Open Access
Add to BibSonomyAbstract
We propose a non-orthogonal multiple access (NOMA) scheme combined with orthogonal frequency division multiplexing access (OFDMA) for visible light communications (VLC), which offers a high throughput, flexible bandwidth allocation and a higher system capacity for a larger number of users. Bidirectional NOMA-OFDMA VLC is experimentally demonstrated. The effects of power allocation and channel estimation on the bit error rate performance are investigated. The experiment results indicate that accurate channel estimation can eliminate the inter-user interference more effectively. The optimum power allocation ratios for uplink and downlink are both about 0.25 in the case of two users.
© 2017 Optical Society of America
1. Introduction
Visible light communications (VLC) based on commercial white light-emitting diodes (LEDs) has attracted much attention from academic and industry, due to its advantages such as license free spectrum, free from electromagnetic interference and inherent high security [1,2]. Most research works reported on VLC systems are for point to point transmission using equalization techniques, advanced modulation schemes, and multiple-input-multiple-output (MIMO) technology to achieve higher date rate [3–6]. However, multiple access (MA) support for VLC is essential to provide multi-user wireless services. The conventional MA techniques such as time division multiple access (TDMA), frequency division multiple access (FDMA) and code division multiple access (CDMA) together with some optical MA techniques such as wavelength division multiple access (WDMA) and space division multiple access (SDMA) have already been proposed for VLC [7]. Orthogonal frequency division multiplexing access (OFDMA) and interleaved frequency division multiple access (IFDMA) are also invesitgated as promising MA schemes for uplink VLC with high spectral efficiency and high tolerance against multipath induced distortion and synchronization error [8].
Power domain multiple accesss, also known as non-orthogonal multiple access (NOMA) has recently been proposed as a promising solution to enhance the spectral efficiency for the 5th generation (5G) wireless networks [9–11]. In contrast to orthogonal multiple access (OMA) techniques such as TDMA and FDMA, where users are allocated individual time-frequency (TF) resources, NOMA superposes user messages in the power domain and uses successive interference cancellation (SIC) at the receivers to separate the users, so that all of the users can use the whole TF resources. The NOMA has also been adopted for downlink VLC systems to enhance the throughput and improve the system capacity [12–14]. However, these above works were mainly simulation based with no experimental verifications. In addition, no work has been reported on the uplink VLC based on NOMA.
In this paper, we propose a NOMA-OFDMA scheme for bidirectional VLC transmission, which offers a high throughput, high tolerance against multipath induced distortion, high spectral efficiency and a higher system capacity for a larger number of users. The feasibility of the bidirectional NOMA-OFDMA VLC is verified with experiment demonstration. As shown in the experiment results, the optimum power allocation ratios (PARs) for uplink and downlink VLC are both about 0.25. We also investigate the effect of channel estimation on the bit error rate (BER) performance. Since intra symbol frequency averaging (ISFA) and minimum mean square error (MMSE) perform better channel estimation than least square (LS) [15], they can eliminate the inter-user interference more effectively
2. Technique principle
Figure 1 shows the schematic diagram of downlink NOMA-OFDMA VLC with N users. For simplicity, we assume each user uses the whole TF resources. In combination with OFDMA where each user uses a set of subcarriers to transmit or receive its data, more flexible bandwidth allocation can be achieved to support more users. In the transmitter (Tx), the source data for each user is mapped and encoded into OFDM symbol (x1, x2, . . ., xN) prior to power allocation, respectively. Then all the OFDM signals are combined with a total transmitted power of P. The final transmitted time-domain signal can be written as:
where pi is the allocated power for user i, xi is the transmitted time-domain OFDM signal for user i. The combined digital OFDM signal is converted into analog signals using a digital-to-analog converter (DAC). Following the inclusion of the direct current (DC) bias, the DC- OFDM is used for intensity modulation (IM) of a LED. After the wireless optical channel, the received signal can be represented as:the frequency-domain representation of which can be written as:where h and H are channel coefficients represented in time-domain and frequency-domain respectively, w and W are noises represented in time-domain and frequency-domain respectively. denotes the convolution operation. Xi is the frequency-domain representation of xi. We assume that p1 > p2 > p3…> pN. At the receiver (Rx), the optical signal is detected by a photo-detector and then converted into a digital format using an analog-to-digital converter (ADC). The output of the ADC is then passed through a frame synchronization module prior to removing the cyclic prefix (CP). After the discrete Fourier transform (DFT) operation, the received signal for user 1 can be obtained by dividing Y by , which can be written as:Note that H can be calculated from channel estimation using the training sequence inserted in the preamble. The transmitted signal of user 1 (i.e., s1) is recovered after demapping Y1. After removing the term of in (3), the received signal is divided by , which can be written as:The transmitted signal of user 2 (i.e., s2) can be recovered after demapping Y2. The decoding order of SIC is in the order of increasing channel gain (i.e., ). Finally, the received signal for user N can be written as:The transmitted signal of user N (i.e., sN) can be obtained after demapping without inter-user interference.
