CHINESE JOURNAL OF GEOPHYSICS Vol.49, No.

6, 2006, pp: 16581664

BOOSTING SIGNAL TO NOISE RATIO OF SEISMIC SIGNALS USING

THE PHASED-ARRAY VIBRATOR SYSTEM
JIANG Tao1,2

LIN Jun1,2

LI Tong-Lin3

CHEN Zu-Bin1,2

ZHANG Lin-Hang1,2

1 College of Instrument Science and Electrical Engineering, Jilin University, Changchun 130026, China
2 Key Laboratory of Geo- exploration Instrumentation of Ministry of Education, Jilin University,
Changchun 130026, China
3 College of GeoExploration Science and Technology, Jilin University, Changchun 130026, China

Abstract When noise in the field is very strong, we can not get data with proper signal to noise ratio (SNR) even
if combination vibroseis is used. To solve this problem we proposed to apply a phased-array vibrator, which can
form a directional seismic wave. We have made numerical simulation of multi-track seismograms and computed
reflected waves SNR for combination and phased-array sources on a given horizontal stratified model. The result
shows that the average SNR of reflected waves increases by 1.549.90dB with respect to the combination source
when 4,8,15 and 20-unit phased-array vibrator are used. We conclude that phased-array source is competent to
improve seismic signal SNR. Moreover, the phased-array source can boost SNR obviously when more vibrators
are applied but there is no similar result for the combination source.
Key words

Vibroseis, Combination, Phased-array, Signal to Noise Ratio (SNR), Numerical simulation.

1 INTRODUCTION
Since it is a nondestructive source, electrodynamic vibroseis[13] is widely used in reflection seismic
prospecting. Because of its low power, it is sometimes difficult to get high SNR (signal to noise ratio) seismic data, especially when noise in the fields is strong. Some techniques to improve SNR of data have been
studied[47] . Here, we propose a new PAVS (phased-array vibrator system) technique for acquisition of high
SNR seismic signals.
The combination vibroseis[8] , is a special vibrator array that vibrates simultaneously. It is often used to
get high SNR seismic signals. Generally, the number of vibrators used in the combination source is less than 3
or 4. It cannot reach obvious improvement of SNR of seismic signals, even though more vibrators are used. The
reason is that combination sources direct the seismic wave to a fixed direction (generally the vertical direction),
whereas the reflected seismic waves received from seismometer groups are not close to that direction. Therefore,
we have developed a PAVS, which can form different directional seismic waves and improve SNR of seismic
signals.
In PAVS, more concerns are focused on control parameters of PAVS, layout of the observation system and
complex underground structure. Here, based on basic phased-array seismic theory and the principle of seismic
wave propagation, we synthesize seismic records on a horizontal layered strata model by PAVS and combination
source, and calculate seismic signal SNR. According to the results, we conclude that PAVS is feasible and
superior for seismic prospecting.
2 THEORY
The phased-array technique originated from phased-array radar[911] , which can make electromagnetic
waves concentrate in any direction. Now, the idea of beam forming is widely used in other wave areas, such
as microwave[12] , ultrasonic[13] and light-wave[14] . Vibrator array[4] with solo frequency signals was used to
testify the existence of directional seismic wave. Based on above research, we propose a method to make sweep
frequency seismic waves concentrate in certain direction.
E-mail: jiangtao13@126.com

Jiang T et al.: Boosting Signal to noise Ratio of Seismic Signals Using the Phased-array Vibrator System

1659

PAVS[15] is composed of multi-vibrators and a phased-array control system. All units of the source and
geophones are kept in the line. Vibrators are labeled in ascending order. Distances between adjacent units
are same. Because all units have different locations, reflected seismic waves from different units arrive at each
geophone asynchronously. In order to eliminate wave path difference, we let vibrators vibrate one by one with
the regular interval . Therefore, reflected seismic signals at geophones are stacked in phase, and SNR is
improved.
In order to describe direction of seismic wave, the phased-array direction coefficient Fa () is defined as:
Fa () =

sin n[(kd cos + )/2]

