each of the I hypotheses, both
`the test vorume 10 and test cerl geometry are specified a
`priori. Thus, the number of cerrs M and the position of each
`test cell are specified constants for all i = Lr2r...f. As the
`search for the actuar target progresses, the test vorume 10
`will shift in some prescribed way through the insonified volume
`during each transmission cycle.
`By adopting this Eulerian (fixed set of test cells)
`approach, the randomness of the signal s.(t) is characterized
`in the modet of equation (2') solely by the statistical
`properties of the (zero-mean and complex) coefficienls {am}
`m=1 to M, i.e., by a M x M complex covariance matrix &i, i =
`L,2,...rI. &i is not restricted to be of fu1l rank; therefore,
`the model of equation (2) al1ows for an arbitrary
`degree of coherence among the random waveforms scattered from
`each m-th test ceIl.
`In practice, the dinensions and
`
`I'
`
`It'
`
`ff
`
`1 2 3 4 5 5 7
`
`i ))[
`
`ll
`
`I 9
`
`l
`
`10
`
`11
`.)
`L2 .)
`
`13
`t4
`
`15
`
`16
`
`L7
`
`ii
`
`t.-.;
`
`18
`
`.,/
`19')
`,{""d
`_) .) _
`2Ol) ) )
`2L r) t-t
`.i1'\
`22r
`
`./
`
`i
`
`23
`
`24
`
`IPR2020-00038
`MM EX1050, Page 17
`
`

`

`decomposition of test volume 10 and estimates of e wiII be
`deterrnined directly from both experimental data and
`calculations from scattering models of hypothesized targets.
`The resulting M x l{ covariance matrices for each of the I
`hypothesized targets may then be stored in a database.
`In order to determine which of the f hypotheses is
`residj-ng within the test volume 10, a Bayesian approach is used
`to generate a classification scheme. This approach is
`described in great detail by H. L. Van Trees in Detection.
`Estimation, and Modulation Theorv- Volume 1 - pages 24-34, John
`Wiley and Sons, fnc., New York, L969, which is herein
`incorporated by reference. rt is sufficient for the method of
`the present invention that the Bayesian approach is applied to
`yield estimates Q of the scattering coefficient vector Bi, { =
`1 to I. The estimated scattering coefficient vector Er is then
`used in the signal model set forth in equation (3) to yield
`estimates of the signal component vector !i,, i = 1 to I.
`Finally, the estimated signal component vector gi i" correlated
`with the output vector x(t) of receive array L7. A high
`correlation wourd be indicative of the target residing within
`test volume 10.
`Specifically, the estimated scattering coefficient vector
`Ei, i = 1 to I is determined by:
`
`10
`
`1 2 3
`
`4
`
`.t"
`
`',,'
`
`I
`
`5 6 7 Is
`
`i"
`I10 1 ," ,-t
`11 l
`
`I {
`
`L2
`
`r-lq'
`
`I
`1
`
`13
`)l
`L4 it
`15
`15 ,'.'i
`IL7 I
`
`18
`
`19
`
`20
`
`2L
`
`22
`23
`
`')
`,,.ir 71n
`
`IPR2020-00038
`MM EX1050, Page 18
`
`

`

`ut:-, ffrut * tfi[i'fr, ( r) x( t) dt)
`
`17,
`
`were,
`
`ID is an M x M identity matrix,
`No is a noise spectrum level due to background
`acoustic or electronic noise,
`o is an M x I'{ waveform correlation matrix described
`by the equation
`
`, r rV.-
`
`#l::d,(e).8, (t)dt
`
`(8)
`
`"
`It(t) is
`element such that
`
`a N x M matrix having an n-th and m-th
`
`1 2 3 4 5 6
`
`7 I 9
`
`l'>1, .' t& I it
`
`,lrLiO
`
`t.Ej ( t) I *,=f (9 1(t-t ,r) )
`
`(e)
`
`i{
`
`'r.1
`
`,\t
`
`10
`11 f
`72
`13 r itl.
`L4
`
`L5
`
`16
`
`L7
`
`18 utr
`
`19
`
`ana S
`
`'.\
`
`(t) is the complex conjugate-transpose of
`
`&(t).
`Thus, the estimated scattering coefficient vector Ei, i = 1 to
`r is a function of the known covariance matrices, the size of
`receive array L7, and the transnitted acoustic waveform 11
`appropriately tine delayed and ditated (or contracted). As
`mentioned above, the estimated signar component vector $ is
`then generated by the signal model of equation (3). The
`correlation required by the Bayesian structure may be defined
`
`11
`
`IPR2020-00038
`MM EX1050, Page 19
`
`