Fig. 1 Block diagram of downlink NOMA-OFDMA VLC (DFT: discrete Fourier transform, DAC: digital-to-analog converter, ADC: analog-to-digital converter, DC: direct current, CP: cyclic prefix).
Figure 2 shows the schematic diagram of uplink NOMA-OFDMA VLC. In the uplink, the OFDM signal for each user (xi) is generated and then DC-level shifted prior to IM of a LED, respectively. After free space transmission, the received signal can be written as:
the frequency-domain representation of which can be written as:where hi and Hi are channel coefficients represented in time-domain and frequency-domain for user i, respectively, which can be calculated from the inserted training sequences in the preambles respectively. We assume the received optical power from a user with a lower index is greater than that with a higher index (i.e., H1>H2…>HN). At the Rx, following optical to electrical conversion, the regenerated OFDM signal is converted into a digital format using an ADC. The output of the ADC is then passed through a frame synchronization module prior to removing the CP. After DFT operation, the received signal for user 1 can be obtained by dividing Y by , which can be written as:The transmitted signal of user 1 (i.e., s1) is recovered after demapping Y1. After removing the term of in (8), the received signal is divided by , which can be written as:The transmitted signal of user 2 (i.e., s2) can be recovered after demapping Y2. The decoding order of SIC is in the order of increasing channel gain (i.e., ). Finally, the received signal for user N can be written as:After demapping, the transmitted signal for user N is obtained without inter-user interference. Figure 3 shows the power and frequency resources allocation for our proposed NOMA-OFDMA VLC. As shown in Fig. 3, each user can use all the subcarrier frequencies or a set of subcarriers to transmit or receive its data. Therefore, NOMA-OFDMA can accommodate significantly more users than other MA schemes.3. Experiment setup and results
Figure 4 shows the experimental setup for downlink NOMA-OFDMA VLC with two users. At the Tx, two 1.7-Mbaud baseband quadrature phase shift keying (QPSK) OFDM signals are three times up-sampled and then up-converted to 1.25 MHz by means of digital I-Q modulation. The two OFDM signals are combined after power allocation and then uploaded to an arbitrary waveform generator (AWG) operating at 5 MS/s. The DFT and CP sizes are 256 and 8, respectively. The generated waveform is converted into analog streams and then DC-level shifted using the bias Tee prior to IM of a commercially available phosphorescent white LED. At the Rx, a commercial optical Rx (THORLABS PDA10A) is used to convert the optical signal back into the electrical signal. The optical Rx output is passed through ADC and captured using a real-time digital oscilloscope for offline signal processing in order to recover the transmitted data. All the key system parameters are provided in Table 1.
Table 1. System parameters
Figure 5 shows the bit error rate (BER) as a function of the distance between the transmitter and receiver. The PAR between the two users is set to 0.09, 0.16, 0.25, 0.36, and 0.49, respectively. Each BER is calculated from the average of the two users, which is based on more than 1 105 bits. As shown in Fig. 5, the best BER performance is achieved with a PAR of 0.25. Figure 6 shows the BER performance of the two users with LS, ISFA, and MMSE channel estimation methods. In our previous work [15], we have shown that both MMSE and ISFA perform better channel estimation than LS. As shown in Fig. 6, both MMSE and ISFA can eliminate the inter-user interference more effectively than LS. At the Tx, user 1 is allocated with more power. At the Rx, the data of user 1 is decoded prior to decoding the data of user 2. If the data demodulation of user 1 cannot be accurately realized, the data of user 2 cannot be recovered with error free. As such, the BER performance of user 1 is better than that of user 2.
The experimental setup for uplink NOMA-OFDMA VLC with two users is shown in Fig. 7. The two OFDM signals are generated and uploaded into the AWGs respectively. The outputs of the AWGs are converted into analog streams and then DC-level shifted using the bias Tee prior to IM of two commercially available phosphorescent white LEDs, respectively. At the Rx, the optical signal is detected by a commercial optical Rx. Since user 1 is more close to Rx as shown in Fig. 7, so the received optical power from user 1 is greater than that from user 2. We can change the position of the Rx in the x-direction to adjust the received power ratio of user 1 to user 2. The x-coordinate of user 1 is zero. The electrical OFDM signal is captured by the scope and then decoded offline. All the key system parameters are provided in Table 1. Figure 8 shows the preamble structure for uplink VLC. For each user, the preamble includes two synchronization sequences and a time-multiplexed training sequence, which are used for frame synchronization and channel estimation, respectively.
Figure 9 shows the average BER performance for uplink VLC based on NOMA-OFDMA. In the experiment, the distance between Tx and Rx and the distance between users are set to 15 cm and 10 cm, respectively. The position of the Rx is changed in the x-direction within the range of −6 to 4 cm. LS channel estimation method is used. Since user 1 is more close to the Rx, more optical power is received from user 1. As such, the data of user 1 is decoded prior to decoding the data of user 2. With the increase of x coordinate value, the PAR between the two users increases. The best BER performance is achieved with the Rx being placed at the point with an x coordinate of 2 cm as shown in Fig. 9, in which the PAR is about 0.25. Figure 10 shows the BER performance as a function of the distance between Tx and Rx with the Rx being placed at the point with an x coordinate of 2 cm. The BER performance of user 1 is better than user 2, which is similar to downlink transmission.
