SONAR BUOYS: AN IMPROVED DESIGN APPROACH

K. Balasubramanian

Department of Computer Engineering,Faculty of Architecture and Engineering

European University of Lefke, Turkish Republic of Northern Cyprus

Mersin 10, Turkey

Keywords: Buoy, marine exploration, replica correlation, underwater surveillance, microprocessor control.

Abstract: Design approach for improved system performance of a microprocessor controlled sonar buoy performing

surveillance of underwater objects is proposed. When launched into sea or ocean the microprocessor

controlled buoy sets into action for automatic scanning of the underwater as to extract the object

information and transmit the same by wireless to a remote ground station for further processing and taking

final control action. System design outline for sonar buoy incorporating 11-cell replica correlation resulting

in improved system performance is presented in this paper. Although the complexity of the hardware

replica correlator is minimized using the recent digital delay lines the proposed microprocessor controlled

buoy performs replica correlation through software and extracts object information conceding improved

system performance.

1 INTRODUCTION

In several underwater exploring systems sonor

buoys are kept on the surface of the ocean on

permanent basis as to detect the underwater objects

and present the information continuously for longer

time. Major types of noise affecting the performance

of such sonar systems, under these environment

conditions, are the impulse noise and colored noise

which could affect detection process and create false

alarms. Attempts have been made by the researchers

to get rid of the effect of impulse noises as to reduce

the false alarm probability and to enhance the

probability of detection. A signal processing

technique dominant in this area is replica correlation

(N.Sarkar, 1999, Elements of Digital signal

processing, Khanna Publishers, New Delhi. Taub

and Schilling, 1990, Balasubramanian et

al,1993,1999 and 2000) which would boost the SNR

to a high level for enabling easy detection. Replica

correlation technique involves in using binary phase

shift keyed signal (BPSK) commanded by a packet

of ‘n’ cells incorporated in the duration of the

ongoing pulse. The received signal is appropriately

delayed and summed up in an array of adders with

weighted signs assigned in the reverse pattern as to

achieve signal boosting. In the past, in certain mini

type sonar systems a 7-cell hardware replica

correlator having analog delay lines were employed

to have a reasonable system performance. The use of

acoustic delay lines in any system makes it bulky

and heavy. (M.I.Skolnik,2000). The performance

was enhanced further by using 9-cell corrlators with

the installation of digital delay lines. The recent

digital delay lines are promising to be useful for

such purposes( Balasubramanian et al 1994 and

1999). As the cell word size for replica correlation

increases, it increases the SNR but also increases the

complexity of the delay line structure contributing to

its physical size. Alternatively, the replica

correlation could be performed by software

techniques where the usage of delay lines is avoided.

In such cases, the increase in the word size for cell

pattern calls for the use of high speed processor and

large memory. The portable type of sonar used in

mini-vessels and buoys demand the system hardware

to be minimized as to achieve compactness and low

weight for the unit.

2 SONAR BUOY SYSTEM

OUTLINE

The essential constituents of the proposed sonar

buoy are, 1. Acoustic transmitters and receivers with

piezo electric transducers (PZTs), 2. Stepper motor

and its drive circuit, 3. Continuous Wave (CW) AM

modulator and transmitter and 4. Microprocessor

161

Balasubramanian K. (2006).

SONAR BUOYS: AN IMPROVED DESIGN APPROACH.

In Proceedings of the Third International Conference on Informatics in Control, Automation and Robotics, pages 161-167

DOI: 10.5220/0001201201610167

 SciTePress

based control unit which governs all units of the

buoy. Fig.1 shows the simplified block schematic.

When the buoy is thrown into the sea for marine

exploration certain parts of the buoy such as the

antenna of the mini AM transmitter are staying

above the water surface so as to keep it ready for

transmission. The acoustic PZT crystals with

holders are set inside the water ready for performing

the propagation of the acoustic waves. When

launched into the sea, after the predetermined time

set in the buoy with a monostable multivibrator the

microprocessor sets the stepper to rotate

continuously. At the same time in each step of its

rotation acoustic packet of pulses are transmitted and

reflected echo is received in conventional manner.

As the time elapsed between the transmitted pulse

and the received echo concedes the range

information relating to the current azimuth angle this

is accumulated in memory of the microprocessor

unit. After storage this information is converted into

analog form and driven to the AM transmitter for

broadcasting to the ground station. The remote

ground station at the seashore in turn receives the

information and performs further processing as to

extract useful information needed for subsequent

decision making purposes.

