Outline Outline 1 Basics Introduction Basic Physics Introduction to Underwater sound INF-GEO4310 2 Sonar Theory Sonar types Roy Edgar Hansen Position Estimation Signal processing Department of Informatics, University of Oslo 3 Sonar Applications Fish finding September 2012 HUGIN AUV Mapping Imaging 4 Summary

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 1 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 2 / 77

Sonar Basics Sonar Basics Introduction Outline History of 1 Basics Introduction Basic Physics Underwater sound If you cause your ship to stop and place the head of a long tube in the water and place the 2 Sonar Theory outer extremity to your ear, you will hear ships at Sonar types a great distance from you. Position Estimation Leonardo da Vinci, 1490 Signal processing In 1687 Isaac Newton wrote his Mathematical 3 Sonar Applications Principles of Natural Philosophy which included Fish finding the first mathematical treatment of sound HUGIN AUV In 1877 Lord Rayleigh wrote the Theory of Mapping Sound and established modern acoustic theory Imaging 4 Summary Image from wikipedia.org.

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 3 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 4 / 77 Sonar Basics Introduction Sonar Basics Introduction History of SONAR The masters in sonar SOund Navigation And Ranging A lot of activity after the loss of British passenger liner RMS Titanic in 1912 English meteorologist Lewis Richardson, April/May 1912: Patent on Iceberg detection using acoustic echolocation in air and water German physicist Alexander Behm 1913: patent on sounder Canadian engineer Reginald Fessenden, 1914: Demonstrated depth sounding, underwater communications (Morse Code) and echo ranging detecting an iceberg at two miles range French physicist Paul Langevin and Russian immigrant electrical From wikipedia.org. engineer, Constantin Chilowski 1916/17: US patents on ultrasonic submarine detector using an electrostatic method

Image from wikipedia.org.

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 5 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 6 / 77

Sonar Basics Introduction Sonar Basics Introduction Similar technologies Literature

Course text: sonar_introduction_2012.pdf Course presentation: sonar_presentation_2012.pdf SONAR = Sound Navigation And Ranging Xavier Lurton, An introduction to underwater acoustics = Radio Detection And Ranging Springer Praxis, First edition 2002, Second edition 2010 Medical , higher , shorter range and more www.wikipedia.org complex medium I sonar I underwater acoustics Seismic exploration, lower frequencies, more complex medium I side-scan sonar I biosonar, echolocation I Ocean Acoustics Library http://oalib.hlsresearch.com/

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 7 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 8 / 77 Sonar Basics Physics Sonar Basics Underwater sound Basic Physics Underwater sound

Sound is waves travelling in pressure perturbations Acoustics is the only long range information carrier under water Or: compressional wave, longitudal wave, mechanical wave The pressure perturbations are very small The acoustic vibrations can be characterized by Obtainable range is determined by I Wave period T [s] I free space loss and absorption I f = 1/T [Hz] I the sensitivity to the receiver I Wavelength λ = c/f [m] I Sound speed c [m/s] The ocean environment affects sound propagation: I sea surface I seafloor I temperature and salinity I currents and turbulence Underwater sound propagation is frequency dependent

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 9 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 10 / 77

Sonar Basics Underwater sound Sonar Basics Underwater sound Geometrical spreading loss - one way Geometrical spreading loss - two way

The acoustic wave expands as a spherical wave The acoustic wave expands as The acoustic intensity a spherical wave to the decreases with range in reflector inverse proportion to the The reflected field expands as surface of the sphere a spherical wave back to the The acoustic wave amplitude receiver A decreases with range R In homogeneous media, the The intensity I = A2 two way loss becomes In homogeneous media 1 1 1 ∼ = I 2 2 4 1 R R R I ∼ R2

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 11 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 12 / 77 Sonar Basics Underwater sound Sonar Basics Underwater sound Absorption Transmission loss Transmission loss is geometrical spread + absorption Seawater is a dissipative medium through viscosity and chemical Logarithmic (dB) scale: IdB = 10 log10(I) processes A certain frequency will have a certain maximum range Acoustic absorption in seawater is Frequency is a critical design parameter frequency dependent Lower frequencies will reach longer than higher frequencies

f [kHz] R [km] λ [m] 0.1 1000 15 1 100 1.5 10 10 0.15 100 1 0.015 1000 0.1 0.0015

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 13 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 14 / 77

Sonar Basics Underwater sound Sonar Basics Underwater sound The Ocean as Acoustic Medium Reflection and Refraction in Acoustics

