Underwater Communication Employing High-Sensitive Magnetic

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/.

This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI10.1109/ACCESS.2020.3026623, IEEE Access

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.

Digital Object Identifier 10.1109/ACCESS.2017.DOI

Underwater Communication EmployingHigh-Sensitive Magnetic Field DetectorsMAURICE HOTT1, AND PETER A. HOEHER2, (Fellow, IEEE)1Faculty of Engineering, Kiel University, D-24143 Kiel, Germany (e-mail: [email protected])2Faculty of Engineering, Kiel University, D-24143 Kiel, Germany (e-mail: [email protected])

Corresponding author: Maurice Hott (e-mail: [email protected]).

This EU.SH project is funded by the European Union – European Regional Development Fund (ERDF), the Federal Government and LandSchleswig-Holstein, Germany under Grant LPW-E/1.2.2/1075.

ABSTRACT Magnetic communication is receiving significant interest in RF-challenging environments.Particularly in underwater environments and underground wireless sensor networks, magnetic communica-tion is an emerging research area. In this paper, a new approach for magnetic underwater communicationis presented and evaluated. In this approach, the receiver coil of a conventional magnetic inductioncommunication system is replaced by a high-sensitive low-noise wideband magnetic field sensor. Thisconcept enables a good detection sensitivity and, under certain conditions, an extended communicationrange. Most magnetic field sensors are small compared to equivalent search coils and offer a high bandwidth.Hence, they can be assembled in order to provide multiple-input multiple-output processing. Based onsuitable channel modeling, trade-offs between system parameters are analyzed and the channel capacity isderived. Analytical results are supported by a prototype implementation. Potential application scenarios arestudied, where emphasis is on mobile applications.

INDEX TERMS Communication channels, Communications technology, Magnetic field measurement,Magnetic sensors, Mobile communication, Underwater communication, Underwater technology.

I. INTRODUCTION

OCEANS cover more than 70 percent of the Earth’ssurface, but only a fraction of the seabed is already

explored. At least locally, new insights can be gained throughunderwater robotics and sensor networks. Digital communi-cation is a key technology for data exchange between mobilenodes and/or fixed installations, both in near/mid-range andfar-range applications. A bottleneck is the lack of fast yetrobust wireless underwater communication techniques. Mostcommercial wireless underwater communication systems areeither based on acoustic waves or optical links. Acousticmodems offer the largest communication range among theseoptions, but the bandwidth is small and the latency maybe problematic. Reliable acoustic communication is difficultparticularly in shallow waters. Multipath propagation causesa large delay spread, and mobility a significant Dopplerspread. Optical communication is strongly affected by thevisibility between transmitter and receiver, as well as bydaylight in shallow water depths.

A promising alternative for near/mid-range communica-tion is to exploit magnetic fields for digital communica-tion [1]–[8]. Unlike radio waves, which are radiative and

hence subject to multipath propagation, operation is in thenon-radiative near field. Typically, coils are used both atthe transmitter side as well as at the receiver side. Thetransmitter coil is driven by a modulated AC current. Bymeans of magnetic induction (MI), the information-carryingsignal is reproducible at the receiver side. Magnetic near-field communication has distinctive features. Compared toacoustical communication, advantages of MI include a scal-able bandwidth over a wider spectral range, a negligiblepropagation delay, and less susceptibility to surroundings.This technique is insensitive to water turbidity (includingsurf zones, tidal flow, and rivers), water depth, reflections bymaterials and surfaces, interference by sound and light, aswell as Doppler spread and delay spread. In the non-radiativenear field, critical issues like multipath propagation andfading are negligible. Due to the scalable bandwidth, multi-user communication and networking are more efficient. Inautonomous underwater vehicle (AUV) and remotely oper-ated vehicle (ROV) applications, acoustic communication isaffected by vibrations of the thrusters. Coils can be integratedinto non-metal hulls, whereas acoustic transducers protrudeinto the water. Furthermore, fast-decaying magnetic fields are

VOLUME 4, 2016 1



Maurice Hott et al.: Underwater Communication Employing High-Sensitive Magnetic Field Detectors

more environment-friendly for marine mammals than soundwaves. In contrast to optical underwater communication, in-tervisibility and ambient light are no bottleneck. The pointingand acquisition problem of collimated optical beams does notexist in MI communication. Narrowband relaying is simplebecause it can be achieved with passive resonant circuits[1]. In summary, communication in harsh environments likeshallow water, turbulent water, and harbors is feasible. Mag-netic communication is one of the few techniques that workbelow and simultaneously above the sea surface (or aboveand simultaneously below the sea bottom). Magnetic fieldscan be used jointly for communication, localization, energyharvesting, and for tracing metal objects. These synergeticeffects are beyond the scope of this contribution.

However, a fundamental physical drawback of MI is thedistance law. In the non-radiative near-field regime, the ratiobetween the received power and the transmitted power isproportional to d−6, whereas in the radiative far-field regimeit is d−2. Both figures hold for line-of-sight propagation in airor purified water. In salty waters with conductivity σm > 0,attenuation is even worse for MI. This problem can be relaxedto some extend by using large-size coils (or low frequencies),for example. However, these are bulky and hence not suitablefor mobile applications.

