Page 154 - Proceedings of the 2017 ITU Kaleidoscope

2017 ITU Kaleidoscope Academic Conference

the certain communication distance, the SISI noise reduces array and the lens’ focal length f. According to the imag-
the BER performance remarkably. In order to define the ing laws, the SISI noise intensity is directly correlated to the
0
optimal detection threshold, the distribution coefficient σ of average expectation value of the spacing distance r , which
ij
the stray light PSF can be estimated by fitting analysis tool is the distance of two adjacent LEDs in the imaging plane.
so that the average SISI noise intensity is able to be obtained Consequently, the SISI noise intensity is able to be estimated
as well. Hence, the optimal detection threshold can be calcu- with the given d, f and the certain communication distance
lated. In Fig.6, the simulation result indicate that compared u. The simulation result and analysis is given.
with the fixed detection threshold condition, the BER curve Furthermore, an adaptive threshold detection method is in-
with optimal detection threshold moves closer to the ideal troduced and the performance is analyzed. The SISI noise
BER performance curve. In other words, by means of esti- impacts the system BER performance in the form of addi-
mating the SISI noise intensity and adaptively adjusting the tive noise. By means of estimating the SISI noise power,
detection threshold, the system BER performance improve the optimal detection threshold can be obtain and the sys-
significantly. tem BER performance enhances significantly. The simula-
tion result and analysis indicates that σ b \σ SISI , the ratio of
the background noise intensity and the SISI noise intensity,
determines the normalized optimal detection threshold value.
In summary, the modeling and analysis of SISI is significant
to improve the MIMO-IS-VLC system performance. In the
future work, the optical power to pixels intensity conversion
will be taken into consideration so that the optimal image-
binary threshold can be applied in the practical application
to improve the system performance.

5. REFERENCES

[1] Parth H Pathak, Xiaotao Feng, Pengfei Hu, and Prasant
Mohapatra, “Visible light communication, networking,
and sensing: A survey, potential and challenges,” IEEE
Fig. 7: The BER curves of different focal length of the lens. Communications Surveys & Tutorials, vol. 17, no. 4,
pp. 2047–2077, 2015.
The Fig.7 shows that there exists a optimal normalized de-
tection threshold that make the system achieve the best BER [2] Nirzhar Saha, Md Shareef Ifthekhar, Nam Tuan Le, and
performance. In the MIMO-IS-VLC system, the background Yeong Min Jang, “Survey on optical camera communi-
noise mainly comes from the ambient light while the SISI cations: challenges and opportunities,” IET Optoelec-
0
noise is related to the received LED light power P LED , the tronics, vol. 9, no. 5, pp. 172–183, 2015.
spacing distance d on the LED array and the lens’ focal
length. As a result, the background noise component and [3] Yuki Goto, Isamu Takai, Takaya Yamazato, Hiraku
the SISI noise component are independent. The simula- Okada, Toshiaki Fujii, Shoji Kawahito, Shintaro Arai,
tion result shows that the σ b \σ SISI influences the detection Tomohiro Yendo, and Koji Kamakura, “A new auto-
threshold value. If the σ b \σ SISI increases from 9.5dB to motive vlc system using optical communication image
12dB, the normalized detection threshold value increases ap- sensor,” IEEE Photonics Journal, vol. 8, no. 3, pp. 1–
proximately 0.05 when the system SNR is 20dB. In addition, 17, 2016.
it is indicated that along with the decreasing of the system
[4] Byung Wook Kim and Sung-Yoon Jung, “Vehicle posi-
SNR, the optimal normalized detection threshold increases.
tioning scheme using v2v and v2i visible light commu-
nications,” in Vehicular Technology Conference (VTC
4. CONCLUSIONS Spring), 2016 IEEE 83rd. IEEE, 2016, pp. 1–5.

In this paper, a MIMO image sensors based visible light com- [5] Katsunori Ebihara, Koji Kamakura, and Takaya Ya-
munication is designed and researched from the perspective mazato, “Layered space-time coding using led array
of optical communication system. The formation mecha- for image-sensor-based visible light communications,”
nism of the stray light in the imaging system and the influ- in 2015 IEEE International Conference on Communi-
ence of the spatial inter-symbol interference is analyzed. The cations (ICC). IEEE, 2015, pp. 5048–5053.
stray light, caused by the incident light’s diffraction, refrac-
tion and reflection, produces the spatial inter-symbol inter- [6] Yuki Ohira, Shintaro Arai, Tomohiro Yendo, Takaya
ference. The SISI noise intensity is related to the received Yamazato, Hiraku Okada, Toshiaki Fujii, and Koji Ka-
0
LED light power P LED , the spacing distance d on the LED makura, “Novel demodulation scheme based on blurred

– 138 –