The sound velocity - environmental dependency Recall from first lecture on optical Layering and refraction - waveguides imaging The sea floor and the sea surface - scattering The reflection angle is equal to the incident angle Noise sources The angle of refraction is given by Snell’s law sin θ sin θ 1 = 2 c1 c2

The index of refraction n = c2/c1 Snells law can be derived from Fermats principle or from the

general boundary conditions From Wikipedia

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 15 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 16 / 77 Sonar Basics Underwater sound Sonar Basics Underwater sound Refraction and the underwater sound velocity Deep sound velocity variation

Medvins formula:

c = 1449.2 + 4.6T − 0.055T 2 + 0.00029T 3 + (1.34 − 0.010T )(S − 35) + 0.016D The surface layer The sound velocity depends on 3 major parts: The seasonal thermocline Temperature T in degrees Celsius The permanent thermocline Salinity S in parts per thousand The deep isothermal layer Depth D in meters The sound velocity contains information about the ocean environment. Example: T = 12.5 ◦C, S = 35 ppt, D = 100 m gives c = 1500 m/s

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 17 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 18 / 77

Sonar Basics Underwater sound Sonar Basics Underwater sound Sound refraction Underwater sound channel waves are trapped in a guide The sound will refract towards areas of slower speed The energy spreads in one dimension instead of two I ∼ 1/R SOUND IS LAZY Much longer range : Map the effect of the medium on underwater acoustics

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 19 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 20 / 77 Sonar Basics Underwater sound Sonar Basics Underwater sound Coastal variability Coastal variability

Factors that affect sound propagation: The sound is trapped in a The sound velocity profile waveguide The sea surface The boundaries of the waveguide changes the Internal waves properties of the sound Turbulence wave Ocean current

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 21 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 22 / 77

Sonar Basics Underwater sound Sonar Basics Underwater sound Reflection: basic physics Reflection: the sea surface

Characteristic impedance Z0 = ρc 3 I ρ is the density [kg/m ] The characteristic impedance is I c is the sound speed [m/s] a material property The sea surface (sea-air interface) Reflection coefficient (normal Material Impedance Air: Z = 415 6 incidence) Air 415 Seawater: Z0 = 1.54 × 10 Seawater 1.54 × 106 Z − Z Reflection coefficient R(f ) = 0 Clay 5.3 × 106 Z + Z0 6 Z − Z Sand 5.5 × 10 R = 0 ≈ −1 6 Z + Z Transmission coefficient (normal Sandstone 7.7 × 10 0 6 incidence) Granite 16 × 10 The sea surface is a perfect reflector Steel 47 × 106 2Z T (f ) = 0 Z + Z0

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 23 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 24 / 77 Sonar Basics Underwater sound Sonar Basics Underwater sound Reflection: the sea floor (or bottom) Scattering - smooth surfaces

The sea floor (sea-bottom interface) 6 Sand: Z = 5.5 × 10 Scattering from rough surfaces 6 Seawater: Z0 = 1.54 × 10 The sea surface Reflection coefficient The seafloor Z − Z Other scattering sources R = 0 ≈ 0.56 Z + Z0 Volume scattering from fluctuations Sandy seafloors partially reflects, partially transmits Scattering from marine life Estimated reflection coefficient can be used in classification of bottom A smooth surface gives mainly specular reflection type

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 25 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 26 / 77

Sonar Basics Underwater sound Sonar Basics Underwater sound Scattering - rough surfaces Ambient Noise

The ocean is a noisy environment Scattering from rough surfaces Hydrodynamic

The sea surface I Tides, ocean current, storms, The seafloor wind, surface waves, rain Other scattering sources Seismic I Movement of the earth Volume scattering from (earthquakes) fluctuations Biological Scattering from marine life I Produced by marine life Man made

A rough surface gives specular reflection and diffuse scattering I Shipping, industry

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 27 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 28 / 77 Sonar Basics Underwater sound Theory Marine Life and Acoustics Outline 1 Basics Introduction Basic Physics Underwater sound and use acoustics 2 Sonar Theory for echolocation and communication. Sonar types songs are in the frequency From wikipedia.org. Courtesy of NASA. Position Estimation between 12 Hz and a few kHz. Signal processing Dolphins use a series of high 3 Sonar Applications frequency clicks in the range from Fish finding 50 to 200 kHz for echolocation. HUGIN AUV Mapping Imaging

From wikipedia.org. Author Zorankovacevic. 4 Summary

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 29 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 30 / 77

Theory Sonar types Theory Sonar types Active sonar Active sonar

Transmits a signal Transmits a signal The signal propagates towards the The signal propagates towards the object of interest object of interest The signal is reflected by the target The signal is reflected by the target The signal is recorded by a receiver The signal is recorded by a receiver

