ULTRASONIC OFDM PULSE FOR BEACON IDENTIFICATION

AND DISTANCE MEASUREMENT IN REVERBERANT

ENVIRONMENTS

Daniel F. Albuquerque, Jos

e M. N. Vieira, Carlos A. C. Bastos and Paulo J. S. G. Ferreira

Signal Processing Lab – IEETA/DETI, University of Aveiro, 3810-193 Aveiro, Portugal

Keywords:

Indoor location system, Asynchronous communication, OFDM, Ultrasound, Time-of-ﬂight.

Abstract:

In this work we propose a frame architecture for asynchronous data transmission using ultrasonic OFDM

pulses in reverberant environments. The frame has two different OFDM pulses modulated with BPSK. The

ﬁrst pulse plays an important role, it is used for time synchronization and to demodulated the unknown data in

the second pulse by a differential demodulation scheme. The proposed frame architecture proved to be robust

to the multipath in different scenarios. Results have demonstrated that it is possible to keep the bit error rate

low in the presence of strong signal echos where other techniques fails, moreover, the simulations show that it

would increase the reliability of ultrasonic indoor location systems.

1 INTRODUCTION

Location is an active area of research in the signal

processing community with a large potential from the

point of view of applications (Sayed et al., 2005; Liu

et al., 2007). The GPS is the most popular system for

outdoor location achieving an accuracy between 20 to

30 m (Sayed et al., 2005).

Recently, there has also been a great interest in us-

ing the existing mobile phone antenna infrastructure

to perform outdoor location without the need of any

additional hardware besides the mobile phones. How-

ever, such systems present in urban areas an accuracy

about 100 m (Lakmali and Dias, 2008; Gustafsson

and Gunnarsson, 2005). This accuracy is not enough

for indoor applications, where the system must pro-

vide the exact position of the object. To perform in-

door location, there are 2 main types of solutions: Ul-

trasonic (US) and Radio Frequency (RF) based sys-

tems. RF based systems are extremely inexpensive

but require the proﬁling of the entire location scenario

to get a RF ﬁngerprint resulting in an accuracy from

1 to 5 meters approximately (Stuntebeck et al., 2008;

Bahl and Padmanabhan, 2000). On the other hand,

the ultrasound technology is the best suited to achieve

the necessary accuracy level in three dimensions, that

can be less than 1 cm in some cases (Gonzalez and

Bleakley, 2009; Prieto et al., 2007).

1.1 LocUS Location System

LocUS is an ultrasonic based location system in de-

velopment, with the main goal of perform indoor

location using only ultrasonic signals. These ultra-

sonic signals will be used to get distance information,

from time-of-ﬂight (TOF) measurements, and also to

implement data communication. Unfortunately, al-

most all of the known ultrasonic location systems use

an auxiliary RF channel for measuring the propaga-

tion delay from the source to receiver (except the M.

Hazas and A. Hopper’s system (Hazas and Hopper,

2006) that presents an accuracy less than 25 cm in

95% of the cases). Although this auxiliary RF channel

allows very simple clock synchronization and delay

measurement solutions, it also gives away two impor-

tant advantages that US-based systems bear in refer-

ence to RF-based ones: the immunity to RF interfer-

ence, and the ability to safely operate in the presence

of critical electronic instrumentation such as medi-

cal or life-support systems. Therefore, one way to

avoid the use of an auxiliary RF signal to measure the

TOF is to synchronize the clocks of the nodes (Skeie

et al., 2001). To achieve this, the nodes should be

able to send to each other the clock information us-

ing the ultrasonic channel. Due to the reﬂection of

the ultrasonic signals on the walls the acoustic com-

munication channel presents a strong multipath effect

causing inter-symbolic interference, for that reason,

OFDM (Orthogonal Frequency Division Multiplex-

124

F. Albuquerque D., M. N. Vieira J., A. C. Bastos C. and J. S. G. Ferreira P..

ULTRASONIC OFDM PULSE FOR BEACON IDENTIFICATION AND DISTANCE MEASUREMENT IN REVERBERANT ENVIRONMENTS.