`

`1
`
`'{r t"
`
`{
`
`as the likelihood ratio
`Tiextr GI;:d( t)s j (t) dtl
`
`( 10)
`
`for i = 1 to I. Since scattering coefficient vectors are
`unigue and unrelated for each of the I hypotheses, a high
`correration wourd be indicative of the estimated scattering
`coeffj-cient vector for the target to be classified.
`The advantages of the present invention are numerous. The
`method of the present invention inprements a si-gnal noder that
`is a function of the scattering coefficient vector and the
`transmitted waveform. The signar moder aIlows for scattering
`from objects with any amount of cross range extent and is not
`limited to a single scatterer at each down range rocation. The
`scattering coefficient vector is estimated via a Bayesian
`estimate/correrate structure. The resurting estimated signal
`component vector is then correlated with the actual output
`vector.
`Thus, it will be understood that many additional changes
`in the details, materials, steps and arrangement of parts,
`which have been herein described and irrustrated in order to
`explain the nature of the invention, may be made by those
`skilled in the art within the principle and scope of the
`invention as expressed in the appended claims.
`
`L2
`
`2
`
`3 4 5 6 7 I 9
`
`10
`
`11
`
`L2
`
`13
`!4
`
`15
`
`t-6
`
`L7
`
`18
`
`L9
`
`20
`
`2L
`
`22
`
`IPR2020-00038
`MM EX1050, Page 20
`
`

`

`| ,,\
`
`.'"'1.!t
`
`l! ,j
`
`yi
`
`,,
`
`,l'i 1t1 i.i
`
`')l
`
`,'l
`
`i"...
`
`What is claimed is:
`1. A method of using active sonar in an underwater
`a set of I possible
`
`targets, comprising the steps of:
`transmitting an acoustic waveform f(tr) having finite
`a)
`energty E, over a period of time tr = o to T, into a
`test volume of water, said test volume being divided
`into M test cells, each of said test celLs occupying
`a location m, m = 1 to M, in said test volume, each
`of said test cells at said location m further having
`exactly one scattering coefficient a*, i = 1 to I,
`wherein the i-th target has a scattering coefficient
`vector q such that
`
`arr= la rT ^i
`
`b)
`
`c)
`
`providing a database
`i = 1 to I, for each
`
`of
`of
`
`M x M covariance matrices Q,
`said I possible targets;
`
`receiving an acoustic waveform vector x(tz), over a
`period of time tz = Tr to T, where Tr ) T, with an
`acoustic receive array having N acoustic sensors,
`said recei-ved acoustic waveform vector x(tz) having a
`corresponding signal component vector s(tz) and a
`
`13
`
`IPR2020-00038
`MM EX1050, Page 21
`
`

`

`corresponding noise component vector n(tz), wherein
`said signal component vector s(tz) at an n-th sensor
`of said N acoustic sensors for the i-th target is
`sh (tz) and is represented by a signal model
`
`! a-f (0 ir( t2-t-,) )
`
`M
`
`Itt-L
`
`i=ltoIandn=1toN,
`wherein ro. is a time delay for said transmitted
`acoustic waveform to travel to an m-th test ceII
`Iocatj-on and back to the n-th sensor, B* is the
`Doppler effect of any relative motion temporal
`adjustment in the received waveform and f(pa (tz-r,_) )
`is said transnitted acoustj.c waveform compensated by
`said relative motion temporal adjustment;
`
`d) estimating a scattering coefficient vector e,, i = 1
`to f, for each of said I possible targets according
`to the equation
`
`Ku,trD+,ff,U1-, thl}rt i(q\ x(t2) dt27
`
`wherein,
`ID is an M x M identity matrix,
`No is a noise spectrum leve1 due to backqrround
`acoustic or electronic noise,
`
`t4
`
`.l
`l!
`
`{ II
`
`ji
`
`'' / ,r , , .'
`
`,.)
`t't t)
`s g t7*' ti
`
`i ; {.r 1,.
`
`.'
`
`)t'l
`
`-;i
`
`' l: il
`I I i\,''
`
`tt
`l,I"
`)rt{tt
`
`',,'
`
`IPR2020-00038
`MM EX1050, Page 22
`
`