3 REPLICA CORRELATION

DETECTION

Replica correlation technique improves the

probability of detection and the false alarm

probability. As explained before, each acoustic

pulse being transmitted is contributed by a packet of

‘n’ cells. In the proposed project ‘n’ is optimally

chosen as 11. The pattern of cells is evaluated as a

bit stream shown in Fig. 2 where + indicates in-

phase and – indicates the out-phase of the CW sine

wave.

The Continuous Wave signal being transmitted is

binary phase shift keyed according to the cell pattern

and acoustically transmitted in the under water for

regional. The acoustic echo returned from the

objects also posses the same phase reversals and by

replica correlation operation we identify and boost

these echo in presence of noise. Noise being random

in nature it does not undergo these pattern of phase

reversals and does not get boosted up. The nature of

the CW signal which is BPSK keyed and transmitted

according to cell pattern is shown in Fig.3

Replica correlation could be performed by hardware

and also by software. The schematic of the hardware

replica correlator is shown in Fig.4.

Figure 1: Simplified Block Schematic of the Buoy.

Figure 2: 11-Cell pattern.

MPU

CW AM

Modulator

Transmitter

Drive

Acoustic

Transmitter &

Receiver

Stepper

Drive

Motor

PZT

- - + + - + - - + + +

ICINCO 2006 - SIGNAL PROCESSING, SYSTEMS MODELING AND CONTROL

162

0 50 100 150 200 250

-1

-0.8

-0.6

-0.4

-0.2

0.2

0.4

0.6

0.8

Figure 3: CW signal in one packet of acoustic pulse.

Figure 4: Replica Correlator Circuit.

The echo BPSK signal enters into the digital

delay line and propagates up to the end. The recent

digital delay line [8,9] comprises pipe lined 8-bit

shift registers terminated with its front end by an

ADC (Analog to Digital Converter) which converts

the incoming BPSK into a stream of digital bytes.

The ADC used is 8-bit flash ADC constructed using

advanced architecture (Balasubramanian, 2003). The

bytes entering at the front end shift register

progresses to the end with the clock applied to delay

line. In order to get the delayed analog output at

different taps, DACs (Digital to Analog Converters)

are fixed at the Shift registers at regular intervals. At

some taps of the delay line analog inverters are

connected depending upon the inverse sequence of

the cell pattern. The delayed signals after due

inversion are summed up to obtain the replica

correlated output. The received BPSK echo

propagates through the delay line, and when its first

cell reaches the last tap there occurs a significant

boost in the signal level making an amplification to a

factor of 11 corresponding to the cell number.

The signal boosting process is evaluated as

follows. If x

denotes the bi-phase value of the

received signal falling in i

cell and pj

denotes the

polarity of the amplifier attached to the j

tap of the

delay line the output realized at the summing

amplifier at various cells during the propagation of

the signal in the delay line is expressed in (1)

wherein K

denotes the amplitude of the summing

amplifier at the i

cell of the arriving incoming

signal. For the 11-cell replica correlator K

would

have a magnitude 11 and all other K

would be

limited to a maximum of 3.

= Σ p

(i+1-q)

(1)

q=1

When the BPSK has propagated to 11

cell the

signal amplitudes in all cells in the order is given in

(2)

= { -1 -2 -1 +2 +3 0 -3 0 -3 0 +11} (2)

When only noise is entering the delay line there

would be additions and cancellations randomly

causing low output at the summing amplifier. At the

output of the replica correlator there is a threshold

voltage comparator which detects easily the

presence of the echo. Fig.5 shows the replica

correlated output for an echo BPSK signal.

Inverting amplifier

ack of

tapped

digital

delay line

Summing

Amplifier

Resistors - all are of same value

Incoming

BPSK echo

nal

SONAR BUOYS: AN IMPROVED DESIGN APPROACH

163

0 50 100 150 200 250

-15

-10

-5

Figure 5: Replica correlated output.

Table 1: Best designed cell pattern.

No.of

cells

Pattern Unwanted

Max

signal

Cell signal output

7 - + - - + + + 1u -1 0 -1 0 -1 0 +7

9 - + - + - - + + + 2u -1 0 -1 +2 -1 -2 +1 -2 +9

11 - - + + - + - - + + + 3u -1 -2 -1 +2 +3 0 -3 0 -3 0

+11

13 - - - + + + - + - - + + + 4u -1 -2 -3 0 +3 +4 +1 -2 -3 -4 -1 +2

+13

15 - - - + + + + - - + - - + + + 5u -1 -2 -3 0 +3 +4 +3 -1 -1 -4 -5 -2 -3 +4

+15

Figure 6: Schematic of the sonar with microprocessor implementation.