In [9], we have presented a novel magnetic communicationapproach: The receiver-side coil of a classical MI systemhas been replaced by a high-sensitivity low-noise widebandmagnetic field sensor. Related work has independently beenpublished in [10], where the emphasis is on diver commu-nication employing a speech codec, and in [11], [12], wheremagnetic field detectors are used for localization purposes inair and fluids. This cheap and reliable approach offers severalbenefits and, under certain circumstances, an extended com-munication range. In our prototype receiver, several high-sensitive wideband low-noise anisotropic magnetoresistance(AMR) magnetic field sensors are implemented. The re-sistance of AMR sensors, typically operated in a Wheat-stone bridge, is a function of an external magnetic field.Among the advantages is that these magnetic field sensorsare very small and suitable for multiple-input multiple-output(MIMO) processing. It should be noted that the measurementprinciple of magnetic field sensors is not based on MI. Hence,the system performance, as well as the trade-offs betweenrelevant system parameters, are different for the magneticcommunication approach under investigation. In this work,detailed underwater channel modeling for magnetic commu-nication employing magnetic field detectors is provided. Theterms magnetic field sensor and magnetic field detector areused alternately. Novel contributions include the followingaspects:

• The magnetic underwater channel is examined both fora homogeneous water column as well as for shallowwater depths and surfaces.

• The trade-offs between relevant system parameters arestudied analytically.

• The SNR enhancement by using several sensors is stud-ied.

• Our analysis is supported by an experimental prototype.The remainder is organized as follows. In Section II, the

channel model and the trade-offs between relevant systemparameters are studied analytically. A prototype implemen-tation demonstrating the feasibility of the system concept ispresented in Section III. Potential applications are suggestedin Section IV. Finally, conclusions are drawn in Section V.

II. THEORETICAL ANALYSISA. NEAR-FIELD COMMUNICATIONA fundamental advantage of magnetic near-field communi-cation is the insensitivity to reflection/scattering/diffractioncaused by the seafloor and other obstacles. However, it isimportant to note that this characteristic is only valid in thenon-radiative near field [13]. The terms “near field” and “farfield” mentioned before describe different spatial areas inantenna technology that surround the radiator. The spatialareas are usually divided into three areas:• The reactive near field is the area in the immediate

vicinity of the radiator. In this region, transmission isnon-radiative. In the case of a coil, a magnetic field isobtained. Magnetic field lines are closed.

• The radiating near field is the transition field, also re-ferred to as the Fresnel region.

• The far field, also called the Fraunhofer region, is thearea in which an electromagnetic wave propagates inspace as a wave. In the far field, magnetic commu-nication suffers from multipath propagation and theassociated fading at the receiver side.

The boundaries between the individual regions depend onthe wavelength λ and are radiator-dependent. Frequently,the reactive near field is defined for distances up to λ/(2π)around the radiator and the far field is said to start at about4λ. The distance

dnf =λ

2π=

cm

2πf(1)

at which the non-radiative near field passes to the radiatingnear field is an essential parameter with respect to systemdesign. This boundary depends on the frequency f of thetransmitted signal and the speed of light cm in the transmis-sion medium.

The speed of light is a material-dependent parameter andcan be calculated as follows:

cm =1√µε

=1√

µ0µrε0εr=

c0√µrεr

, (2)

where c0 describes the speed of light in vacuum, µthe absolute permeability, µr the relative permeability,µ0 = 4π · 10−7 H/m the vacuum’s absolute permeability,ε the absolute permittivity, εr the relative permittivity andε0 = 8.85 · 10−12 F/m the vacuum’s absolute permittivity.

The relative permeability of water is similar to the one ofair and is defined as µr = 1. The relative permittivity in

2 VOLUME 4, 2016




water is different from the one of air and is a frequency-, temperature-, and salinity-dependent parameter. For purewater with a temperature of 10 ◦C and a signal frequency be-tween 0 Hz and 1 GHz, the permittivity is about εr, water ≈ 84[14]. The relative permittivity for water drops down slightlyfor increasing temperature and/or salinity (shielding effect)[15]. However, this effect is very small for variations intemperature and salinity of common water types, so that itcan be neglected here. For example, given a signal frequencyof f = 100 kHz and a speed of light of cwater = 3.27 ·107 m/sin pure water with εr = 84 yields dnf = 52 m. This valuerepresents the maximum non-radiative near-field communi-cation distance for the considered underwater channel andtransmission frequency, independent of transmitter powerand receiver sensitivity.

B. CHANNEL MODELING FOR A HOMOGENEOUSWATER COLUMNAccording to Biot-Savart’s law, for a planar coil with Nw

windings, excitation current I , and azimuth angle α withrespect to the main coil axis, at a distance d the magnetic fluxdensity can be expressed in non-conductive environments(like air) as

B =µ0NwIr

2 cos (α)

2(r2 + d2)3/2, (3)

where r = (ro + ri)/2 with ro being the outer and ri theinner radius of the coil. Note that for d � r the magneticflux density is proportional to 1/d3.