DOI: 10.5220/0003375601240132

In Proceedings of the 1st International Conference on Pervasive and Embedded Computing and Communication Systems (PECCS-2011), pages

124-132

ISBN: 978-989-8425-48-5

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

ing) could be a viable solution for location and com-

munication. This technique has already been used in

ultrasonic underwater communications (Mason et al.,

2007; Nakashima et al., 2006). OFDM is a very ﬂexi-

ble modulation technique, robust to multipath and that

simpliﬁes the channel equalization. It is also very sen-

sitive to synchronization, which may be an advantage

when the application requires the measurement of the

TOF (Levanon and Mozeson, 2004).

In this paper it will be presented an asynchronous

data transmission with OFDM pulses (section 2). In

section 3 is presented some concerns in the pulse

choice like the shape, size, etc. Some simulated re-

sults and a comparison with other techniques will be

presented in section 4. At the end it will be presented

brief conclusion.

2 ASYNCHRONOUS DATA

TRANSMISSION WITH OFDM

An architecture for asynchronous data transmission

using OFDM pulses is proposed. There will be used

two pulses, one for time synchronization (e.g. time-

of-ﬂight measurement) and another for some data in-

formation transmission (e.g. source identiﬁcation).

The group of the synchronization pulse and the data

information pulse is called frame.

2.1 Frame prototype

Figure 1 presents the proposed prototype for the

frame, as mention before, there are two different main

pulses in the frame: The OFDM Sync. and the OFDM

Data. The ﬁrst pulse is known by the receiver besides

to allow synchronization it will be used to demodulate

the second pulse, the OFDM Data, by a differential

demodulation scheme (Haykin, 2001). This method

was chosen manly because the OFDM pulse is robust

to environments with multipath however it does not

produce a very high resolution in time synchroniza-

tion, how will be seen further. Therefore, the differ-

ential demodulation needs to be robust to this small

jitter.

Guard

FFT

OFDM Sync.

Guard

Time

OFDM Data

Figure 1: Asynchronous data transmission with OFDM,

frame prototype.

The Guard FFT, presented in Figure 1 is a protection

in the demodulation process due to the time synchro-

nization jitter. The Guard Time has the same function

of Guard FFT plus the reduction of the inter-symbolic

interference (Schulze and Luders, 2005).

The frame building for the asynchronous data

transmission with OFDM pulses will be presented us-

ing an example with 17 bit modulated BPSK (Binary

Phase Shift Keying) per pulse. The BPSK is a good

modulation for differential demodulation due to the

phase difference (Haykin, 2001), 17 bits were chosen

because this data may be enough to identiﬁcation and

some data communication. But the extrapolation for

other data sizes and modulations would not be a prob-

lem.

2.2 OFDM Pulse

In the OFDM the information is divided in N blocks

and each value of this block is sent using a differ-

ent carrier, in order to avoid the ISI (inter-symbolic-

interference) the OFDM uses carriers that are orthog-

onal to each other, reducing this way the total signal

bandwidth (Schulze and Luders, 2005).

Moreover, the FFT (Fast Fourier Transform) and

IFFT (Inverse Fast Fourier Transform) can be used

to improve the efﬁciency of the system and the im-

plementation simplicity. Figure 2 presents the block

diagram of a possible OFDM communication system

using the FFT.

S(k)

S/P

... ...

...

P/S

...

Re() DAC

Channel

ADCS/P

...

...... ...

P/S

X(k)

s(n)

x(n)

Figure 2: Block diagram of a OFDM communication sys-

tem.

The system input, S(k), can be symbols from some

classic modulation like PSK (Phase Shift Keying) or

QAM (Quadrature amplitude modulation) (Schulze

and Luders, 2005). Moreover, the system only mod-

ulates the carriers of the pass-band transmission sys-

tem, as shown in Figure 2.