, [0 , 180 ],
n sin[(kd cos + )/2]

(1)

2
where is horizontal angel of any underground observation spot, n is number of vibrators, k =
, d is distance

between vibrators, and = 2f is phase-difference due to the delay time . Extreme angle = M is called
seismic wave direction, in which the strongest seismic wave is. According to the theory of phased array[9] , we
have
v 
.
(2)
M = arccos
d
According to the direction characteristic of seismic wave, the amplitude of seismic wave of PAVS is Ap =
n|F ()| times that of the single source.
3 METHOD
It is necessary to determine proper control signals[1,16] of PAVS, noise level and underground structure in
the field before seismic synthetic records are made.
First, let d=2m, =0.89ms, and PAVS have 4,8,15 and 20 units respectively in the experiment.
Second, let sweep control signal s(t) in the first unit be


f2 f1
s(t) = A cos 2 f1 +
t t[1] ,
(3)
2T
where A is signal amplitude, starting frequency f1 =50Hz, ending frequency f2 =200Hz, T =4s, and t is a time
variate.
Because driving signals between adjacent vibrators are delayed successively , the driving signal of each
unit can be expressed as
si (t) = s[t (i 1) ], (i = 1, 2, , n).
(4)
We assume that noise in the field is random, and variance of noise is 2 =1. SNR can be adjusted by
changing amplitude A of driving signals when noise is constant. Here, we take A=0.003, which means SNR at
the source position is 28.23dB in a single source case.
Finally, we design a multi-layer model. The depth is 100m and velocity is 1000ms1 in the first layer, the
depth is 250m and velocity is 1700ms1 in the second layer, the depth is 180m and velocity is 1900ms1 in
the third layer, and velocity is 2200ms1 in the fourth layer.
Let seismic synthetic records for single source be rsi (t), (i = 1, 2, , number of track), where i is trace
number. When noise is neglected, rsi (t) is composed of a direct wave ri,0 (t) and three reflected waves ri,1 (t),
ri,2 (t), and rsi,3 (t), that means
3
X
rsi (t) =
ri,j (t).
j=0

Thus, seismic waves in the track i can be written as

rpi (t) =

3
X
j=0

n|Fa (i,j )|ri,j (t),

(5)

1660

Chinese J. Geophys.

Vol.49, No.6

where i,0 is the direction of direct wave, i,1 , i,2 , and i,3 are directions of reflected waves, and Fa (i,j ) is
function of frequency f which can be seen from Eq.(1). Because f is linearly related to t for chirp driving
signals, naturally, Fa (i,j ) can be written as Fa (i,j , t) and Eq.(5) can be expressed as:
rpi (t) =

3
X

n|Fa (i,j , t)|ri,j (t).

(6)

j=0

Considering noise, real seismic signals for PAVS are

rmixpi (t) = rpi (t) + rN (t).

(7)

After cross-correlation processing[17] , seismic waves in pulse form are made. To PAVS, reference signal


n1

ref(t) = s t
2
is chosen[1517] . Seismogram of track i is cross-correlation between ref(t) and rmixpi (t)
ri (t) = ref(t) rmixpi (t).
So the seismic synthetic records for PAVS can be expressed as
[r1 (t), r2 (t), , ri (t), , rNUM (t)],
where NUM is the number of tracks.

Fig. 1 Direction graphs for combination and phased-array vibrator system

(a) Direction graph for combination source at f =125Hz; (b) Direction graph for combination source at f =200Hz;
(c) Direction graph for phased-array source at f =125Hz; (d) Direction graph for phased-array source at f =200Hz.

