Open Access
Add to BibSonomy
Abstract
In this paper, we propose a performance improvement of non-orthogonal multiple access (NOMA) with a three-dimensional (3D) constellation and a two-dimensional Inverse Fast Fourier Transform IFFT modulator (2D-IFFT) for the passive optical network (PON). Two kinds of 3D constellation mapping are designed for the generation of a three-dimensional NOMA (3D-NOMA) signal. Higher-order 3D modulation signals can be obtained by superimposing signals of different power levels by pair mapping. Successive interference cancellation (SIC) algorithm is implemented at the receiver to remove interference from different users. Compared with the traditional two-dimensional NOMA (2D-NOMA), the proposed 3D-NOMA can increase the minimum Euclidean distance (MED) of constellation points by 15.48%, which enhances the bit error rate (BER) performance of the NOMA. The peak-to-average power ratio (PAPR) of NOMA can be reduced by 2 dB. A 12.17 Gb/s 3D-NOMA transmission over 25 km single-mode fiber (SMF) is experimentally demonstrated. The results show that at the bit error rate (BER) of 3.8 × 10−3, the sensitivity gain of the high-power signals of the two proposed 3D-NOMA schemes is 0.7 dB and 1 dB compared with that of 2D-NOMA under the condition of the same rate. Low-power level signal also has 0.3 dB and 1 dB performance improvement. Compared with 3D orthogonal frequency-division multiplexing (3D-OFDM), the proposed 3D-NOMA scheme could potentially expand the number of users without obvious performance degradation. Due to its good performance, 3D-NOMA is a potential method for future optical access systems.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
With the continuous integration and development of new 5 G services such as metaverse, 8 K video, and industrial Internet, global network traffic has exploded in recent years. The future of data transmission relies on the availability of optical access networks capable of handling massive amounts of data. New requirements are put forward for the bandwidth and flexibility of the next-generation optical access system. Passive optical network (PON) is attractive as a potential access network design for the future because of its huge bandwidth, low-power consumption, low cost, and extensive coverage [1,2]. From the initial time division multiplexing PON (TDM-PON) [3], wavelength division multiplexing PON (WDM-PON) [4], to the current widely used orthogonal frequency-division multiplexing PON (OFDM-PON) [5–7], it has undergone many upgrades. The two main parts of OFDM are constellation mapping and Inverse Fast Fourier Transform (IFFT). Constellation mapping is to convert serial raw bit data into complex signals. IFFT transforms the OFDM multicarrier from the frequency domain to the time domain signal. The basic architecture of OFDM has not changed in recent years [8,9]. However, since the OFDM signal is represented as the superposition of orthogonal subcarrier signals, when the subcarrier signals are exactly added together with the peaks, the OFDM signal will also generate the maximum peak, leading to the high peak average power ratio (PAPR) [10–12], which eventually causes the deterioration of OFDM signal. In order to reduce the PAPR of OFDM signal, a series of research has been carried out. These research schemes are mainly divided into four categories, including signal distortion [13], signal coding [14], symbol scrambler [15] and constellation expansion [16]. The basic principle of signal distortion is to nonlinearly distort the OFDM signal's peak value and its vicinity to reduce the peak power. Clipping the part of the signal which exceeds a specified value is the simplest nonlinear distortion method. However, clipping causes signal distortion and degrades the performance of the system. Signal coding uses different encodings to produce OFDM symbols with smaller PAPR. It is worth noting that the smaller PAPR required, the fewer code groups available. Symbol scrambler is also known as selective mapping (SLM) [17,18] and can be used as a special case of signal coding. The basic principle of this method is to process multiple scrambler signals simultaneously and select the output signal with the smallest PAPR. The generated OFDM signal is also irrelevant for scrambler sequences, leading to a low PAPR. SLM does not introduce additional damage to the signal. It can reduce the probability of high PAPR, but it can’t guarantee that PAPR will be reduced below a certain value. Constellation extension schemes generate low PAPR OFDM signals by changing the position of constellation points, but this method requires a huge amount of calculation [16]. Although these methods effectively reduce the PAPR of OFDM signals, these technologies are still implemented based on 2D signal constellation and 1D-IFFT, which is still the idea of generating OFDM signals first and then reducing PAPR. A 3D signal mapper and 2D-IFFT are used to allocate 3D signals to OFDM subchannels and to modulate the signals, such 3D signals effectively inhibit PAPR from the production mechanism [8]. Based on the 3D signal mapper and 2D-IFFT modulator, the PAPR of the generated OFDM signal is reduced by 4 dB [19]. In addition, compared with the 2D-OFDM, 3D-OFDM can effectively increase the MED of constellation points, thus improving the transmission performance of the system.