4. Conclusions
We proposed an experimental demonstration of a NOMA-OFDMA scheme for bidirectional VLC transmission, which provided a high throughput, flexible bandwidth allocation and a higher system capacity for a larger number of users. The experimental results showed that the optimum PARs for uplink and downlink are both about 0.25. Efficient channel estimation methods such as MMSE and ISFA could eliminate the inter-user interference more effectively. The experimental results reveal that NOMA-OFDMA is a promising multiple access scheme for both downlink and uplink of VLC networks.
Funding
This work was supported in part by the Chunmiao Project of Haixi Institutes, Chinese Academy of Sciences, in part by the National Natural Science Foundation of China under Grant 61601439 and 61501427, in part by External Cooperation Program of Chinese Academy of Sciences under Grant 121835KYSB20160006, in part by State Key Laboratory of Advanced Optical Communication Systems and Networks, China.
References and links
1. Z. Ghassemlooy, W. Popoola, and S. Rajbhandari, Optical Wireless Communications: System and Channel Modelling With MATLAB (Taylor & Francis, 2012).
2. L. Wu, Z. Zhang, J. Dang, and H. Liu, “Adaptive modulation schemes for visible light communications,” J. Lightwave Technol. 33(1), 117–125 (2015). [CrossRef]
3. C.-H. Yeh, H.-Y. Chen, C.-W. Chow, and Y.-L. Liu, “Utilization of multi-band OFDM modulation to increase traffic rate of phosphor-LED wireless VLC,” Opt. Express 23(2), 1133–1138 (2015). [CrossRef] [PubMed]
4. B. Lin, X. Tang, Z. Ghassemlooy, X. Fang, C. Lin, Y. Li, and S. Zhang, “Experimental Demonstration of OFDM/OQAM Transmission for Visible Light Communications,” IEEE Photonics J. 8(5), 7906710 (2016). [CrossRef]
5. T. Fath and H. Haas, “Performance comparison of MIMO techniques for optical wireless communications in indoor environments,” IEEE Trans. Commun. 61(2), 733–742 (2013). [CrossRef]
6. G. Cossu, A. M. Khalid, P. Choudhury, R. Corsini, and E. Ciaramella, “3.4 Gbit/s visible optical wireless transmission based on RGB LED,” Opt. Express 20(26), B501–B506 (2012). [CrossRef] [PubMed]
7. H. Elgala, R. Mesleh, and H. Haas, “Indoor optical wireless communication: potential and state-of-the-art,” IEEE Commun. Mag. 49(9), 56–62 (2011). [CrossRef]
8. B. Lin, X. Tang, H. Yang, Z. Ghassemlooy, S. Zhang, Y. Li, and C. Lin, “Experimental Demonstration of IFDMA for Uplink Visible Light Communication,” IEEE Photonics Technol. Lett. 28(20), 2218–2220 (2016). [CrossRef]
9. Y. Saito, Y. Kishiyama, A. Benjebbour, T. Nakamura, A. Li, and K. Higuchi, “Non-Orthogonal Multiple Access (NOMA) for Cellular Future Radio Access,” in Proceedings of IEEE Conference on Vehicular Technology (IEEE, 2013), pp. 1–5. [CrossRef]
10. Y. Saito, A. Benjebbour, Y. Kishiyama, and T. Nakamura, “System Level Performance Evaluation of Downlink Non-Orthogonal Multiple Access (NOMA),” in Proceedings of IEEE Annual Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC), (IEEE, 2013), pp.611–615. [CrossRef]
11. Z. Ding, Z. Yang, P. Fan, and H. Poor, “On the performance of nonorthogonal multiple access in 5G systems with randomly deployed users,” IEEE Signal Process. Lett. 21(12), 1501–1505 (2014). [CrossRef]
12. H. Marshoud, V. M. Kapinas, G. K. Karagiannidis, and S. Muhaidat, “Non-Orthogonal Multiple Access for Visible Light Communications,” IEEE Photonics Technol. Lett. 28(1), 51–54 (2016). [CrossRef]
13. R. C. Kizilirmak, C. R. Rowell, and M. Uysal, “Non-orthogonal multiple access (NOMA) for indoor visible light communications,” IEEE International Workshop on Optical Wireless Communications, (2015), pp. 98–101. [CrossRef]
14. L. Yin, X. Wu, and H. Haas, “On the performance of non-orthogonal multiple access in visible light communication,” IEEE International Symposium on Personal, Indoor, and Mobile Radio Communications, (IEEE, 2015), pp. 1354–1359.
15. B. Lin, X. Tang, Z. Ghassemlooy, S. Zhang, Y. Li, Y. Wu, and H. Li, “Efficient Frequency Domain Channel Equalization Methods for OFDM Visible Light Communications,” IET Commun. 11(1), 25–29 (2017). [CrossRef]