All known wireless communication principles are subjectto high signal attenuation under water, which makes com-munication at reasonable data rates very difficult. In thecase of magnetic communication in conductive media suchas seawater, an additional signal attenuation is caused byeddy currents. Eddy currents produce a negative moment thatchanges the transmitted moment and the field distribution.The signal attenuation is a function of the skin depth δ. Forgood conductors, the well-known approximation for the skindepth applies:

δ =1√πfµσ

. (4)

For bad conductors, e.g. seawater, a more general expressionmust be used [16]:

δ =1

2πf

√µmεm2

(√1 +

σ2m

(2πf)2ε2m− 1) . (5)

The equation shows that the signal frequency f , the magneticpermeability µm, the permittivity εm, and the conductivity ofthe transmission medium σm contribute to the skin depth. Fora homogeneous conductive transmission medium, the eddy-current-related signal attenuation can be expressed by theterm

e−√d2+r2

δ . (6)

0 5 10 15 20 25 30 35 40 45 500

50

100

150

200

250

300

350

400

450

500

Distance in m

Mag

netic

loss

indB

10 kHz Air10 kHz Tap water (0.057 S/m)10 kHz Baltic Sea (1.2 S/m, mean)10 kHz Kieler Förde (1.1 S/m, measured)100 kHz Air100 kHz Tap water (0.057 S/m)100 kHz Baltic Sea (1.2 S/m, mean)100 kHz Kieler Förde (1.1 S/m, measured)

FIGURE 1. Magnetic loss as a function of distance for different homogeneousmedia.

Altogether, (3) can be extended as

B =µ0NwIr

2 cos (α)

2(r2 + d2)3/2e−√d2+r2

δ . (7)

This manifests that the magnetic flux density decreases withincreasing signal transmission frequency and/or conductivity.The conductivity of water is mainly given by the number ofions and increases with salt concentration. The main typesof ions in seawater are chlorine and sodium. The salinity fordifferent water types like tap water, lake water, or seawatervaries widely. In general, a distinction is made betweenfreshwater (≤ 0.1 %), brackish water (between 0.1 % and1 %), and saline water (≥ 1 %). The Baltic Sea has an averagesalt content of ≈ 0.8 %, which is far below the average saltcontent of the oceans with≈ 3.5 %. The relationship betweenthe salinity and the electrical conductivity has been studied in[17].

To specify the decay of the magnetic field between trans-mitter and receiver in the presence of a certain homogeneousmedium, the magnetic loss is derived from (7) as

MLhom = 20 log10

(2(r2 + d2)3/2

e−√d2+r2

δ

)− 20 log10

(2r3

e−rδ

)

= 20 log10

(2(r2 + d2)3/2e−

rδ

2r3e−√d2+r2

δ

). (8)

It represents the loss of the magnetic field amplitude indB for an optimal alignment (α = 0) depending on thetransmitter coil radius and the skin depth of the mediumand is independent of the number of windings and the coilcurrent. The ML is normalized to the magnetic field strengthin the physical center of the coil, constituting d = 0.

Fig. 1 depicts the magnetic loss for magnetic communica-tion as a function of distance for the conductivity of air, tapwater, Kiel Fjord water, and Baltic Sea water. The magneticloss is plotted for the signal frequencies 10 kHz and 100 kHz.

VOLUME 4, 2016 3




For d � r, in non-conductive environments like air themagnetic loss is proportional to d3 and is independent ofthe signal frequency. For conductive media, i.e. σ > 0, themagnetic loss further increases with higher conductivity andsignal frequency. The plot also shows that the influence ofincreasing conductivity can be compensated by reducing thesignal frequency.

C. MAGNETIC LOSS FOR NEAR-SURFACECOMMUNICATIONConventional models for calculating the performance ofmagnetic communication systems are based on the assump-tion of a homogeneous, infinitely extended transmissionmedium. This assumption neither holds near the surface nornear the bottom. There are approaches in the area of magneto-inductive underground communication, which model the sig-nal propagation in heterogeneous media, see for example[18]. Frequently, however, only the distance-dependent vari-ations of the medium between the sender and receiver areconsidered. For the case of underwater communication, thevertical variation of the medium is important, whereby the in-terface between water and air represents the greatest changein the medium properties. If the transmitter or receiver islocated at a relatively shallow depth below the water surfaceand the communication distance is sufficiently large, themagnetic field lines partly pass through the air. Consequently,these are less attenuated, which leads to an underestimationof the communication range by most existing models.

In [19], a scenario is examined where a coil is placedexactly on the boundary between two media. The coil issimplified as a magnetic dipole in the referenced publica-tion. Consequently, this approximation is valid for distancesmuch larger than the coil radius (d � r). Moreover, thisequation supports only a 2-layer model and is not able totake into account a coil placed at different depths belowthe sea surface. For this reason, we applied a finite elementmethod (FEM) analysis using COMSOL Multiphysics. Inthis context, a deep water scenario and a shallow waterscenario are investigated.

1) Deep Water ModelIn deep water, where the water depth is much larger thanthe communication distance, the properties of the seabed arenegligible. This makes the examination of a 2-layer modelsufficient. We assume that the main coil axis is aligned inparallel to the sea surface. Different coil depths are emulated.The scenario is illustrated in Fig. 2.

Fig. 3 shows the corresponding results of the FEM calcula-tion for selected communication depths. For seawater, a con-ductivity of σ = 1.2 S/m is assumed. The signal frequencyis selected as f = 100 kHz. As a reference, the magneticloss for homogeneous air and seawater with an infinite ex-pansion is shown. The results support the assumption that themagnetic loss gradually increases with depth under the seasurface. It can be seen that all curves for the different waterdepths initially follow the same function. When the distance

d

depth

Water (µr2 = 1, εr2 = 84, σ2 = 1.2 S/m))

Air (µr1 = 1, εr1 = 1, σ1 = 0 S/m))

Tx Rx

FIGURE 2. Deep water 2-layer model for communication near the watersurface.