(8)

Jiang T et al.: Boosting Signal to noise Ratio of Seismic Signals Using the Phased-array Vibrator System

1661

Combination synthetic records can be got similarly to PAVS records. In fact, the combination source is
a special case of PAVS, when =0. Direction graphs of the combination source and PAVS are shown in Fig. 1,
where n=8, d=2m. From Fig. 1, we can see that the directions of seismic waves become much narrower when
the frequency of driving signals increases.
4 RESULTS
4.1 Synthetic Records for Combination and PAVS Sources
For given experiment conditions and the numerical simulation method above, we synthesized seismograms
for combination and phased-array sources as shown in Fig. 2. The corresponding seismic wave direction graphs
are shown in Fig. 1a and 1b with n=8.
Figure 3 shows a synthetic seismogram for PAVS, where the maximum radiation direction of seismic wave
M = 54.53 and direction graphs are shown in Fig. 1a and 1b with n=8.

Fig. 2 Synthetic seismograms for combination vibroseis system

(a) n=4; (b) n=8; (c) n=15; (d) n=20.

1662

Chinese J. Geophys.

Vol.49, No.6

Fig. 3 Synthetic seismograms for phased-array vibrator system

(a) n=4; (b) n=8; (c) n=15; (d) n=20.

4.2 Comparison of Synthetic Record SNR Between Combination and PAVS Sources
Using known signal and noise, we calculate SNR of seismic signals from different tracks and reflectors in
Figs. 1 and 2. Results are shown in Tables 1 and 2. Furthermore, we can work out how many PAVS can improve
SNR with respect to the combination case, which are shown in Table 3. As SNR values in different tracks are
not same, we use average SNR, which means arithmetic mean of SNR in all tracks.

Jiang T et al.: Boosting Signal to noise Ratio of Seismic Signals Using the Phased-array Vibrator System

1663

Table 1 Average SNR in synthetic seismograms for combination vibroseis system (dB)
Source

Direct

Reflected

Reflected

Reflected

Mean of three

Mean of direct wave and

position

wave

wave I

wave II

wave III

reflected waves

three reflected waves

n=4

16.19

35.27

34.94

32.73

34.75

33.81

34.10

n=8

10.17

40.33

39.28

29.52

30.10

31.14

32.20

n = 15

4.71

37.12

36.52

31.88

29.06

31.35

32.21

n = 20

2.21

37.85

38.02

31.27

30.59

32.00

32.86

Table 2 Average SNR in synthetic seismograms for phasedarray vibrator system (dB)
Source

Direct

Reflected

Reflected

Reflected

Mean of three

Mean of direct wave and

position

wave

wave I

wave II

wave III

reflected waves

three reflected waves

n=4

16.19

31.71

31.86

31.92

34.51

32.27

32.05

n=8

10.17

30.52

29.09

26.29

29.08

27.36

27.93

n = 15

4.71

35.57

32.76

22.00

25.49

23.62

24.74

n = 20

2.21

32.31

31.63

20.64

24.86

22.10

23.15

Table 3 Average SNR improvement in synthetic seismograms using phased-array vibrator system
(Comparing with the situation of combination source) (dB)
Direct

