Research on The Methods of Multipath Separation for
Shallow Water Wideband Signal Based on Warping
Transformation
Chuanxi Xing
School of electrical and information engineering
Yunnan Minzu University
Kunming, China
xingchuanxi@163.com
Lingling Kong
School of electrical and information engineering
Yunnan Minzu University
Kunming, China
kongling0104@163.com
Guofang Dong
School of electrical and information engineering
Yunnan Minzu University
Kunming, China
dongguofang1@163.com
Tianyu Gao
Institute of Harbin Technology
Space Control and Inertial Technology Research Center
Harbin, China
hit_gaotianyu@hotmail.com
Abstract For the research on the multipath characteristics of
shallow water wideband signal, the methods of multi-path
separation were explored. After the summary of all kinds of
normal mode decomposition method and the combination with a
new kind of non-linear signal processing transformation method
Warping transformation, a new method of multi-path
separation for shallow water wideband acoustic signal based on
Warping transformation was proposed. With the transmitted
signals limited on a time window and the received signal of the
processing of pulse compression, the processed received signal had
a processing of warping transform and time-frequency analysis, in
this way, a method which has wide applicability and high precision
on decomposition and information ex-traction of the normal mode
was achieved. Finally, a conclusion was verified through a
simulation, the new method can be used in the normal mode
separation of the chirp transmitted signal with a certain length of
time, and has a greater advantage compared with conventional
Warping transform processing results. Meanwhile, the separation
signals extracted were substantially identical with the original
signal after inverse operation restore, which proved the accuracy
of this method.
Keywords
shallow water wideband signal; normal mode
decomposition; Warping transformation; pulse compression
I. INTRODUCTION
Because of the influence of factors such as nature conditions,
geography conditions and so on, shallow water wideband signal
is very complex.
When sound signal transmits in shallow water, uneven of
medium, inner structure of ocean and other factors cause the
signal to reach received point in different time and to form multi-
path effect. Multi-path effect will make signal distortion, time
domain and frequency domain wideband of signal and channel
are selective fading. Meanwhile, multi-path structure is usually
time-varying and relates to sound source and position of
received point[1]. Multi-path effect is always hot issue in
underwater acoustics research and common multi-path
decomposition technologies mainly included diversity
technology, spread spectrum technology, adaptive equalization
technique and time reversal technology, etc. These methods aim
to remove mutli-path effect. But in the underwater acoustic
propagation theory in order to understand the characteristics of
the channel, it must also analyze multipath delay and amplitude.
Transmitted signal of every path of mutli-path involves
sound field environment, sound source position and other
information so that mutli-path structure can not only describe
basic principle of sound propagation but also be used as array
mutli-path signal reaching time can be used to study location and
inverse sound velocity profile, etc. Therefore, research on multi-
path separation methods are of great significance[2-4].
In recent years, using Warping transformation to analyze
characteristics of normal mode frequency dispersion in
underwater waveguide has become more and more popular.
N.R.Chapman used Warping transformation to research on
inversion of bottom geoacoustic parameters. Haiqiang Niu used
Warping transformation to separate each order normal mode and
took advantage of dispersion characteristics to inversion on
sound velocity profile, and according to BDRM theory, deduced
and improved Pekeris waveguide dispersion formula again, gave
a Warping transformation when thermocline exits in seawater.
Normal mode decomposition based on Warping
transformation is a frequency analysis method based on
frequency analysis in nature which is easy but still has some
disadvantages. For example, Warping transformation only can
be used in wideband pulse signal and some ocean environment
models, there is folding phenomenon among each normal mode.
978-1-5090-0690-8/16/$31.00 © 2016 IEEE