In order to better understand the OFDM pulse a

simple example was chosen, an OFDM pulse with

only three bits, 0 1 0, of information that can be mod-

ulated with BPSK resulting in the symbols 1 -1 1.

Therefore, for this three symbols, carriers 5, 6 and

7 of an IFFT with size 1000 were chosen. Figure 3

ULTRASONIC OFDM PULSE FOR BEACON IDENTIFICATION AND DISTANCE MEASUREMENT IN

REVERBERANT ENVIRONMENTS

125

presents the three individual chosen carriers modu-

lated BPSK. Resulting in the OFDM pulse of Fig-

ure 4.

0 20 40 60 80 100

−0.001

0.001

Sample

Amplitu d e

k = 5

k = 6

k = 7

Figure 3: OFDM example:3 carriers coded BPSK.

0 20 40 60 80 100

−0.003

0.003

Sample

Amplitu d e

Figure 4: OFDM example: The OFDM pulse.

The different combination of carrier codes cre-

ates different pulse shapes and pulse amplitudes. The

systems are normally limited in amplitude and not

in energy due to the DAC(Digital-to-Analog Con-

verter) and ampliﬁers limitations. Therefore, it will

be important to maximize the pulse energy for the

same amplitude or, in other words, we need to min-

imize PMEPR (Peak-to-Mean Envelope Power Ra-

tio) (Schulze and Luders, 2005). In Figure 5 we com-

pare two different OFDM pulse envelopes where all

the 17 carriers where BPSK modulated. The ﬁrst

case (Figure 5(a)) shows an OFDM pulse envelope

with the lowest PMEPR and the second case (Fig-

ure 5(b)) shows an OFDM pulse envelope with the

greatest PMEPR. As can be seen the two examples

are very different and give an idea of how different

the pulses envelope can be.

0 N−1

RMS

(a) Best PMEPR value.

0 N−1

RMS

(b) Worst PMEPR value.

Figure 5: Greatest and lowest PMEPR for an OFDM pulse

with 17 carriers (RMS - root mean square).

3 OFDM PULSE CHOICE

In order to maximize the OFDM pulse detection the

system must use a matched ﬁlter (Levanon and Moze-

son, 2004). Moreover, it is possible to demonstrate

that the probability of pulse detection is proportional

to the pulse energy, (each mens that the most impor-

tant thing for pulse detection will be energy of it and

not the envelope of it) (Levanon and Mozeson, 2004).

Due to the maximum signal amplitude limitation

impose by the hardware (DACs, ADCs, ampliﬁers,

etc.) the OFDM pulse must have the PMEPR as low

as possible in order to maximize the probability of de-

tection.

However, to have good TOF measurement with

less error as possible the OFDM pulse must have an

autocorrelation function similar to an impulse (Lev-

anon and Mozeson, 2004). Therefore, the mainlobe

heigh (at the origin) and the relationship between the

this lobe and the sidelobes must be as big as possible.

However, the mainlobe heigh have a direct relation

with the energy of the OFDM pulse, but the peak-to-

peak amplitude are limited by the source.

Due to the receiver and/or source movement

Doppler effect will occur (Haykin, 2001), therefore,

the correlation between the received pulse and the

matched ﬁlter’s pulse will change with the doppler

shift. As a result of this we must look not only to

the autocorrelation function but also to the output of

the match ﬁlter for different frequencies shifts. To the

output of match ﬁlter in function of time delay and

doppler shift will be called ambiguity function (Lev-

anon and Mozeson, 2004). Consequently, the ideal

ambiguity function is a function that has a single inﬁ-

nite spike at the origin and is zero elsewhere (Levanon

and Mozeson, 2004).

To explain how to chose an OFDM pulse, a pulse

with 17 carriers modulated BPSK was chosen. Al-

though, the extrapolation for other pulse sizes would

not be a problem, it only take more time to ﬁnd the

best pulse. These OFDM pulses must have an ambi-

guity function similar to the ideal ambiguity function.