0 5 10 15 20 25 30 35 40 45 500

50

100

150

200

250

300

350

400

450

500

Distance in m

Mag

netic

loss

indB

Air 0 m 1 m 2.5 m 5 m10 m Baltic Sea

FIGURE 3. Magnetic loss as a function of distance d in deep water fordifferent depths given the scenario in Fig. 2. The signal frequency is specifiedas f = 100 kHz and the conductivity is σ = 1.2 S/m.

is about twice as large as the depth below the sea surface, themagnetic loss curve converges to the d3 distance law. Thesefindings confirm that known homogeneous channel modelsunderestimate the channel when the depth is less than thetransmission distance.

2) Shallow Water Model

In shallow waters, the communication distance can besmaller than the water depth. In this case the properties ofthe seabed should be considered in the model for an accuratecalculation of the magnetic loss. Hence, a 3-layer FEM modelis developed, which represents the media air, water and theseabed. Fig. 4 illustrates the scenario and shows the chosenmagnetic and electrical parameters of air, water and theseabed. The main coil axis is aligned in parallel to the watersurface and is located at half the water depth. The signalfrequency is f = 100 kHz.

Fig. 5 shows calculated magnetic loss based on the FEManalysis as a function of distance for different water depths.As expected, the magnetic loss in deeper waters increases

4 VOLUME 4, 2016




Water (µr2 = 1, εr2 = 84, σ2 = 1.2 S/m))

Air (µr1 = 1, εr1 = 1, σ1 = 0 S/m))

Seabed (µr3 = 1, εr3 = 10, σ3 = 0.05 S/m)

d

depth

depth

Tx Rx

FIGURE 4. Shallow water 3-layer communication model.

0 5 10 15 20 25 30 35 40 45 500

50

100

150

200

250

300

350

400

450

500

Distance in m

Mag

netic

loss

indB

Air 1 m 2.5 m 5 m 10 m 15 mBaltic Sea

FIGURE 5. Magnetic loss as a function of distance d in shallow water fordifferent depths given the scenario in Fig. 4. The signal frequency is specifiedas f = 100 kHz and the conductivity is σ = 1.2 S/m.

faster with distance. In direct comparison to the results ofthe deep water 2-layer model in Fig. 3, it can be seen thatthe seabed slightly reduces the magnetic loss compared toseawater due to its greater skin depth. Overall, the use of amulti-layer model can improve the accuracy of the resultsin both scenarios. It should be noted that the assumptionof a homogeneous infinitely extended medium in low com-munication depths and shallow water scenarios leads to anunderestimation of the channel.

D. TRADE-OFFS BETWEEN SYSTEM PARAMETERS

In this section, the influence of adjustable system parameterson the proposed magnetic communication system is investi-gated. The adjustable parameters are separately investigatedfor the transmitter and receiver side. The analysis serves asa design guide for magnetic communication systems. Themaximum communication distance under consideration isset to 50 m. This corresponds approximately to the non-radiative near-field boundary in water for a signal frequencyof 100 kHz.

FIGURE 6. SNR at the sensor output as a function of the radius of thetransmitter coil and the distance between transmitter and receiver.

1) SNRThe signal-to-noise ratio (SNR) is a fundamental parame-ter in communications, besides the bandwidth. The SNR isdetermined by the ratio between the received signal powerand the noise power, or the corresponding root mean square(RMS) voltages at the detector output:

SNR =PS

PN=U2

RMS, S

U2RMS, N

. (9)

To calculate the RMS voltage of the wanted signal URMS, S,Eqn. (3) can be used, which gives the magnetic flux densityas a function of the link distance for a homogeneous media.By additionally considering the sensitivity S of the magneticfield sensor and the amplifier gain g, the RMS voltage at thedetector output can be determined:

URMS, S = gSBRMS. (10)

Since URMS, S ∼ B applies, PS ∼ B2 holds. This clearlydemonstrates that the power decreases by d−6.

The RMS noise voltage URMS, N can be calculated bysumming of all noise sources. For this purpose, all contribut-ing circuit components must be taken into account. For theproposed communication system, the AMR sensor and theinstrumentation amplifier are the main noise sources on thereceiver side:

URMS, N = gERMS, AMR + ERMS, AMP. (11)

Note that the noise of the AMR sensor is amplified by theinstrumentation amplifier.

2) Transmitter DesignIn Fig. 1 the magnetic loss is shown for different frequenciesand conductivities as a function of the distance. The magneticloss describes the attenuation of the signal which occurs in

VOLUME 4, 2016 5




the link between transmitter and receiver. To ensure that theSNR at the receiver side is still sufficient, the amplitude ofthe generated signal must be sufficiently large. In the case ofmagnetic communication, this signal is generated by a coil.The properties of the coil and the current flowing throughthe coil indicate the strength of the magnetic field generated,as shown in (3). For a better understanding the number ofwindings, the current through the coil and the coil radius canbe substituted in the magnetic moment

m = NwIπr2. (12)

To maximize the magnetic moment generated by the coil, thenumber of turns, the current, and/or the area enclosed by thecoil can be increased. The magnetic moment is proportionalto the square of the coil radius. However, it should be notedthat additional power is required to keep the current throughthe coil constant, as the length of the conductor, and thus theelectrical resistance increases. The same applies when thenumber of windings is increased. The coil current does notaffect the other parameters of the magnetic moment.

