*4.2. Frame Format*

Figure 9 shows the frame format of the information bits, which consists of four parts: The synchronization header of the frame, the training level, the information bits, and the terminator. The synchronization header of the frame uses a high level of six time slots and a low level of one time slot. To eliminate the effects of ambient light, the training levels are added before the information bits. Finally, the signal 0 × 0D is used to represent the end of a frame.


**Figure 9.** The frame format of the information bits.

The character sent by the serial debugging assistant at the transmitter is displayed in an ASCII format. The eight information bits are conveyed with (16,1,2)GSM-MPAPM. Figure 10 shows the processes for generating and receiving information bits at the transmitter and receiver.

**Figure 10.** The process for the generation and reception of information bits (**a**) at the transmitter; and, (**b**) at the receiver.

#### **5. Results and Analysis**

Measurements were made in a laboratory that was 4 m long, 3 m wide, and 4 m high, in the presence of other light sources in an indoor environment. It was supposed that the light emitting from the LEDs was on line of sight. At the transmitter, there were five LEDs, each with 2 W of power. The maximum output current of the pins on the C8051f330 was 25 mA, and the drive capability of the designed LED drive circuit was limited; the luminous power in the measurement was only 1.5 W. Furthermore, the C8051f330 clock influenced signal generation at the transmitter. Thus, 25 kb/s was chosen as the transmission rate of the information bits in the experiment.

This experiment firstly analyzed two activated LEDs which transmitted eight information bits at a time in the system. The different peak values of the voltage were obtained by changing the distance between the photodetector and the LEDs. The photodetector faced LED1, but all LEDs were at the same level and each was a different distance from the photodetector. Therefore, the different values were obtained according to LED combinations. In Figures 11–14 seven high-level synchronization headers were followed by three training levels. The measured peak value of the voltage decreased from 4.15 V, 2.30 V, 1.40 V, to 0.8 V when the distance between the photodetector and the LEDs increased from 10 cm, 20 cm, 30 cm, to 40 cm. It could be seen clearly that in Figure 14, when the distance between the photodetector and the LEDs was 40 cm, the measured peak value of the voltage was almost close to that of the noise. The received signals were drowned in the noise at this moment. The BER performance was also analyzed in this experiment. Figure 15 shows that the BER remained below the value of 1.5 <sup>×</sup> 10−<sup>4</sup> when the distance was within 40 cm. However, when the distance exceeded 40 cm, the value rapidly increased. The results of the peak value of the voltage and the BER performance of the received signals both demonstrated that the distance between the photodetector and the LEDs was a very crucial parameter for the proposed scheme in the indoor short-distance VLC. This was mainly due to the use of a single photodetector for the intensity modulation/direct detection demodulation at the receiver. The illumination intensity decreased as the distance increased. Once the distance exceeded a threshold that guaranteed normal communication, multiple levels of interference

appeared in the received signals. From the five figures above, it was drawn that the threshold of the distance was 40 cm. Ultimately, the aliasing signals prevented the original signals from being modulated. Furthermore, the luminous power of 1.5 W in this experiment might have been a factor that influenced the BER performance.

**Figure 11.** The amplified waveform of the received signals (10 cm).

**Figure 12.** The amplified waveform of the received signals (20 cm).

**Figure 14.** The amplified waveform of the received signals (40 cm).

**Figure 15.** The relationship between the BER (%) and the distance.

When the string of the information bits with the value of 10100101110000110011110011100110 was transmitted by Tera\_Term at the transmitter, the receiver correctly received the string. In this test, the BER performances of GSM-MPAPM, V-MPAPM, MPAPM, and ACO-OFDM were compared. Among them, the latter three were mentioned in [5,9,11], respectively. Figure 16 shows that, within the range of the observation distance that was set between 15 cm and 23 cm, as the distance increased, the proposed (16,1,2)GSM-MPAPM, V-MPAPM, and MPAPM all had a stable performance in the BER, while the BER of ACO-OFDM gradually raised. This was because when the OFDM was applied to the VLC, due to the nonlinear response of the LED, it caused a higher BER than the single carrier modulation, such as GSM-MPAPM, V-MPAPM, and MPAPM, which were compared in Figure 16. At the same time, the (16,1,2)GSM-MPAPM had a better BER performance as compared to V-MPAPM and MPAPM obviously. That is because, as the distance increased, the channel fading was getting more and more serious, the (16,1,2)MPAPM provided an increase in spatial diversity via the two activated LEDs, which causes an improvement in the BER.

**Figure 16.** Comparisons of the bit error ratios (BERs) of different modulations.

Three dimming coefficients were chosen and an illuminance curve was drawn through measurements in Figure 17 to verify the relationship between the illuminance and the distance of the (16,1,2)MPAPM. It can be seen from the curve that the illuminance decreased as the distance increased over the three different dimming coefficients. This caused a decrease in the peak value of the voltage and an increase in the BER. For p1, the system had a stable dimming effect when it was kept within the measurement distance, but the illuminances of all of the distances were slightly lower than the other two values. For p2 and p3, there were sudden changes at some values, i.e., 14 cm, 20 cm. Therefore, p1 was the best choice when a stable dimming was considered.

**Figure 17.** The relationship between the illuminance and the distance under three dimming coefficients.

#### **6. Conclusions**

In this paper, an efficient data transmission scheme was successfully implemented in a designed indoor VLC system. The experiments demonstrated that the proposed GSM-MPAPM modulation worked normally so long as the distance between a transmitter and a receiver was within 40 cm, and the BER performance was inversely proportional to the distance. Furthermore, GSM-MPAPM had a better BER performance than VMPAPM, MPAPM, and ACO-OFDM when the same distance was used. This was because activated LEDs, of which there are usually multiple, provide an increase in spatial diversity to resist communication fading. Additionally, the differentiated dimming strategy caused a stable dimming effect on the variety of light in the environment.

**Author Contributions:** This research article contains three authors, including J.-J.B., who graduated from Communication and Information System, was responsible for conceiving and designing the modulation algorithm, designing the embedded hardware system together with C.-L.H., doing the experiments, and writing the paper and the replies. C.-L.H., was charged of analyzing the data and designing the embedded hardware system together with J.-J.B., J.-F.T., who was charged for proposing many helpful suggestions on the conception, format and writing of this paper. With the author's help, this study can be completed as schedule.

**Funding:** This research was funded by [Social Science and Technology Development of Dongguan Science and Technology Bureau] grant number [2019507156746], [School-Enterprise Cooperation Horizontal Research of Dongguan Polytechnic] grant number [2018H64], [Technical R&D and service team of Dongguan Polytechnic] grant number [CXTD201802].

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