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 31 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 31 / 77 Theory Sonar types Theory Positioning Passive sonar Range Estimation

Passive sonar only records signals Estimation of time delay (or two way travel time) τ Relate time delay to range cτ R = 2 Sound velocity must be known

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 32 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 33 / 77

Theory Positioning Theory Positioning Range Resolution Pulse forms 1 - active sonar The minimum distance two echoes can be seperated Different pulse forms for different applications Gated Continuous Wave (CW) Related to the pulse length Tp for non-coded pulses Simple and good Doppler sensitivity but does not have high BT cT δR = p Linear Frequency Modulated (LFM) (or chirp) 2 Long range and high resolution but cannot handle Doppler Related to bandwidth B for coded pulses c δR = 2B

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 34 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 35 / 77 Theory Positioning Theory Positioning Pulse forms 2 - active sonar Directivity Different pulse forms for different applications Hyperbolic Frequency Modulated (HFM) pulses Transducers (or antennas or loudspeakers) are directive Long range and high resolution and Doppler resistive The beamwidth (or field of view) of a disc of size D is Pseudo Random Noise (PRN) BPSK Coded CW λ High resolution and good Doppler sensitivity but low efficiency β ≈ D The beamwidth is frequency dependent. Higher frequency gives narrower beam.

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 36 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 37 / 77

Theory Positioning Theory Positioning Bearing estimation - arbitrary Rx positions Bearing estimation - array sensor

Direction of arrival can be calculated from the time difference of By delaying the data from each element in an array, the array can arrival cδt  be steered (electronically) θ = sin−1 L

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 38 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 39 / 77 Theory Positioning Theory Positioning Bearing estimation - array sensor Imaging sonar / beamforming

Direction of arrival from several reflectors can be estimated by Echo location is estimation of range and bearing of an echo (or using several receivers. target) Imaging sonar is to produce an image by estimating the echo strength (target strength) in every direction and range

Algorithm for all directions for all ranges estimate echo strenght in each pixel end end

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 40 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 41 / 77

Theory Positioning Theory Positioning Beamforming defined Beamforming algorithm in time domain

Beamforming Algorithm Processing algorithm that focus the array’s signal capturing ability in a for all directions particular direction for all ranges for all receivers Calculate the time delay Beamforming is spatio-temporal filtering Interpolate the received time series Beamforming turns recorded time series into images Apply appropriate amplitude factor (from time to space) end sum over receivers and store in result(x,y) Beamforming can be applied to all types of multi-receiver : end active, passive, towed array, bistatic, multistatic, synthetic aperture end

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 42 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 43 / 77 Theory Positioning Theory Positioning Imaging sonar resolution Imaging: Performance measures

Range resolution given by Detail resolution pulse length (actually Geometrical resolution - minimum resolvable distance bandwidth) Contrast resolution Azimuth resolution given by Value resolution, echogenicity, accuracy array length measured in Temporal resolution wavelengths Number of independent images per unit time Field of view is given by Dynamic range element length measured Resolvability of small targets in the presence of large targets in wavelengths Sensitivity Detection ability of low level targets

Array signal processing in imaging is the primary topic in INF 5410

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 44 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 45 / 77

Theory Signal processing Theory Signal processing Sonar signal model Active sonar processing

The basic active sonar processing consists of Preprocessing Pulse compression (range) Beamforming (azimuth) Detection Parameter estimation - position Classification

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 46 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 47 / 77 Theory Signal processing Applications The sonar equation Outline The sonar equation is an equation for energy conservation for 1 Basics evaluation of the sonar system performace. Introduction In its simplest form: Signal − Noise + Gain > Threshold Basic Physics More detailed (for active sonar): Underwater sound 2 Sonar Theory SL − 2TL + TS − NL + DI + PG > RT Sonar types Position Estimation SL is source level Signal processing TL is transmission loss 3 Sonar Applications TS is target strenght Fish finding NL is noise level HUGIN AUV Mapping DI is directivity index Imaging PG is processing gain 4 Summary RT is reception threshold

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 48 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 49 / 77

Applications Fish finding Applications Fish finding Echosounders Stock abundance and species characterisation

The echosounder is oriented vertically The target strength is estimated in every range (depth) The ship moves forward to make a 2D map of fish density The target strength is related to fish size (biomass) Different frequencies can be used for species characterisation

From www.simrad.com. Courtesy of Kongsberg Maritime

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 50 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 51 / 77 Applications Fish finding Applications HUGIN AUV Fish detection range The HUGIN autonomous underwater vehicle

Modern echosounders can detect a single fish at 1000 m range. Some fish have a swimbladder (air filled) which gives extra large target strength

From www.simrad.com.