Therefore, the mainlobe heigh (at the origin) and the

relationship between the this lobe and the sidelobes

must be as big as possible (similar to the autocorre-

lation function). However, the mainlobe heigh have a

direct relation with the energy of the OFDM pulse, but

the peak-to-peak amplitude are limited by the source.

Therefore, we must minimize the PMEPR and

maximize the MSR (Mainlobe-to-Sidelobe Ratio).

With these 17 carriers modulated BPSK we only have

131072 possibilities, to test all off them will not be a

big problem. Moreover, we only need to test half of

them because one pulse and its complement will

produce the same ambiguity function. And we do not

PECCS 2011 - International Conference on Pervasive and Embedded Computing and Communication Systems

126

need to test all of the half set because a pulse and it

carriers inverse produce the same pulse envelope and

ambiguity function (i.e. considering one pulse with 3

carriers coded BPSK the pulse [1 1 -1] has the ambi-

guity function of pulse [-1 1 1]).

The values for MSR and PMEPR are shown in

Figure 6 and 7 respectively. Note that, the MSR tests

were only performed for zero Doppler shift.

Figure 6: Mainlobe-to-Sidelobe Ratio for OFDM with 17

carriers coded BPSK.

Figure 7: Peak-to-Mean Envelope Power Ratio for OFDM

with 17 carriers coded BPSK.

The PMEPR values are between 1.73 and 16.82

and the MSR are between 1.97 and 7.08. In order to

chose the best OFDM pulse with a big MSR and a low

PMEPR we sort the MSR and PMEPR values by the

ratio of MSR and PMEPR from the shortest to biggest

(Figure 8) and is represented the last eight values in

Figures 6 and 7 by a grey circle.

0 1 2 3 4 5 6

x 10

0.5

1.5

2.5

Sort by MSR/P ME P R

MS R /PMEPR

Figure 8: Sort of the MSR and PMEPR ratio.

It is logic that the best value is the last one. It has

a bigger MSR (4.95) and a smaller PMEPR (1.91). In

that case, the BPSK codes are [00111100100101010],

[11000011011010101], [10101011011000011] or

[01010100100111100]. How can be seen the codes

are complemented in phase and frequency.

For compare this pulse with others we chose three

more pulses and we normalize all pulses to have the

same maximum amplitude. We chose: the pulse

with the biggest PMEPR [01101010111111100], the

pulse with the biggest MSR [00000101010101010]

and the pulse with the lowest MSR and PMEPR ra-

tio [00000000000000000]. The ambiguity function of

these examples are shown in Figures 9, 10, 11 and 12.

Where N is the size of the pulse and d the distance

between two OFDM adjacent carriers. For a better

comparative we normalize all the ambiguity function

to the maximum value of the ambiguity function with

the biggest PMEPR.

Figure 9: Ambiguity function of the pulse with the lowest

PMEPR.

Figure 10: Ambiguity function of the pulse with the biggest

MSR.

Figure 11: Ambiguity function of the pulse with the biggest

MSR-PMEPR ratio.

3.1 Pulse Size

To choose the best pulse size is necessary to know the

limitations of the system that will be used to transmit

the information, the limitations come mainly from the

channel conditions (air) and the waves propagation

(ultra-sounds).

For detection point of view a pulse must be as big

as possible (for the same amplitude the energy in-

crease with the length of the signal) (Levanon and

ULTRASONIC OFDM PULSE FOR BEACON IDENTIFICATION AND DISTANCE MEASUREMENT IN

REVERBERANT ENVIRONMENTS

127

Figure 12: Ambiguity function of the pulse with the lowest

MSR-PMEPR ratio.