`

`Q is an M x M waveform correlation matrix described
`by the equation
`
`#l;:il,(t2t Ei(t2t dtz
`
`wherein Ir(tz) is a N x M matrix having an n-th and
`ur-th element such that
`lEr( t) ) *,= f (9 ir( tr-c *,) \
`
`I'J i;"'')l;i
`
`i.l/f
`
`! /
`
`t,)l!,.
`
`I ,... l-l
`, rit
`|
`
`'y,
`
`7 ,t t\A
`
`_)
`
`t ' {
`
`:
`
`t
`l-t
`
`-,.t to
`
`{,r
`
`),li
`
`It"
`
`lt
`
`ana gli 1tr1 is the complex conjugate-transpose of
`4 (tz) ;
`e) estinating a signal component vector g, (tz), i = L to
`I, for each of said I possible targets from said
`estimated scattering coefficient vector ai using said
`signal model; and
`
`f)
`
`correlating said estinated signal component vector
`gi (tr) , i : t- to r, with said received acoustic
`waveform vector x(tz) for each of said I possible
`targets wherein said correration is indicative of the
`target to be classified.
`
`15
`
`IPR2020-00038
`MM EX1050, Page 23
`
`

`

`2- A method of using active sonar in an underwater environment
`to classify a far fierd target using a set of known target
`hlpotheses, comprising the steps of:
`
`transmitting an acoustic waveform of known energy into a
`test vorume of water, the test volume of water being
`divided into test ce11s wherein each test cell
`contains one scatterer that refrect,s the transmitted
`acoustic waveform according to a scattering
`coefficient wherein a unique and unrerated scattering
`coefficient for a test ceII exists for each of the
`known target hypotheses and wherein a known
`covariance rerationship exists between the test cerrs
`for each of the known target hypotheses;
`
`receiving a portion of the refrection from each scatterer
`at an acoustic receive array wherein said received
`portion is a function of: f-) the scattering
`coefficients of the target to be classified and 2)
`the transmitted acoustic waveform appropriatery tirne
`delayed by the time it takes the transrnittea acoustic
`waveform to travel to the scatterer and reflect to
`the receive array inctuding a further temporal
`adjustnent due to the relative motion between the
`sonar and the test volumel
`
`15
`
`t)
`
`I
`
`l'l
`
`,.,
`
`I I
`
`1Ii
`
`IPR2020-00038
`MM EX1050, Page 24
`
`

`

`estimating the scattering coefficient at each test cerl
`for the set of known target hypotheses based upon the
`known covariance relationships and said received
`portion of the reflection;
`
`{
`
`estimati-ng the received portion of the reflection for the
`set of known target hypotheses based upon said
`estimated scattering coefficient and said
`appropriately time delayed transmitted acoustic
`waveform; and
`
`correlating said estimated received portion with said
`received portion for
`the set of known target
`hypotheses wherein a hiqh correLation is indicative
`of the target to be classified.
`
`3. A method according to claim 2 wherein said step of
`receiving begins after said a step of transmitting.
`
`4. A method according to clairn 3 wherein said step of
`estimating the scattering coefficient uses a aayesian
`structure.
`
`5. A method according to claim 3 wherein the scattering
`coefficient at each test ceII is unique for each of the known
`target hypotheses.
`
`L7
`
`/,rI, /t\!
`
`' ii'}"l
`
`IPR2020-00038
`MM EX1050, Page 25
`
`