To enable high coil currents, the transmitter coil can beoperated in a resonant circuit. For this purpose, a componentwith capacitive characteristics, such as a capacitor, is con-nected in parallel (parallel resonant circuit) or in series (seriesresonant circuit) with the coil. This provides a frequencywhere the reactances of coil and capacitor cancel each otherout XC = XL, which is called electrical resonance. As aresult, the maximum coil current I in a series resonant circuitat the resonant frequency fr is limited only by the Ohmicresistance of the coil and the internal resistance of the source:

I =V√

R2 + (XL −XC)2. (13)

In this way, strong alternating magnetic fields can be gener-ated efficiently. The so-called quality factor is an expressionof the ratio between the reactance of the coil or capacitor andthe Ohmic resistance at the resonant frequency: Q = XL/R.However, a low resistance is at the expense of bandwidthwhich is defined as ∆f = fr/Q. The upper and lowercut-off frequency is determined by the 3 dB limits of thequality curve of the resonant circuit. This results in a trade-offbetween coil current and bandwidth.

Fig. 6 shows the SNR at the receiver side as a functionof the radius of the transmitter coil and a given distancebetween transmitter and receiver. The amplitude of the coilcurrent is set to I = 7 A and the number of windings toNw = 100. The analysis is performed for air respectivelypure water with a perfect alignment (α = 0) and a bandwidthof ∆f = 5.4 kHz. It is noticeable that the SNR decreaseswith an increasing radius for small transmission distances.This can be explained by the fact that the distance fromthe coil axis to the current-carrying conductor increases. Forlarger distances (d � r), the enlargement of the radius hasthe effect of an increasing SNR. It should be considered thatthe frequency of the alternating field in conductive materials

FIGURE 7. SNR at the sensor output as a function of the receiver bandwidthand the distance between transmitter and receiver.

also contributes to attenuation, this must also be taken intoaccount in the system design.

3) Receiver DesignThe receiver performance is limited by the magnetic fieldsensor sensitivity and the total noise of the detector circuit.The total noise results from the noise contributions of allcircuit components and is a bandwidth-dependent value. Thenoise characteristic of a component is defined by the noisevoltage density EN in the data sheet, which has the unit V√

Hz.

The RMS noise voltage ERMS, N can be calculated by usingthe equivalent noise bandwidth (ENBW) ∆f of the receiver:

ERMS, N = EN√

∆f. (14)

ENBW can be obtained by the filter 3 dB frequency f3 dB andthe steepness s of the used filter type, which compensates thefilter characteristic: ∆f = s · f3 dB. For a brick wall filter,∆f = f3 dB applies. For example, if a pole filter with onepole is used, the value is defined as s = 1.57.

Fig. 7 shows the expected SNR on the receiver output asa function of distance and the communication bandwidth.As expected, the SNR decreases with increasing bandwidthand distance. Therefore, the bandwidth efficiency should betaken into account in the system design, which describesthe information rate that can be transmitted over a givenbandwidth for the communication system.

To describe the performance of a magnetic field sensor,the so-called sensibility Bmin is taken as a measure. Thesensibility describes the sensitivity of the sensor in relationto its inherent noise. In this paper, the sensibility is definedby the magnetic flux density at which the RMS signal level isequal to the RMS noise level (SNR = 0 dB). Fig. 8 shows therelationship between communication range and sensibilityfor different water types. It is important to note that themaximum range refers to an SNR of 0 dB and is not a general

6 VOLUME 4, 2016




10−15 10−14 10−13 10−12 10−11 10−10 10−9 10−8 10−7 10−6100

101

102

103

Sensibility Bmin in T

Ran

gefo

rSN

R=

0dB

inm

AirTap water (0.057 S/m)Baltic Sea (1.2 S/m, mean)Kieler Förde (1.1 S/m, measured)

FIGURE 8. Communication range for an SNR of 0 dB as a function of thesensibility Bmin of the magnetic field detector for selected water types.

upper limit for the sensitivity of the sensor. If the overalldetector would be able to operate at an SNR of less than 0 dB,for example by means of low-rate channel coding, the curvesin Fig. 8 would move towards larger ranges.

4) Channel CapacityThe channel capacity represents the maximum informationrate (in b/s) at which information can be transmitted quasierror-free over a channel. The calculation of the channel ca-pacity varies depending on the type of transmission channel.For the magnetic communication between a coil and an AMRbased receiver, the channel distortions can be modeled by ad-ditive white Gaussian noise (AWGN) in good approximation.Thus, the single-input single-output channel capacity can becalculated by using the Shannon–Hartley theorem:

C = ∆f · log2

(1 +

PS

PN

)= ∆f · log2

(PS + PN

PN

), (15)

where ∆f is the channel bandwidth, PS the received signalpower and PN the noise power. This equation holds forGaussian-distributed channel input symbols. The channelcapacity for a given distance can be determined by measuringthe total signal, which is a superposition of the wantedsignal plus noise (numerator) and by measuring the noiseonly (denominator). Fig. 9 shows the channel capacity as afunction of SNR for different bandwidths. The blue curvedepicts the channel capacity for a bandwidth of 5.4 kHz. Thered curve represents the channel capacity for a bandwidthof 1 MHz, which corresponds to the maximum bandwidthof the AMR sensor used in the prototype. It can be seenthat the channel capacity for an SNR of 0 dB correspondsto the bandwidth and grows for an increasing SNR. Thus,the red curve is shifted by the factor 185 with regard to theblue curve. It should be remembered that the receiver noisealso increases with the bandwidth (see Fig. 7) and the SNRdecreases for that reason. With MIMO signaling the channelcapacity can be boosted.