Mozeson, 2004). On the other hand the length of

the signal is inversely proportional to the distance of

two adjacent carriers and consecutively to the Band-

width of the resultant signal. So in the ﬁrst analysis a

huge pulse will be the best, however it will introduce a

problem, the carriers will be very near and a little rel-

ative speed between the source and the receiver pro-

duce a catastrophic change in the carriers. Therefore

an up and a lower bound to the pulse size is presented:

≤ T

Pulse

≤

c + v

2v f

(1)

where T

Pulse

is the time length of the pulse, B is the

maximum bandwidth for resultant pulse, c is the wave

propagation speed, v is the maximum allowed relativ-

ity speed between the source and the receiver and f

is the central frequency of the resultant desired pulse.

Note that, in this equation only a maximum Doppler

shift of a half of the distance between the adjacent car-

riers will be allowed, moreover, the resultant pulse has

a very narrow band so the Doppler is approximately

equal for all the carriers.

The Guard Time must be greater or equal than

Guard FFT and greater than the minimum time to

avoid the inter-symbolic interference. Normally the

minimum time to avoid the inter-symbolic interfer-

ence is bigger than the Guard FFT. Mathematically

the lower bound to Guard Time is:

≥ max

{

FF T

ISI

min

}

(2)

where the minimum to avoid the inter-symbolic inter-

ference can be approximated, for the omni-directional

transducers, by the solution of the equation:

20log(acT

ISI

min

) + αcT

ISI

min

= log(r) + αr (3)

Where a is the maximum ratio between the reﬂection

wave amplitude which does not interfere in the direct

wave and the direct wave amplitude. The coefﬁcient

α is the attenuation of wave due to the absorption of

ultra-sound by the air and it is given in dB/m. Fi-

nally, r is the maximum distance between the source

and the receiver.

3.2 Asynchronous Receiver

The block diagram of a possible asynchronous re-

ceiver is shown in Figure 13. The Synchronization

part consist in a simple algorithm: after the system de-

tect the pulse, it chooses the maximum of the output

of the detector on the next N samples (the length of

the pulse plus the Guard FFT time if it exists); After

that the system demodulates the data using the syn-

chronization pulse by a differential comparison.

> °

j j

y(n)

maximum

counter

0 to N

Buffer

N samples

de-mod

Stop

ind.max

Start

Data

Start

Data demodulator

Synchronization

x(n)

Figure 13: The block diagram of the asynchronous OFDM

receiver.

4 RESULTS

To test the proposed asynchronous data transmission

with OFDM pulses a practical example was chosen,

the source and the receiver are almost motionless,

therefore, a 0.02 m/s of maximum relativity speed

was chosen. The system works at 40 KHz of cen-

tral frequency with 2 KHz of bandwidth and sampling

rate of 200 KHz. Nevertheless the system must have

at least 10 different channels for communication and

17 bits to transmit. So each channel has a maximum

bandwidth of 200 Hz. Therefore, from the equation

1, we can compute the size of the pulses:

85 ms ≤ T

Pulse

≤ 107 ms (4)

A T

Pulse

= 100 ms can be chosen, as a result of this

the distance between two adjacent carriers is 10 Hz

(d = 10 Hz) and each pulse has 10000 samples.

Too choose T

FF T

the pulse ambiguity function

must be used. It can be seen that the main lobe had

about 6 ms of width. Therefore 3 ms is the distance

in time from the maximum of main lobe to the mini-

mum of it. So we use 5 ms for T

FF T

to allow a poor

accuracy in the received instant.

PECCS 2011 - International Conference on Pervasive and Embedded Computing and Communication Systems

128

Finally to compute the Guard Time, the maxi-

mum distance between the source and receiver must

be used. It is 5 m (r = 5), therefore the attenuation of

wave during propagation due to the absorption of ul-

trasounds by the air is 1 dB/m (α = 1) and a reﬂected

wave with amplitude of one ﬁfth (a = 0.2) of the di-

rect wave does not produce a relevant inter-symbolic

interference. With these values the T

ISI

min

is 15 ms so

15 ms can be used for the Guard Time.

One possible frame for the given example is

shown in Figure 14;

0 5 105 120 220

−1

Time (ms)

Figure 14: Asynchronous data transmission with OFDM,

example of a frame.