Courtesy of Kongsberg Maritime

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 52 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 53 / 77

Applications HUGIN AUV Applications HUGIN AUV The HUGIN autonomous underwater vehicle Acoustic sensors on HUGIN Free swimming underwater vehicle Preprogrammed (semi-autonomous) Used primarily to map and image the seafloor Runs up to 60 hours, typically in 4 knots (2 m/s) Imaging sonar Maximum depth: 1000, 3000, 4500 m Altimeter Anti collision sonar Doppler velocity logger Subbottom profiler Acoustic communications

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 54 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 55 / 77 Applications Mapping Applications Mapping Multibeam echosounders MBE Example 1 Data collected by HUGIN AUV Multibeam echosounders maps the seafloor by estimating the Maps from the Ormen Lange field range in different direction The peaks are 50 m high The map resolution is determined by the 2D beamwidth and the range resolution

Courtesy of Geoconsult / Norsk Hydro.

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 56 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 57 / 77

Applications Mapping Applications Mapping MBE Example 2 MBE Example 3 Data collected by HUGIN AUV Data collected by HUGIN AUV Maps from the Ormen Lange field Example area with large sand ripples The ridge is 900 m long and 50 m high.

Courtesy of Geoconsult / Norsk Hydro. Courtesy of Kongsberg Maritime / FFI.

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 58 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 59 / 77 Applications Mapping Applications Mapping MBE Example 3 MBE Example 4 Hull mounted MBE 70 - 100 kHz Hull mounted MBE Colour coded seafloor depth 70 - 100 kHz Magic T (Mills cross) layout

From www.simrad.com. Courtesy of Kongsberg Maritime

From www.simrad.com. Courtesy of Kongsberg Maritime

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 60 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 61 / 77

Applications Imaging Applications Imaging Sidescan sonar Sidescan sonar area coverage Range resolution is given by the pulse length (or bandwidth) Sidescan sonar: sidelooking sonar to image the seafloor Along-track resolution is range dependent Typical platform: towfish, hull mounted, AUV An image is created by moving and stacking range lines Typically frequency 100 kHz - 500 kHz Typical range 100 - 500 m

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 62 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 63 / 77 Applications Imaging Applications Imaging Sidescan sonar example principle

Collect succesive pulses in a large synthetic array (aperture) Increase the azimuth (or along-track) resolution Requires accurate navigation - within a fraction of a wavelength Very similar to Synthetic Aperture Radar (SAR)

Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 64 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 65 / 77

Applications Imaging Applications Imaging Synthetic aperture sonar principle Sidelooking Example - very high resolution (SAS) The length of the synthetic aperture increases with range Along-track resolution becomes independent of range Along-track resolution becomes independent of frequency

Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 66 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 67 / 77 Applications Imaging Applications Imaging Resolution matters Example: Fishing boat

Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 68 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 69 / 77

Applications Imaging Applications Imaging Properties in a sonar image Example large scene with small objects

Geometry: Range and elevation Resolution Random variability - speckle Signal to noise Object highlight and shadow

Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 70 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 71 / 77 Applications Imaging Applications Imaging Comparison of sonar image with optical image Comparison of sonar image with optical image

Sonar range: 112 m Sonar range: 73 m Optical range: 4.5 m Optical range: 5 m

Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI Image collected by HUGIN AUV. Courtesy of Kongsberg Maritime / FFI

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 72 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 73 / 77

Summary Summary Outline Summary 1 Basics Introduction Acoustics is the only long ranging information carrier under water Basic Physics Underwater sound Sound velocity variations cause refraction of acoustic waves The ocean is lossy: higher frequencies have shorter range 2 Sonar Theory Sonar types SONAR is used for Position Estimation I positioning I velocity estimation Signal processing I characterisation 3 Sonar Applications Applications:

Fish finding I Fish finding HUGIN AUV I Imaging of the seafloor Mapping I Mapping of the seafloor Imaging I Military 4 Summary

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 74 / 77 RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 75 / 77 Summary Pa˚ Norsk

Engelsk Norsk beam strale˚ beamwidth stralebredde˚ range avstand bearing retning echosounder ekkolodd sidescan sonar sidesøkende sonar multibeam echosounder multistrale˚ ekkolodd

RH, INF-GEO4310 (Ifi/UiO) Sonar Sept. 2012 76 / 77