A series of advanced digital signal processing techniques improve the performance of OFDM, but the above schemes are all based on orthogonal multiple access (OMA) schemes. In OMA, different optical network units occupy different orthogonal resource blocks in time or frequency domain, while non-orthogonal multiple access (NOMA) loads different users’ information to different power levels by digital domain superposition encoding to achieve the improvement of the spectral efficiency [20]. Successive interference cancellation (SIC) at the receiver is used to demodulate users’ information at different power levels. NOMA can increase the spectrum efficiency of the system and serve more users. It has been considered a potential candidate for the next optical access network and has attracted much attention from researchers. Compared with OFDM-PON, NOMA has been proven to achieve twice the data rate under the same bandwidth [21]. A NOMA multi-stack structure with synchronous downlink and asynchronous uplink is proposed, which allocates power according to user path loss and greatly improves the flexibility of the access system [22]. In our previous work, we proposed a new NOMA scheme that multiplexes spectrally overlaid signals in power domain to improve spectral efficiency and serve more users [23]. In addition, we also explored how to increase the number of users in NOMA and carried out sparse coding for signals of different power levels, thus increasing the number of access users by 150% [24]. All the above NOMA schemes are realized based on the 2D constellation mapping and 1D-IFFT. It still suffers from high PAPR and small constellation point MED. Researches on 3D-NOMA based on the 3D constellation and 2D IFFT are relatively few. The main reason is that 2D-NOMA is based on two power level Quadrature Phase Shift Keying (QPSK) superposition to get a square 16 Quadrature Amplitude Modulation (QAM) constellation. In contrast, the constellation point distribution in 3D constellation space is no longer suitable for the NOMA power superposition method, which needs to re-optimize 3D constellation distribution.
Therefore, this paper proposes a performance improvement of NOMA with 3D Constellation and 2D-IFFT Modulator for the first time. It is worth noting that the proposed 3D-NOMA scheme is implemented based on 2D-IFFT and 3D constellation pair mapping, which is perfectly compatible with TDM and OFDM. The proposed 3D-NOMA is not limited to the two schemes presented in this paper. Other three-dimensional constellations and even higher-order three-dimensional constellations can be applied in 3D-NOMA as well. We creatively explore the pair mapping scheme based on 3D square according to different quadrants and the pair mapping method based on regular tetrahedrons in 3D space. At the receiver, the SIC algorithm is used to demodulate multi-user information. Compared with the traditional QPSK mapping, the MED of regular tetrahedral constellation points increases by 15.48%, effectively improving the system's performance. 3D-NOMA can optimize transmission performance by adjusting different power distribution ratios. What’s more, this 3D-NOMA can also significantly reduce the PAPR of the system. To verify the performance of the proposed 3D-NOMA in the intensity modulation direct detector (IMDD) system, a 12.17 Gb/s 3D-NOMA signal transmission in 25 km SMF is successfully demonstrated in the experiment. Experimental results show that the proposed 3D-NOMA pair mapping scheme can effectively improve the NOMA's performance and the access system's flexibility.
2. Principle
The schematic diagram of 3D-NOMA is shown in Fig. 1, which mainly includes the 3D-OFDM modulation process and the superposition of two signals with different power distributions. It is noticed that the proposed 3D-NOMA scheme can be applied to the superposition of three levels of power signals or even more levels of different power signals. In traditional 2D-NOMA, a 16QAM square constellation is obtained by the superposition of different power after QPSK mapping and IFFT, as shown in Fig. 2. It is worth noting that for the three NOMA schemes, P1 represents high-power level and P2 represents low-power level. There are two main differences between 3D-NOMA and 2D-NOMA. First, the constellation points are no longer the traditional QPSK; Second, 3D constellation requires 2D-IFFT to obtain time domain signals. Specifically, data represents binary information sequences. These sequences were first converted into C low-rate parallel binary sequences through serial-parallel (S/P) conversion, where C is the number of subcarriers. Each low-rate bit sequence will then be mapped to a 3D constellation point. Then the signal on the kth subcarrier can be expressed as
The three elements in the matrix represent the vectors of the 3D constellation points in the x, y, and z axes. An OFDM symbol in the frequency domain can be achieved by a collection of C subcarrier signals, which can be represented as
After 3D mapping, the frequency domain OFDM symbol is converted into a time domain signal by 2D IFFT [8].