To validate the proposed frame we will compare

it with other used method. This method uses a very

common pulse for synchronization, a chirp (Levanon

and Mozeson, 2004), and an usually differential mod-

ulation (DBPSK) to data transmission which allows

some synchronization jitter (Haykin, 2001). As a re-

sult of using DBPSK for transmit the same data we

will need one more extra bit. This extra bit is go-

ing to be used as a reference for differential demod-

ulation. The chirp ambiguity function is presented in

Figure 15 and the block diagram of the frame in Fig-

ure 16.

Figure 15: Ambiguity function of a chirp pulse.

Chirp DBPSK Data

Figure 16: Asynchronous data transmission with chirp and

DBPSK, frame prototype.

The ﬁrst primordial characteristic to be set is the

size of the frame, therefore both frames must have the

same length (220 ms). The chirp has almost the dou-

ble of energy compared to the chosen OFDM pulse, so

it will produce the double of main peak in the match

ﬁlter output. Consequently, in this comparative, we

are going to use a chirp with a half size of the OFDM

pulse (50 ms), as a result we maintain similar capa-

bility of detection for both cases. We do this because

we want give more time (170 ms) to the data in order

to reduce the total bandwidth. But even with this data

length, it is impossible to get the same bandwidth that

we can obtain using the OFDM pulses. Moreover, we

want to transmit 18 bits (17 bits plus one bit for dif-

ferential demodulation) in 170 ms so we need about

212 Hz of bandwidth instead of the 170 Hz in the case

of the OFDM pulse. One possible frame of this exam-

ple is shown in Figure 17;

0 50.08 78.4 106.72 135.04 163.36 191.68 220

−1

Time (ms)

Figure 17: Asynchronous data transmission with chirp and

DBPSK, example of a frame.

The main goal of the proposed frame is the robust-

ness to multipath that exist in the rooms when we use

ultrasonic waves to communicate (reverberant envi-

ronment). All tests will be performed to demonstrate

how robust the proposed frame is and how better or

worst it is comparatively to other common technique.

To test our system we use two different ap-

proaches. In the ﬁrst one we use a synthetic impulse

response to simulate the multipath, in the second one

we use an ultrasonic room acoustics simulator.

4.1 Test with a Synthetic Impulse

Response

To simulate the robustness to multipath we implement

a synthetic impulse response. The system begins to

receive the direct pulse at instant 0 and receives three

more attenuated echoes. This impulse response is

shown in Figure 18 where N (µ,σ) represent a Gaus-

sian independent variable with µ mean and σ

vari-

ance.

N(7.5,1)N(10,1) N(15,1)

N(0.01,10

-4

)

N(0.05,10

-4

)

N(0.4,10

-2

)

time (ms)

Amplitude

Figure 18: The synthetic impulse response for test the sys-

tem.

ULTRASONIC OFDM PULSE FOR BEACON IDENTIFICATION AND DISTANCE MEASUREMENT IN

REVERBERANT ENVIRONMENTS

129

The data to send was chosen randomly for OFDM

and Chirp plus DBPSK, in DBPSK the ﬁrst bit was

set to one in all simulations. Moreover, it was added

white noise to resultant signal and the threshold was

set to have a probability of false alarm of the 10

−9

The Bit error rate (BER) was computed for each sig-

nal (OFDM or Chirp plus DBPSK) and 1 million sim-

ulations per signal amplitude to noise standard devia-

tion were performed. The result of this test is shown

in Figure 19. How you can be seen the OFDM had

an excellent behavior comparatively to the Chirp plus

DBPSK, which has a BER ﬂoor greater than 10

−2

−40 −30 −20 −10 0 10 20 30 40 50 60

−6

−4

−2

Signal amplitude to noise standard deviation ratio (dB)

BER

OFDM

Chirp

Figure 19: The Bit error rate in a multipath channel for a

theoretical impulse response.