−50 −40 −30 −20 −10 0 10 20 30 40 50100

101

102

103

104

105

106

107

108

SNR in dB

Cha

nnel

capa

city

inbi

tspe

rse

cond

Bandwidth ∆f = 1 MHzBandwidth ∆f = 5.4 kHz

FIGURE 9. Channel capacity as a function of SNR for a bandwidth of 5.4 kHzand 1 MHz.

5) SNR EnhancementAn improved SNR at the receiver side leads to fewer trans-mission errors and thus increases the channel capacity of thetransmission channel. In addition to boosting the transmis-sion power, the noise can also be reduced by using a smallerbandwidth for a better SNR. Another approach to minimizethe receiver noise exploits the statistical independence of thenoise. While the signal power increases linearly by averag-ing over n sequences, the noise power increases only withthe square root of the number of averaged sequences. Thisimproves the SNR according to the central limit theorem bythe factor

n√n

=√n. (16)

If the channel noise is very small, approximately all noisecontributions at the receiver are produced in the detectorcircuit. In this case, spatial averaging of the measured signalis a preferable alternative to time averaging. This can beachieved by using n > 1 magnetic field sensors for signaldetection. The symbol rate remains the same for physicalaveraging, whereas the symbol rate decreases for time-basedaveraging. Due to the small footprint of the AMR, a largenumber of sensors can be mounted on a mobile-friendlyreceiver circuit for realizing an SNR enhancement. This as-pect is investigated by using the developed prototype systemintroduced in Section III.

III. PROTOTYPE IMPLEMENTATIONIn this section, the suitability of the communication conceptfor underwater environments is verified. For this purpose, asmall-scale underwater communication system is developed(Fig. 10), and measurements are taken in the Kiel Fjord,an area in the southwestern region of the Baltic Sea, andanalyzed.

The magnetic communication approach under investiga-tion is based on a direct measurement of the magnetic field,which is produced by a transmitter coil. The setup includes

VOLUME 4, 2016 7




FIGURE 10. Small-scale magnetic communication experiment setup. Thetransmitter circuit including the coil is shown on the left, the magnetic fielddetector circuit on the right. Transmitter and receiver are pressure-neutralencapsulated.

a planar transmitter coil with a radius of r = 1.6 cm andNw = 10 windings, that operates in resonant mode andis driven by a half-bridge inverter. The resonant frequencyis tuned to 100 kHz. The coil current amplitude is chosenas I = 3 A. The receiver coil of a conventional MI com-munication system is replaced by a custom-made detectorcircuit, which is based on a Sensitec AFF755 high-sensitivewideband low-noise AMR magnetic field sensor with a sizeof about 25 mm2 including packing. The typical sensitivityof this sensor is specified as 15 mV/V

kA/m , which allows measure-ments of weak magnetic flux densities in the frequency rangefrom 0 Hz up to 1 MHz. The sensor signal is amplified by awideband low-noise instrumentation amplifier with a fixedgain-factor of 2000. The developed small-scale prototypeachieves for the resonant frequency an SNR of 12 dB at adistance of 80 cm between transmitter and receiver. For anAWGN channel this corresponds to a channel capacity of21.64 kb/s.

As theoretically shown in Section II-D5, an SNR enhance-ment can be achieved by using multiple sensors for spatialaveraging according to the central limit theorem. Here, theinfluence of the physical number of magnetic field sensorsworking jointly on the SNR is investigated practically. Forthis purpose a detector circuit with four independently oper-ating AMR sensors was developed. The circuit includes fourAMR sensors and four instrumentation amplifiers. The signalof the AMR sensors is amplified by one instrumentationamplifier each. This allows separate measurement data acqui-sition for each sensor. The measuring signal can be digitallysummed and evaluated for a varying number of sensors. Inthe measurement setup, a signal of 100 kHz was generatedwith a transmitting coil. This signal was measured with thedetector array and digitally evaluated.

Fig. 11 presents the power spectrum as a result of the com-bination of the signals from a different number of sensors.The spectrum depicts that the signal power increases by 6 dB

80 85 90 95 100 105 110 115 120−55

−50

−45

−40

−35

−30

−25

−20

Frequency in kHz

Pow

ersp

ectr

umin

dB

1 Sensor2 Sensors3 Sensors4 Sensors

FIGURE 11. Power spectrum measured for sensor array with a maximumnumber of four sensors working jointly.

when doubling the number of sensors. The noise level on theother hand only increases by 3 dB when doubling the numberof sensors. This gives a total SNR gain of 3 dB when doublingthe number of sensors which are working jointly. Since themeasurement results are in accordance with the central limittheorem, the magnetic channel noise can be neglected in theconsidered frequency range. Thus, spatial averaging with alarge number of sensors is as a useful way to enhance theSNR significantly.

IV. POTENTIAL APPLICATIONSIn conventional MI communication, the same coil can beused as both transmitter and receiver (transceiver). Thus,by its basic concept, an MI system offers bidirectional datatransmission. If one coil is replaced by a magnetic field sen-sor, the bidirectionality is lost at the cost of newly obtainedadvantages. However, there are many scenarios in which onlyunidirectional communication is needed and other interests,such as receiver size or weight, have priority. The benefitsof a small, lightweight receiver introduce new possibilities inthe field of underwater communication. But even in scenarioswhere bidirectional communication is required, an additionalintegration of a magnetic field sensor in the transmitter andreceiver circuit can provide further advantages, like an SNRimprovement. Furthermore, a broadband signal detection isenabled. Hence, the received signals are not limited by thetuned resonant frequency of the transmitter coil. In the fol-lowing, two scenarios are presented, which are enabled bythe use of one or several high-sensitive low-noise widebandmagnetic field sensors applied for data reception.