Reflected

Reflected

Reflected

Mean of three

Mean of direct wave and

wave

wave I

wave II

wave III

reflected waves

three reflected waves

n=4

3.56

3.08

0.81

0.24

1.54

2.05

n=8

9.81

10.19

3.23

1.02

3.78

4.27

n = 15

1.55

3.76

9.88

3.57

7.73

7.47

n = 20

5.54

6.39

10.63

5.73

9.90

9.71

5 CONCLUSIONS
From the above seismic synthetic records, we can get following conclusions:
Firstly, SNR of combination source is lower than that of PAVS.
Secondly, the SNR of reflected waves from near-source-geophones increases and that from far-sourcegeophones decreases with the increasing number of vibrators, which is accordant with the previous conclusion
that the number of vibrators in the combination source should be limited. In PAVS, the SNR values of reflected
waves in all geophones are increased. Furthermore, more vibrators in PAVS can bring higher SNR, PAVS is
helpful for higher SNR and there is no limitation in unit number for PAVS.
Based on the quantitative results of SNR improvements in Tables 1,2, and 3, SNR values of all reflected
wave signals for PAVS are higher than those for the combination source case when the same number of vibrators
is used. For the combination system, there is little influence on SNR when vibrators are more than 4, and it
goes worse when 20 elements are used. For PAVS, there is great improvement to SNR when more vibrators are
used. Moreover, there is a significant SNR improvement for the reflected wave from the second reflector. The
reason is that we controll the direction of the main beam to make reflected wave II strongest.
In general, PAVS is feasible to seismic survey; and it is more helpful for high SNR seismic signals than the
combination source. When SNR in the field is low, PAVS is a good choice.
ACKNOWLEDGEMENTS
This work was supported by the National Natural Science Foundation of China (40674096) and the Ministry
of Science and Technology of the Peoples Republic of China (A project of Hi-tech Research and Development
Program of China (2002AA135240).

1664

Chinese J. Geophys.

Vol.49, No.6

REFERENCES
[1] Nigel A. Vibroseis. New Jersey: Prentice Hall PTR, 1991
[2] Lin J, Chen Z B. Study on modal on coupling for vibroseis with ground. Academic Periodical Abstracts of China.
(in Chinese), 1999, 5(6): 811812
[3] Chen Z B, Lin J. Design for vibrator field experiment based on vibrator-earth system. Journal of Geoscientific
Research in Northeast Asia, 2000, 3(1): 107113
[4] Hong F, Hu T Y, Zhang W P, et al. Attenuating multiples for low signal-to-noise ratio seismic data using optimal
beamforming. Chinese J. Geophys. (in Chinese), 2004, 47(6): 11061110
[5] Hu T Y, Wang R Q, White R E. Beamforming in seismic data processing. Chinese J. Geophys. (in Chinese), 2000,
43(1): 105115
[6] Luo M Q, Cao J, Xie X B, et al. Comparison of illumination analyses using one-way and full-way propagators.
Expanded Abstract of 74th Annual International SEG Meeting, 2004. 6770
[7] Xie X B, Jin S W, Wu R S. Wave Equation based illumination analysis. Expanded Abstract of 74th Annual International SEG Meeting, 2004. 933936
[8] Arnold M E. Beam forming with vibrator arrays. Geophysics, 1977, 42(7): 13211338
[9] Mosca S, Bilotti F, Toscano A, et al. A novel design method for Blass matrix beam-forming networks. Antennas and
Propagation, IEEE Transactions on, 2002, 50(2): 225232
[10] Miura R, Tanaka T, Chiba I, et al. Beamforming experiment with a DBF multibeam antenna in a mobile satellite
environment. Antennas and Propagation, IEEE Transactions on, 1997, 45(4): 707714
[11] Kraft U R. Gain and G/T of multielement receive antennas with active beamforming networks. Antennas and
Propagation, IEEE Transactions on, 2000, 48(12): 18181829
[12] Denno S, Ohira T. Modified constant modulus algorithm for digital signal processing adaptive antennas with microwave analog beamforming. Antennas and Propagation, IEEE Transactions on, 2002, 50(6): 850857
[13] Hatfield J V, Chai K S. A beam-forming transmit ASIC for driving ultrasonic arrays. Sensors and Actuators, A:
Physical, 2001, 92(1-3): 273279
[14] Morgan S P. Detection performance of a diffusive wave phased array. Applied Optics, 2004, 43(10): 20712078
[15] Jiang T, Lin J, Chen Z B, et al. Seismic beam-forming in phased array of vibroseis. Journal of Jilin University
(Information Science Edition) (in Chinese), 2004, 22(1): 181184
[16] Lin J, Chen P C, Jiang T, et al. The signal design of electrodynamic vibroseis for surface imaging. Progress in
Geophysics (in Chinese), 2004, 19(4): 807811
[17] Jiang T, Lin J, Chen Z B, et al. The correlation detect ion method of signals in phased array of vibroseis system.
Chinese Journal of Scientific Instrument (in Chinese), 2005, 26(4): 336339