4.2 Tests in a Room Simulator

To test and compare the two methods we use a simula-

tor implemented by the authors and presented in (Al-

buquerque et al., 2010), it is an acoustic simulator that

assumes specular reﬂections and aimed to simulate

ultrasonic communications. Moreover, it takes in to

account the attenuation due the air propagation losses

(as a function of the signal frequency, room tempera-

ture, air pressure and humidity), wall reﬂections and

source and receiver beams. The proposed simulator

was used to model a real room and it closely matched

the experimental observations.

In this simulator two test are performed in the fol-

low room conditions:

• The room has 5x5x3m of dimension;

• The reﬂection coefﬁcient of the wall is 0.7;

• The reﬂection coefﬁcient of the ﬂoor is 0.5;

• The reﬂection coefﬁcient of the ceil is 0.9;

• No noise was added to the incoming signal;

• For each position were sent 170 bits (10 frames);

• For the source was used an approximated beam

function from the Murata transmitter MA40B4T;

• For the receiver was used an approximated beam

function from the Murata receiver MA40B4R;

• The source and the receiver were placed at 1m

from the ﬂoor;

• The room temperature was set to 22

◦

• The atmospheric pressure inside the room was set

1 atm;

• The relativity humidity inside the room was set to

33%.

The ﬁrst test was conducted in the room of Figure

20, the source was located at 1cm of the “top” wall

and 10cm of the “left” wall. The main propagation

direction of the source was to “down” and parallel

to “left” wall. The receiver started “walk” at 2 m

from the “top” and “left” wall and stopped at about

27.5 cm from the source. Therefore, the simulator get

the probability of bit error with a “step” resolution of

2.75 mm resulting in 900 probabilities (Figure 21).

How can seen the two frames have a good behavior

(the OFDM has a small advantage) in this test.

1 m

0.01 m

from the top wall

and 0.1 m

from the left wall

Figure 20: First simulator test room.

The second test was conducted in the room of Fig-

ure 22, this room is similar to the room of the ﬁrst

test, but this one has a wall with a 1 m of width and

20 cm of deep. The source was placed at the centre of

the “top” and at 1 cm from it. The main propagation

direction of the source was set to “down” and paral-

lel to “left” wall. The receiver started “walk” at 1 m

from the “left” and “bottom” wall and stopped at 1 m

from the “right” and “bottom” walls. The simulator

get the probability of bit error with a “step” resolu-

tion of 3 mm resulting in 1000 probabilities (Figure

23). How can seen there will be in these type of cases

that the proposed prototype will present a better re-

sult in comparative to other usual techniques. With an

appropriated code correction it will be possible to re-

duce the probability of bit error to almost zero where

with other techniques it is impossible.

PECCS 2011 - International Conference on Pervasive and Embedded Computing and Communication Systems

130

100 200 300 400 500 600 700 800 900

0.2

0.4

OFDM

P. error

Position

100 200 300 400 500 600 700 800 900

0.2

0.4

0.6

Chirp+DBPSK

P. error

Position

Figure 21: The probability of error in the ﬁrst simulator test.

1 m

R R RRR

0.01 m

from the wall

Figure 22: Second simulator test room.

5 CONCLUSIONS

In this paper is proposed an asynchronous data trans-

mission with OFDM pulses which is robust to the

multipath effect. To achieve these goals OFDM with

BPSK modulated carriers was used. It was presented

the OFDM pulse restrictions and it was given some

advices of how to build the proposed asynchronous

data transmission and how to choose the size and the

shape of the OFDM pulse. Therefore, some sim-

ulations to compute the bit error rate of the pro-

posed system were performed. The performance of

the proposed technique, in the presence of multipath,

is undoubtedly better than the use of the chirp with

DBPSK. Moreover, some considerations can be made

100 200 300 400 500 600 700 800 900 1000

0.2

0.4

OFDM

P. error

Position

100 200 300 400 500 600 700 800 900 1000

0.2

0.4

0.6

Chirp+DBPSK

P. error

Position

Figure 23: The probability of error in the second simulator

test.

about the Doppler effect. The OFDM pulse may be

robust than the chirp pulse. Because the ambiguity

function is more similar to the ideal ambiguity func-

tion which mays allow to recovery the instant and the

speed of the source with a better precision.

