Page 13 - Spring 2019
P. 13

wavelengths per unit distance in the direction of propaga-

tion.) A conventional delay-and-sum beamformer, where

the weights are set to unity, is opti.rnal in the sense that the

output signal-to-noise ratio is a maximum for an incoherent

noise field. However, when the noise field includes coher-

ent sources (such as interference), then an adaptive beam-

former that is able to maximize the output signa.l-to-noise

ratio by applying a set of weights that are complex numbers Supevlav pevlormanve o1 adapllve heamlavmlng .4 submarine ta...» avray am no amen weak
. . . . . 5’ am, ruuluecluml .s lzedxuuriu, am. he burin andn Mr In . iunll ii ...i
L5 1-|T‘P1€1'“€“t9d-This has the 91?€Ct°fSt9€“hS 3 “"11 In the 3...“. man». t..§...7.n..t.s sII9nr:: s.a'..§..s, ,...n'.ns. .5..y"t.:7.."........... ‘ms
direcfion of any unwanted jmerference (3 J-lmmer)‘ staring at nullxln tn. dlvaninnx nl tmularuue, lndllcillutz ms malvstx av.n.ys.w...aisa
Figure 1 shows a comparison of the frequency-wave num- Figure 1. Left: estimated frequency-wave number puwer spectrum
ber spectrum for an actual underwater acoustic field sensed f”' f7 “"5 WW)’ U/_’h)"1"’Ph""95 “SW8 ‘hf? 5""V5""'”"“l/,’9'1“9"5}”‘1"’
by an array using the conventional weight vector (lefl) and ;""t'"‘l7e:;l"f‘;"'f’"g melhui 7-"e”:""X'm“';Vfl;'t1“e_"C}_; “E”: "T15
an adaptive weight vector (right). The adaptive beamformer h” :16; If tmg" fleguefy Zf E army" Fig 'Fsm” W lglggegffl
suppresses the side lobes and resolves the various contribu- an S1 E u fm an a up we mmfwm" mm ergumn '
tions to the acoustic pressure field, which are shown as surfac-

es (ridges) associated with towed array self-noise (structural M ‘iv i it

waves that propagate along the array in both axial directions, I
i.e., aft and forward); tow-vessel radiated noise observed at,  
and near, the forward end-fire direction (i.e., the direction
of the longitudinal axis of the array) for the respective direct ,,,w,,,,,,,,,,,,,,,,_m,,,,,,,,,V
propagation path and the indirect (surface-reflected) mul- “"‘°$‘:_“"_j’,"'i‘§“"::."";‘_'"“° ,
tipath; and three surface ship contacts. The adaptive bea.rn- mm»: beamlormellssenshlrve I “‘ j fit
former better delineates the various signal and noise sources l" “’m'"  '" S'‘"'' V
that compose this underwater sound field (Ferguson, 1998).
TWed_Am}, shape mimafion Fi'gu1z2.Lzft:u tuwed army isstmightbefare the submarine maneu-
. . . . net. but it is bowed during the maneuver Right: variation with bear-
When deep, submarines rely exclusively on their passive so- , . .
_ ingana time ofthe uutput ofthe beumfarmer befvre, aanng, ana after
Mr Wstems to 561159 the uhderwam 5°“"d field ft“ “diam the change afheuding ufthe submarine. The total abservatiun periud
Wise from Ships underway and antisuhrharine active sonar is 20 minutes. anal the aata are real. when the army is straight, the
transmissions. The long-range search sonar on a submarine cuntuct appears an ane l7eui‘irlg befure the maneuver and an unuther
consists of 3 thin flexible neutrally buoyant strem-“er fined bearing after the maneuver when the submarine is U71 its new head-
with a “He may of hydmplmnex which is towed behind mg ana the array became: straight again. estimating the array shape
‘he submaflm Towed mm overcome two problems am (at the eaaniinates ofthe sensurpusitwns in tea ahmensians) aanng
_ _ _ the heudingchunges enables eantaet to be maintainea thraugnuut the
limit the performance of hull-mounted arrays: the noise of submarine maneuver Fmm Fe sun (1 993“)
the submarine picked up by the sensors mounted on the hull ' rgu '
and the size of the acoustic aperture being constrained by the
limned length ofthe submamm to reprocess the hydrophone data so that the shape of the
_ _ _ _ array could be estimated at each instant during a submarine
Unfortunately, submarines cannot travel in a straight line . .
_ _ _ maneuver when the submarine changes course so that it can
forever to keep the towed array straight, so the submarine is . . . .
‘K n _ _ head in another direction (Ferguson, 1990). The estu-nated
deaf when it undertakes a maneuver to solve the left-right . .
_ _ _ shape had to be right because a nonconventiona.l (or adap-
ambiguity problem or to estu-nate the range of a contact by .
_ _ _ _ tive) beamformer was used to process the at-sea data. Uncer-
triangulation. Once the array is no longer straight but bowed, . . . . .
_ . tain knowledge of the sensor positions results in the signal
sonar contact is lost (Figure 2). b . . .
eing suppressed and the contact being lost. The outcome is
Rather than instrumenting the length of the array with head- that submariners maintain their situational awareness at all
ing sensors (compasses) to solve this problem, the idea was times, even during turns.
Spring 2ai9 | Assnuunsa 1-may | 1 1

   11   12   13   14   15