A. DEPLOYMENT BETWEEN STATIONARY AND MOBILENODE(S)Fig. 12 illustrates a communication scenario for coastal andharbor areas, where common techniques have severe prob-lems to establish a reliable data transmission. Here, acousticcommunication struggles with reflections from quay walls

8 VOLUME 4, 2016




FIGURE 12. Scenario 1: Inshore communication using a powerful landtransmitter and repeater node anchored on the seabed. 3D magnetic fieldsensors in the AUVs ensure a permanent downlink. Optionally, a small 3Dtransmitter coil integrated into the AUVs serves as a transmitter for atemporary uplink with short communication range.

and optical communication with poor visibility conditionsin the port basin. A promising solution could be magneticcommunication, especially by employing magnetic field de-tectors. In this case, a large onshore 3D coil with a permanentpower supply can generate a strong magnetic signal. Thereby,magnetic near-field propagation is not affected by the air-water boundary. In addition, a communication node equippedwith a large 3D coil and a magnetic detector is placed inshore,which acts as a repeater. Ideally, this transceiver is suppliedwith permanent power. The AUVs are equipped with a 3Dmagnetic field detector, which enables a continuous down-link. A small 3D transmitter coil with a short communicationrange can additionally be used. This ensures a temporaryuplink, for example when an AUV is located close to acommunication node. Furthermore, this allows an AUV tocontinuously receive data and transmit the collected dataduring the periods when the node is within range of thetransmitting coil. A further advantage by using magneticfield detectors as a receiver is that the transmitter coils canbe tuned on different resonant frequencies. This enables thefull bandwidth for all transmitters. In addition, this allowsthe associated transmitter to be distinguished by knowingthe corresponding frequency without the use of identificationcodes. Besides, this simplifies the realization of localizationalgorithms. The wideband receiver allows all participants toobserve multiple signals simultaneously.

B. DEPLOYMENT BETWEEN MOBILE NODESFig. 13 presents an offshore communication scenario for datatransmission and localization purposes. Here, spatially dis-tributed battery-operated low-power communication nodesare placed on the seafloor. Each communication node pro-vides a magnetic field detector as a receiver and a transmitterunit based on a small 3D coil. The transmitter resonantcircuits are tuned to different frequencies so that the trans-mission frequency is unique for each communication node.

FIGURE 13. Scenario 2: Offshore communication using several low powercommunication nodes anchored on the seabed. The communication nodesand AUVs provide a 3D coil as transmitter and a 3D magnetic field detector asreceiver.

Due to the large bandwidth of the magnetic field detector,each network participant can observe all transmitted signalssimultaneously and discriminate the signal sources. This isa great advantage compared to MI communication wherea small bandwidth has to be shared when several partic-ipants are active. As in the first scenario, the AUVs areequipped with a 3D magnetic field detector, which enablesa continuous downlink. A small 3D transmitter coil with ashort communication range is used to ensure a range-limitedtemporarily uplink connection.

In summary, the advantages of using a wideband magneticfield sensor are:• Each node can use the full transmitter bandwidth since

the transmission frequencies are divided into the band-width from 100 Hz to 1 MHz, for example.

• Each node can operate as a repeater.• No identification code is needed and localization is

simplified, due to unique transmit frequencies.• Only small coils have to be integrated into mobile nodes

(AUVs or ROVs) if an uplink is needed.

V. CONCLUSIONIn this work, we introduce a new approach for underwatercommunication, where the receiver coil of a conventional MIcommunication system is replaced by a high-sensitive low-noise wideband magnetic field sensor. A detailed theoreticalanalysis of the magnetic underwater communication channelis made for homogeneous water columns, extended by anFEM analysis for communication near water surfaces. In thiscontext, the influence of the water depth on the magneticloss for magnetic communication is examined. Besides, theadjustable parameters of both the transmitting and the receiv-ing side is elaborated and the influence of these parametersare graphically presented. The results provide a guide for thedesign of a magnetic communication system. A small-scaleprototype system for magnetic communication is presented,which justifies the theoretical analysis. The results are sup-ported by first measurements that are taken in the Kiel Fjord.

VOLUME 4, 2016 9




It is shown that a further increase of the SNR is possibleby combining the signal of multiple sensors, which operatesimultaneously. In addition, two underwater scenarios suit-able for magnetic communication are pointed out. In thesescenarios, the use of magnetic field detectors offers some ad-vantages with respect to communication. Thereby, the largebandwidth of the detector enables simultaneous detectionof signals from multiple transmitters without interference.Furthermore, the same sensor can be utilized for positioningand localization purposes. In summary, the findings showthat magnetic field communication from a transmitter coilto a magnetic field sensor like an AMR is a promisingcandidate for low- to mid-range communication purposes inmobile underwater applications, where a small receiver sizeis required.

ACKNOWLEDGMENTThis EU.SH project is funded by the European Union –European Regional Development Fund (ERDF), the FederalGovernment and Land Schleswig-Holstein, Germany.

REFERENCES[1] Zhi Sun and Ian F Akyildiz. Magnetic induction communications for

wireless underground sensor networks. IEEE Transactions on Antennasand Propagation, 58(7):2426–2435, 2010.

[2] Johnson I Agbinya. Principles of Inductive Near Field Communicationsfor Internet of Things. River Publishers, 2011.

