29-06-2012, 02:26 PM
Improved Decoding Methods of Visible Light Communication System for ITS using LED Array and High-Speed Camera
[attachment=26123]
INTRODUCTION
Light emitting diodes (LEDs) are expected as lighting
sources for next generation. LEDs are superior to conventional
incandescent lights due to their low power consumption, strong
directivity, good visibility, long life, and low heat generation.
This is why, in Japan, many incandescent traffic lights is on
the road to replacing with LED traffic lights. Also, LEDs
are able to control their intensity electrically at fast rate,
since they are semiconductor devices. That is, LEDs can be
used not only illuminating device but also communication
device. For examples, LED traffic lights broadcast driving
assistance data to vehicles: road-to-vehicle communications,
LED vehicle brake lights transmit warning data to vehicles
behind: vehicle-to-vehicle communications. Because of that,
visible light communication systems using LED are well
studied [1]–[4].
SYSTEM OVERVIEW
block diagram of the system model.
A. Transmitter
The transmitter consists of 256 LEDs in the form of a
16 × 16 square matrix, a hierarchical encoder, and an error
correcting encoder. In this paper, we use a turbo code as
the error correcting scheme. The transmitter LEDs generate
a nonnegative pulse with a width of Tb, where Tb is the bit
duration. By changing the width of Tb, LEDs can express the
luminance. Let the data rate be Rb(= 1/Tb), then the bit rate
of the transmitter becomes 256Rb since each LED transmits
a different bit.
HIERARCHICAL CODING
the received image for the
communication distance of 10m and 70m. We can distinguish
each LEDs in Fig. 3(a). But, in Fig. 3(b), we cannot distinguish
them because the LEDs are lumped in with neighboring ones.
In other words, as communication distance is longer, high
spatial frequency components of images degrade severely.
To take advantage of these channel characteristics, we employed
the hierarchical coding scheme using two-dimensional
fast Haar wavelet transform (2D FHWT) in our previous
research [6]. 2D FHWT is used to allocate high-priority data
to low spatial frequency components and low-priority data to
high spatial frequency components. Hence, we can decode
primary data in long distance and also obtain additional data
in short distance.
APPLYING AN ERROR CORRECTING CODE
Hierarchical coding has enabled us to divide transmitted
data into high and low spatial frequency components, and we
confirmed the advantage of the hierarchical coding. But the
bit error performance is not enough to realize error-free. To
improve the performance, we apply turbo coding which is an
error correcting code.
If we apply turbo coding to all data including high and
low spatial frequency components, the degradation of high
spatial frequency components will affect the correcting capacity.
Hence, we apply turbo coding to data divided into each
priority, and improve the bit error performance.
[attachment=26123]
INTRODUCTION
Light emitting diodes (LEDs) are expected as lighting
sources for next generation. LEDs are superior to conventional
incandescent lights due to their low power consumption, strong
directivity, good visibility, long life, and low heat generation.
This is why, in Japan, many incandescent traffic lights is on
the road to replacing with LED traffic lights. Also, LEDs
are able to control their intensity electrically at fast rate,
since they are semiconductor devices. That is, LEDs can be
used not only illuminating device but also communication
device. For examples, LED traffic lights broadcast driving
assistance data to vehicles: road-to-vehicle communications,
LED vehicle brake lights transmit warning data to vehicles
behind: vehicle-to-vehicle communications. Because of that,
visible light communication systems using LED are well
studied [1]–[4].
SYSTEM OVERVIEW
block diagram of the system model.
A. Transmitter
The transmitter consists of 256 LEDs in the form of a
16 × 16 square matrix, a hierarchical encoder, and an error
correcting encoder. In this paper, we use a turbo code as
the error correcting scheme. The transmitter LEDs generate
a nonnegative pulse with a width of Tb, where Tb is the bit
duration. By changing the width of Tb, LEDs can express the
luminance. Let the data rate be Rb(= 1/Tb), then the bit rate
of the transmitter becomes 256Rb since each LED transmits
a different bit.
HIERARCHICAL CODING
the received image for the
communication distance of 10m and 70m. We can distinguish
each LEDs in Fig. 3(a). But, in Fig. 3(b), we cannot distinguish
them because the LEDs are lumped in with neighboring ones.
In other words, as communication distance is longer, high
spatial frequency components of images degrade severely.
To take advantage of these channel characteristics, we employed
the hierarchical coding scheme using two-dimensional
fast Haar wavelet transform (2D FHWT) in our previous
research [6]. 2D FHWT is used to allocate high-priority data
to low spatial frequency components and low-priority data to
high spatial frequency components. Hence, we can decode
primary data in long distance and also obtain additional data
in short distance.
APPLYING AN ERROR CORRECTING CODE
Hierarchical coding has enabled us to divide transmitted
data into high and low spatial frequency components, and we
confirmed the advantage of the hierarchical coding. But the
bit error performance is not enough to realize error-free. To
improve the performance, we apply turbo coding which is an
error correcting code.
If we apply turbo coding to all data including high and
low spatial frequency components, the degradation of high
spatial frequency components will affect the correcting capacity.
Hence, we apply turbo coding to data divided into each
priority, and improve the bit error performance.