MPU

Keyboard

Dis

RAM

PROM

Interrupt Interface

Oscillator

PISO

Controlled

Inverte

Transmitter

Amp ADC

EOC

IO ports

DAC

AM Modulator &

Transmitter

Stepper Drive

Motor

From

MPU

Output

Port

Timer

ICINCO 2006 - SIGNAL PROCESSING, SYSTEMS MODELING AND CONTROL

164

3.1 Algorithm for Cell Pattern

Design

The SNR materialized in a sonar system improves

with the number of cells used in the packet of bit

stream. Therefore, alternative to 11 cell pattern one

would attempt to include 13 cells or 15 cells as to

improve the SNR. For this consideration optimum

cell pattern has to be chosen for best performance

conceding lowest correlated output during all cell

periods except the last one which gives n-times the

amplitude in voltage level. The optimum cell

patterns evaluated for various cell word lengths are

given in Table 1. The best 7 cell pattern being

conventionally used - + - - + + + would occur as

last 7 cells of any higher order word length. The

algorithm determines the first few ‘m’ cells as

follows. The number ‘m’ is divided by 2 and

actions are taken according to the value. If it is

unity it assigns the first two cells to be – and + such

that the 9-cell word length would be - + - + - - + + +.

It always starts with the ‘ –‘ whatever may be

preceding cells ‘m’. If m/2 is 2, two cells are

given for – and another two for + to have preceding

cell pattern as (--++). This trend continues with

m/2 equal to 3 where first three cells are fixed with –

and the next three with +. When m/2 is more than 3,

we limit the successive identical cells to 3. For

example if m is 8, first three are -, next three are +

and the final two cells would be + -. Therefore, by

applying this program one could find the optimum

cell pattern for any number of cell word length.

4 SONAR BUOY SYSTEM

IMPLEMENTATION

The block schematic of the sonar with

microprocessor implementation is shown in Fig.6.

As outlined in Fig.1 the microprocessor controls all

operations such as controlling the PZTs for rotating,

performing replica correlation by software, detecting

the targets and arranging a coordinate system taken

to memory and driving the AM wireless unit for

radio transmission. A sine wave oscillator is

responsible to produce the acoustic CW wave. This

signal is BPSK keyed by a controlled inverter

(shown in Fig.7)working in accordance with the

polarity of the bit stream stored in PISO (parallel In

Serial Out) register. The sine wave input ‘vi’ is

inverted at the output ‘vo’ depending upon the

polarity of the control bit vc. When vc is ‘high’ the

transistor comes into saturation causing virtual earth

to exist at the non inverting input terminal of the op-

amp resulting in analog inversion. On the other hand

when vc is ‘low’ the transistor would be in cut off

working as a buffer for vi to reach vo.

Figure 7: Controlled inverter circuit.

A Timer is initiated by the microprocessor to

generate periodic input to load the bits stream of cell

pattern to PISO at the rate PRF (Pulse repetition

Frequency) of the acoustic wave. As known, PRT

the Pulse Repetition Time of the periodic pulses

used for the acoustic transmission depends on the

maximum desirable range Rmax set for marine

exploration.

PRT = 1/ PRF = ( 2* Rmax/c ) (3)

where c is the velocity of propagation of the

acoustic waves in underwater.

The timer generates this periodic waveform as a

function of timer-in derived from the source of sine

wave form the oscillator. The received echo signal

from the target is amplified and digitized with 8-bit

ADC and read to the microprocessor. The clock

from the microprocessor is suitably divided in

frequency and given as the timer-input and this also

serves as the Start Conversion (SC) input for the

ADC. The ADC when it completes its conversion

interrupts (Interrupt-1) the microprocessor to read

the sample to it and put it into the RAM memory

organized as FIFO. This FIFO memory substitutes

the use of delay line. At each time of inputting a

byte to the FIFO, it is first updated and then

computation for the replica correlated output is

made. This process takes the data from the definite

memory locations at equal address spacing that

represents the taps in the delay line, assigns polarity

weights and sums them up. The first data is taken at

47K

100

47K

SONAR BUOYS: AN IMPROVED DESIGN APPROACH

165

the location 30 of FIFO and the last tap is at 330.

The replica correlated output is represented by

Eo = ∑ pi *M

30*i

(4)

i= 1

where pi is the polarity assigned to various taps.

Fig.8 shows its simple schematic.

The current data is compared with a threshold to

know whether target is present or not. The timer

which initiates the PISO also interrupts (Interrupt-2)

the microprocessor to initialize a range counter

organized in a register. The range counter is

incremented each time by the software by the

incoming sample with Interrupt-1 and it will

terminate at the instant the replica correlated output

crosses the threshold. The range data is loaded in the

vectored array memory for creating coordinate

display further. At each angular position of the PZT

a vector is initialized to record target in that

direction. While the regional coordinate information

being gathered and dumped in memory it is also sent

to the output port where it is converted into analog

form and driven the AM modulator and transmitter.