REFERENCES

Albuquerque, D., Vieira, J., and Bastos, C. (2010). Room

Acoustics Simulator for Ultrasonic Robot Location.

Rob

otica, (77):10–14.

Bahl, P. and Padmanabhan, V. N. (2000). RADAR: an in-

building RF-based user location and tracking system.

In Nineteenth Annual Joint Conference of the IEEE

Computer and Communications Societies, volume 2,

pages 775–784 vol.2.

Gonzalez, J. R. and Bleakley, C. J. (2009). High-Precision

Robust Broadband Ultrasonic Location and Orienta-

tion Estimation. IEEE Journal of Selected Topics in

Signal Processing, 3(5):832–844.

Gustafsson, F. and Gunnarsson, F. (2005). Mobile posi-

tioning using wireless networks: possibilities and fun-

damental limitations based on available wireless net-

work measurements. Signal Processing Magazine,

IEEE, 22(4):41–53.

Haykin, S. (2001). Communication Systems. John Wiley &

Sons, 4th edition.

Hazas, M. and Hopper, A. (2006). Broadband ultrasonic lo-

cation systems for improved indoor positioning. IEEE

Transactions on Mobile Computing, 5(5):536–547.

Lakmali, B. D. S. and Dias, D. (2008). Database Correla-

tion for GSM Location in Outdoor & Indoor Environ-

ments. In Information and Automation for Sustain-

ability, 2008. ICIAFS 2008. 4th International Confer-

ence on, pages 42–47.

Levanon, N. and Mozeson, E. (2004). Radar Signals. JOHN

WILEY & SONS.

ULTRASONIC OFDM PULSE FOR BEACON IDENTIFICATION AND DISTANCE MEASUREMENT IN

REVERBERANT ENVIRONMENTS

131

Liu, J., Chu, M., and Reich, J. E. (2007). Multitarget Track-

ing in Distributed Sensor Networks. Signal Process-

ing Magazine, IEEE, 24(3):36–46.

Mason, S., Anstett, R., Anicette, N., and Zhou, S. (2007). A

Broadband Underwater Acoustic Modem Implemen-

tation Using Coherent OFDM. In The National Con-

ference On Undergraduate Research, California.

Nakashima, Y., Matsuoka, H., and Yoshimura, T. (2006).

Evaluation and Demonstration of Acoustic OFDM. In

Fortieth Asilomar Conference on Signals, Systems and

Computers, pages 1747–1751.

Prieto, J. C., Jimenez, A. R., and Guevara, J. I. (2007).

Subcentimeter-accuracy localization through broad-

band acoustic transducers. In IEEE International Sym-

posium on Intelligent Signal Processing, pages 1–6.

Sayed, A. H., Tarighat, A., and Khajehnouri, N. (2005).

Network-based wireless location: challenges faced

in developing techniques for accurate wireless loca-

tion information. Signal Processing Magazine, IEEE,

22(4):24–40.

Schulze, H. and Luders, C. (2005). Theory and Applica-

tions of OFDM and CDMA. John Wiley & Sons, ﬁrst

edition.

Skeie, T., Johannessen, S., and Holmeide, O. (2001).

Highly accurate time synchronization over switched

Ethernet. In 8th IEEE International Conference

on Emerging Technologies and Factory Automation,

pages 195–204 vol.1.

Stuntebeck, E. P., Patel, S. N., Robertson, T., Reynolds,

M. S., and Abowd, G. D. (2008). Wideband powerline

positioning for indoor localization. In Proceedings of

the 10th international conference on Ubiquitous com-

puting, pages 94–103, Seoul, Korea. ACM.

PECCS 2011 - International Conference on Pervasive and Embedded Computing and Communication Systems

132