Among them, $0 \le {n_1} \le 2$ and $0 \le {n_2} \le C - 1$, where ${n_1}$ and ${n_2}$ are the indices for columns and rows of the 2-D time domain matrix ${s_1}$, respectively. It can be seen from Eq. (3) that the 2D-IFFT can be achieved by two steps 1D-IFFT. Therefore, 2D-IFFT can be simplified to
where $M_3^{ - 1}$ and $M_C^{ - 1}$ represents the IFFT matrix of 3 × 3 and C × C, respectively. Due to the signal after the above change is still a complex signal, it is not suitable for the optical access system of IMDD. Therefore, we adopt the Hermitian symmetry to transform 3D- OFDM signals into real value. Eq. (4) can be turned into:N is the number of IFFT points and ${F_{all}}$ is the frequency domain signal matrix with three rows and N columns which contains the original signal ${\textrm{F}_{3D}}$ and complex conjugate signals $\textrm{conj}({{\textrm{F}_{3D}}} )$. After 2D-IFFT, the time-domain signal can be represented as:
Then, the generated time domain signal is transformed into a series sequence by parallel-serial (P/S) conversion after adding cyclic prefix (CP).
The 3D OFDM signal can be recovered at the receiver by demodulation through the corresponding FFT. It can be found that our 3D-NOMA is compatible with existing OFDM systems except for 2D IFFT. The proposed 3D-NOMA is obtained by superimposing different power levels in the digital domain, and the superimposed signal can be expressed as:
Table 1. 3D NOMA1 mapping rules
It can be seen from Fig. 3(a) that the distribution of constellation points of high-power P1 is similar to that of QPSK, which is a square distribution, and the MED of constellation points is set to 2. To form Chen's constellation [25], a traditional 3D constellation with two different radius concentric hexahedrons, we carried out a pair mapping scheme for the 3D constellation point. Different mapping rules are designed according to the constellation points in different quadrants of a high-power level signal. Case 1 is used as an example whose mapping rules are shown in Table 1. Based on the constellation points in case 1, the positions of constellation points in other cases are flipped according to the constellation point distribution of high-power signals. Supposing the coordinates of case 1 are named as: $({{x_{case1}},{y_{case1}},{z_{case1}}} )$. Therefore, the constellation points of the other cases can be obtained by:
This way, Chen’s constellation can be achieved, as shown in Fig. 3(a). We named this method 3D NOMA1.
Further, a 3D constellation based on regular tetrahedrons is undoubtedly a better choice to maximize the MED of the constellations. Therefore, we propose a 3D-NOMA scheme based on regular tetrahedrons, shown in Fig. 3(b). The regular tetrahedron takes the origin as the body-centered extension of the regular tetrahedron, and the constellation MED is set to 2. The specific distribution of constellation points is shown in Table 2. This method is similar to 2D-NOMA, where different power levels adopt the same constellation mapping method. For a three-dimensional constellation, its transmitting power can be expressed as $\sqrt {{x^2} + {y^2} + {z^2}}$, where x, y and z respectively represent the three-dimensional coordinates of the constellation points. It can be found that the average transmitting power of constellation points in Table 2 is 1.2246, while the average transmitting power of QPSK constellation points is $\sqrt 2$ .The average transmitting power can be reduced by $\left( {\sqrt 2 - 1.2246} \right)/\sqrt 2 = 13.4\%$. Then, under the same transmitting power, the MED of 3D constellation points can be expanded $\sqrt 2 /1.2246 - 1 = 15.48\%$, which can effectively improve the transmission performance of the system. This scheme is named 3D-NOMA2.
Table 2. 3D NOMA2 mapping rules
To further verify the inherent characteristic performance of the 3D-NOMA, the theoretical symbol error probabilities (SEPs) are calculated. According to [26], the SEP of the n-dimensional constellation with a different signal-to-noise ratio (SNR) ≥ 8 dB can be expressed as:
PAPR is an important parameter affecting the transmission performance of NOMA. Therefore, we compared the complementary cumulative distribution function (CCDF) of the PAPR of the 3D-NOMA signal and the 2D-NOMA signal, as shown in Fig. 5. It can be seen that the PAPR of the two 3D-NOMA signals is significantly lower than that of 2D-NOMA schemes. It can be observed that the PAPR of the 3D-NOMA signal is reduced by at least 2 dB at the probability of 10−2 compared with the 2D-NOMA signal, which shows that 3D-NOMA has better nonlinear tolerance. On the other hand, the PAPR difference between the two 3D NOMA schemes is very small, suggesting that the proposed two methods of 3D constellation mapping have little influence on PAPR.