Figure 8: Flow of data stream for replica correlation.

As continuous scanning is required for

surveillance, the stepper drive data is derived from a

ring counter activated by the MPU and provided to

the stepper motor for performing the scanning

process without any interruption.

4.1 Simulated Experiment for Noise

Study

The replica correlated output is generated by

simulated 11-cell signal with noises added at

different amplitudes. Fig.9 shows the output with

random noises added in voltage level by 200% to

500%. This shows that even if noise is around 400%

amplitude of the signal the target could be detected

without any difficulty.

0 50 100 150 200 250

-20

-15

-10

-5

(a)

0 50 100 150 200 250

-20

-15

-10

-5

(b)

0 50 100 150 200 250

-25

-20

-15

-10

-5

(c)

0 50 100 150 200 250

-25

-20

-15

-10

-5

(d)

Figure 9: Replica correlated output ; Amplitude vs time

for one packet duration. a. 200% noise b.300% c.400%

d.500%.

FIFO

Mem

M(330)

M(300)

M(270)

M(240)

M(210)

M(180)

M(150)

M(120)

M(90)

M(60)

M(30)

Replica

Correlation

Software

FIFO

Replica correlated

output

Sam

le strea

Time

ICINCO 2006 - SIGNAL PROCESSING, SYSTEMS MODELING AND CONTROL

166

5 CONCLUSION

The proposed project envisages in exploring the

marine region to identify the objects and

communicate to the coastal station. This is normally

meant for the temporal investigation requirements

such as to identify the infiltration of foreign

elements in the regional territory or the bulk

availability of sea animals like sharks and whales.

Based on this information decision could be made

by the ground station to proceed further with the

objects determined. Replica correlation activity is

reinforced with the optimal 11-cell pattern so as to

improve the probability of detection and reduce the

false alarm probability. It is to be noted that when

the number of cells in a packet is increased more

than 11, it improves the SNR and also increases the

complexity. For enhanced number of cells the

proposed algorithm determines the optimal cell

pattern for reaching maximised SNR. One could

effectively use this algorithm for larger number of

cells used in the packet. When replica correlation is

performed by hardware the complexity increase is a

dominant factor and it is minimized by using recent

digital delay lines. It is further minimized in the

proposed sonar buoy where the microprocessor

performs the replica correlation by software.

ACKNOWLEDGEMENTS

The author acknowledges with thanks all necessary

support extended by the Department of Computer

Science and Engineering for carrying out this

project. He also thanks the Faculty of Architecture

and Engineering and the Rector of European

University of Lefke for providing the funding

support for the project without which it would not

have come to its completion.

REFERENCES

N.Sarkar, 1999, Elements of Digital signal processing,

Khanna Publishers, New Delhi.

Taub and Schilling, 1990, Principles of Communication

systems, Tata McGraw Hill, New Delhi.

K.Balasubramanian and S.Arica, 1993, "On the synthesis

of a nine-point replica correlator for a sonar",

Proceedings of the International Symposium on

Signals Circuits and Systems, Iasi, Romania, pp 148-

151.

K.Balasubramanian, H.Camur and Rajaravivarma, 1999,

“Replica Correlation Technique for Improving the

SNR of Digitally Transmitted signals”, Abstracts of

the 1999 IEEE SoutheastCON Conference, Lexinton,

Kentucky, USA.

K.Balasubramanian, 2000, “Replica correlation detection”,

Electronics World, Vol 106, No 1767, March2000,

pp 238-240.

M.I.Skolnik, 2000, Introduction to Radar Systems,

McGraw Hill, New York.

K.Balasubramanian and C.Cihan, 1994, "An intelligent

digital delay line for replica correlation",

Proceedings of the International Conference on

Robotics, Vision and Parallel Processing for Industrial

Automation, IPOH, Malaysia, pp 191-197.

K.Balasubramanian, K.B.Gayathri and H.Camur,1999,

“Voltage Control for a Digital Delay Line”,

Electronics World, Vol 105, No 1762, pp 826.

K.Balasubramanian, K.B.Gayathri and H.Camur ,

1999,“Delay-length-locked loop”, Electronics World,

Vol 105, No 1763, pp 910-911.

K.Balasubramanian, 2003, “On the design and

development of high performance ADCs”, Journal of

AMSE-Modelling Measurement and Control, A-

series, France, Vol 76, No. 3, pp 35-48.

SONAR BUOYS: AN IMPROVED DESIGN APPROACH

167