`

`Navy Case No. 7225L
`
`ABSTRACT OF THE DISCLOSURE
`A method of acoustic processing for acoustic image
`crassification of an underwater target from a set of known
`target hypotheses is provided. An acoustic waveform is
`transmitted into a test volume of vrater containing the target
`to be crassified. The test vorume is divided into test cerrs
`such that each test celI contains a portion of the target or
`scatterer that wirr scatter the transmitted waveform.
`scattering of the transmitted waveform at each test cerl is
`described by a scattering coefficient that is unique for each
`of the targets in the set of known hypotheses. Accordingry,
`the scattering coefficient serves as an identifier of the
`target to be crassified. The rerationship of the set of
`scattering coefficients for each of the'known target hypotheses
`is provided by a known covariance matrix. An acousti-c receive
`array receives a portion of the scattered waveform from each of
`the test cerrs in the test volume. This received portion is a
`function of the scattering coefficients of the target to be
`crassified and the transmitted waveform appropriately tine
`delayed by the time between transmission and receipt. The
`scattering coefficients at each test cerr are then estimated
`for each of the possibre hypotheses. Estimation is, based upon
`the known covariance matrices and the received portion of the
`
`18
`
`1 2 3 4
`
`^
`
`5 6 7 I 9
`
`10
`
`11
`
`L2
`
`l_3
`
`1.4
`
`15
`
`15
`
`L7
`
`18
`
`t9
`
`20
`
`2L
`
`22
`
`23
`
`24
`
`25
`
`26
`
`IPR2020-00038
`MM EX1050, Page 26
`
`

`

`scattering. Then, an estimation is made of the received
`portion of the scattering for each of the possible hypotheses.
`Estirnation of this received portion is based upon the estimated
`scattering coefficients and the time-delayed, transmitted
`waveform. Fina11y, the, estimated received portion of the
`scattering is correlated with the actual received portion of
`the scattering. A high correlation j.s indicative of the target
`to be classified.
`
`2 3 4 5 6 7 I
`
`)
`
`19
`
`IPR2020-00038
`MM EX1050, Page 27
`
`

`

`? f3 b65*
`
`Nav1r Case No. 72?5L
`Declaration and Powqr of Attorney for Patent Application
`As a below named inventor, I hereby declare that
`My residence, post office address and citizenship are stated
`below next to my name,
`I believe I am the original,
`first and sole inventor (if only
`first and joint
`one name is listed below) or an original,
`in-
`ventor (if p1ural names are listed below) of the subject matter
`whj-ch is claimed and for which a patent is sought on the inven-
`tion entitled
`METHOD OF ACOUSTIC PROCESSING FOR ACOUSTIC TMAGE CLASSTFICATION,
`the specification of which
`(check one) /X/ is attached
`
`/ / was filed on
`Application SerieT No.
`and was amended on
`I hereby state that I have reviewed and understand the contents
`of the above-identified specification,
`inctuding the claims, as
`amended by any amendment referred to above.
`I acknowledge the duty to disclose information which is material
`to the examination of this application in accordance with Title 37,
`Code of Federal Regulations, 1.56(a).
`I hereby claim foreign priority benefits under Title 35, United
`States Code, LL9 of any foreign application(s) for patent or
`inventor's certificate
`listed below and have also identified below
`any foreign application for patent or inventorrs certificate having
`a filing date before that of the application on which priority
`is
`claimed:
`Prior Foreign Application(s)
`
`Priority Claimed
`
`as
`
`(Number)
`
`(Country)
`
`Yes
`
`No
`
`(Day/Month/
`Year Filed)
`f hereby claj-m the benefit under Tit1e 35, United States Code, 1-2O
`of any United States Application(s) listed below and, insofar as
`the subject matter of each of the claims of this application is not
`disclosed in the prior United States application in the manner pro-
`vided by the first paragraph of Tit1e 35, United States Code, LL2, f
`acknowledge the duty to disclose material information as defined in
`Tit1e 37, Code of Federal Regulations, 1-.56(a) which occurred between
`the filing date of the prior application and the national or PCT in-
`ternational filing date of this application:
`
`IPR2020-00038
`MM EX1050, Page 28
`
`

`

`(Application Serial
`Number)
`
`{r-nEs D-ffi-)-
`
`(Status)
`(patented, pending,
`abandoned)
`POMR OF ATTORNEY: As a named inventor, I hereby appoint the'
`following attorney(s) and/or agent(s) to prosecute this application
`and transact all business in the Patent and Trademark office con-
`nected therewith. (List name and registration numbe