[3] Mehrnoush Masihpour. Cooperative Communication in Near Field Mag-netic Induction Communication Systems. PhD thesis, University ofTechnology, Sydney, 2012.

[4] Mari Carmen Domingo. Magnetic induction for underwater wireless com-munication networks. IEEE Transactions on Antennas and Propagation,60(6):2929–2939, 2012.

[5] Ian F. Akyildiz, Pu Wang, and Zhi Sun. Realizing underwater commu-nication through magnetic induction. IEEE Communications Magazine,53(11):42–48, 2015.

[6] Steven Kisseleff. Advances in Magnetic Induction Based UndergroundCommunication Systems. PhD thesis, Friedrich-Alexander-University,Nuremberg, Germany, 2017.

[7] Steven Kisseleff, Ian F Akyildiz, and Wolfgang H Gerstacker. Surveyon advances in magnetic induction-based wireless underground sensornetworks. IEEE Internet of Things Journal, 5(6):4843–4856, 2018.

[8] Yuzhou Li, Shengnan Wang, Cheng Jin, Yu Zhang, and Tao Jiang. Asurvey of underwater magnetic induction communications: Fundamentalissues, recent advances, and challenges. IEEE Communications Surveys& Tutorials, 21(3):2466–2487, 2019.

[9] Maurice Hott, Peter A. Hoeher, and Sebastian F. Reinecke. Magneticcommunication using high-sensitivity magnetic field detectors. Sensors,19(15):3415, 2019.

[10] Boris A Belyaev, Alexander N Babitskii, Nikita M Boev, Andrey VIzotov, Artem A Sushkov, Eugene V Korolev, and Anton V Burmitskikh.Compact non-linear power amplifier for wideband underwater and under-ground near-field magnetic communication systems. In 2019 InternationalSiberian Conference on Control and Communications (SIBCON), pages1–5. IEEE, 2019.

[11] Chao Hu, Mao Li, Shuang Song, Rui Zhang, Max Q-H Meng, et al. Acubic 3-axis magnetic sensor array for wirelessly tracking magnet positionand orientation. IEEE Sensors Journal, 10(5):903–913, 2010.

[12] Stefano Carrella, I Kuncup, Kai Lutz, and Andreas König. 3d-localizationof low-power wireless sensor nodes based on amr-sensors in industrial andami applications. Proceedings of the Sensoren und Messsysteme, 2010.

[13] Hongzhi Guo. Performance analysis of near-field magnetic inductioncommunication in extreme environments. Progress In ElectromagneticsResearch, 90:77–83, 2020.

[14] Donald G Archer and Peiming Wang. The dielectric constant of waterand debye-hückel limiting law slopes. Journal of Physical and ChemicalReference Data, 19(2):371–411, 1990.

[15] Nir Gavish and Keith Promislow. Dependence of the dielectric constantof electrolyte solutions on ionic concentration: A microfield approach.Physical Review E, 94(1):012611, 2016.

[16] Edward C Jordan and Keith G Balmain. Electromagnetic Waves andRadiating Systems. Prentice-Hall, 1968.

[17] Nicholas Paul Fofonoff and RC Millard Jr. Algorithms for the computationof fundamental properties of seawater. Unesco, 1983.

[18] Steven Kisseleff, Ian F Akyildiz, and Wolfgang H Gerstacker. Digitalsignal transmission in magnetic induction based wireless undergroundsensor networks. IEEE Transactions on Communications, 63(6):2300–2311, 2015.

[19] James R Wait. Mutual coupling of loops lying on the ground. Geophysics,19(2):290–296, 1954.

MAURICE HOTT received his M.Sc. in ElectricalEngineering from the Institute of Electrical Engi-neering and Information Technology, Universityof Kiel, Kiel, Germany, in 2017. In his masterthesis, Maurice Hott developed innovative signalprocessing algorithms for sensor arrays in the fieldof radar technology, which was awarded with the“ARGUS Science Award 2018”. Since 2017 heis working as a research assistant at the Chair ofInformation and Coding Theory at the University

of Kiel towards his Dr.Ing. (Ph.D.) degree in electrical engineering. Hisresearch focuses on digital communication for sensor networks and mobilevehicles exploiting modulated magnetic fields in harsh environments. Withinthe EU.SH project “Mobile Autonomous Underwater Vehicles (MAUS)”started in 2019, Maurice Hott is working on a novel magnetic communi-cation link between autonomous underwater vehicles.

PETER A. HOEHER (F’ 14) received theDipl.Ing. (M.Sc.) degree in electrical engineeringfrom RWTH Aachen University, Aachen, Ger-many, in 1986, and the Dr.Ing. (Ph.D.) degreein electrical engineering from the University ofKaiserslautern, Kaiserslautern, Germany, in 1990.From 1986 to 1998, he was with the GermanAerospace Center (DLR), Oberpfaffenhofen, Ger-many. From 1991 to 1992, he was on leave atAT&T Bell Laboratories, Murray Hill, NJ. In

1998, he joined the University of Kiel, Germany, where he is a FullProfessor of electrical and information engineering. His research interestsare in the general area of communication theory and applied informationtheory with applications in wireless radio communications, optical wirelesscommunications, molecular communications, underwater communications,and simultaneous wireless information and power transfer. Since 2014, hehas been a Fellow of the IEEE for contributions to decoding and detectionthat include reliability information. From 1999 to 2006, he served as anAssociated Editor for the IEEE Transactions on Communications.

10 VOLUME 4, 2016

Documents

Underwater Communication Employing High-Sensitive Magnetic