In the proposed 3D-NOMA, different resource blocks are multiplexed through the allocation of power levels, which are non-orthogonal. The interference of the different users needs to be eliminated. SIC is used to demodulate different users’ information from a digital field. Fig. 6 is the flow block diagram of the SIC algorithm. First, the received signals are equalized using the training sequence to obtain the corresponding channel response h1. SIC is then demodulated based on a level-by-level manner. That is, SIC restores the signal of the high-power level firstly, and the low-power level signal is regarded as noise. After demodulating the high-power level signal, its original bit information is re-modulated with the channel response h1, removing interference from the data stream with the higher power. The low-power level signal can be achieved by subtracting the new modulated high-power signal from the receiver signal. Data2 can also be demodulated after channel estimation and equalization. It is worth noting that SIC can support the superposition of user information of more power levels, which can well support the multi-power superposition scheme of NOMA.
3. Experimental setup and results
The experimental setup for evaluating the performance of the proposed 3D-NOMA technique in the next-generation optical access system is depicted in Fig. 7. Offline DSP generates the NOMA signal at the optical line terminal (OLT). Table 3 displays the NOMA signal parameter settings. Two data streams with pseudo-random binary sequences (PRBS) are mapped onto 216 subcarriers using 3D mapping. The point of IFFT used in the 3D-NOMA is 512. 216 conjugate subcarriers are applied to achieve the complex-value 3D-NOMA signal. In order to reduce inter-symbol interference (ISI) and inter-channel interference (ICI), a CP with a length of 1/16 of one NOMA symbol is used. We transmit 240 NOMA symbols, of which 230 transmit useful information, and 10 symbols are used as training sequences for channel equalization. In 3D-NOMA, each constellation point is represented by 3D coordinates, occupying three elements in the NOMA matrix while in 2D-NOMA, each constellation point only needs to be represented in the form of real part plus an imaginary part. Therefore, the signal rate of the 3D-NOMA scheme is three times lower than that of 2D-NOMA. To get the same data rate, the sampling rate of arbitrary waveform generator (AWG) in 3D-NOMA is set to 24 GSa/s, while the sampling rate of AWG in 2D-NOMA is set to 8 GSa/s. Except for the sampling rate, other parameters are consistent to ensure the same transmission rate, which is 12.17 Gb/s. The radio frequency signal emitted by the AWG passes through the electric amplifier (EA) whose gain is 22 dB and is injected into a Mach-Zehnder modulator (MZM) with 40 Ghz bandwidth. At the same time, A tunable laser (TL) with a linewidth of less than 100 kHz generates a wavelength of 1550 nm optical carrier and is also connected to MZM. The electro-optic modulation is completed in MZM, and the electrical signal is loaded into the optical signal. The modulated signal is transmitted in 25 km SMF. At the output of SMF, a power splitter (PS) is placed to distribute the signal to different optical network units (ONUs). A variable optical attenuator (VOA) adjusts the received optical power for each ONU. A photodetector (PD) is equipped to realize photo-electric conversion. After that, a mixed signal oscilloscope (MSO, TekMSO73304DX) with a sampling rate of 50 GSa/s captures the electric signal. Finally, the original data will be recovered in the offline DSP.
Fig. 7. Experimental setup (AWG: arbitrary waveform generator; EA: electric amplifier; MZM: Mach-Zehnder modulator; SMF: Single mode fiber; PS: power splitter; VOA: variable optical attenuator; PD: photodetector; MSO: mixed signal oscilloscope).
Table 3. Parameters for the 2D-NOMA and 3D-NOMA
The power distribution ratio (PDR) is an essential parameter in the NOMA system. System performance under different PDR has a significant impact. The best performance PDR of the system can be found by traversal. The optimal PDR can be obtained by calculating the average BER of the two power level signals. We have tested the BER performance of the 2D-NOMA and 3D-NOMA schemes under different PDR with the received optical power of -19 dBm. Fig. 8(a)–(c) shows the BER curves of two power level signals under different PDR. It can be found that the BER of the three schemes has similar characteristics. First, the BER of high-power level signals decreases with the increase of PDR. It is because with the increase of PDR, the signal power of high-power level signals increases, and the signal has better anti-noise performance, so it has better performance. The BER of low-power signal decreases first and then increases. When the PDR is low, more power is allocated to the low-power signal, resulting in the power advantage of the high-power signal is not obvious, and the overlap of constellation points after superposition leads to a high BER in the judgment of high-power level. However, the judgment of low-power level signals is performed based on remodulation. The error codes of high-power level signals will be introduced into the demodulation of low-power level signals, known as the error propagation effect [27]. Due to the existence of error propagation, the performance of low-power level signals will deteriorate. As shown in Fig. 6, in the process of SIC demodulation, we first demodulate the high power Data1. Then the signal of Data1 is subtracted from the received signal to achieve the signal of Data2. Once the signal in Data1 has bit errors, these errors will be introduced into the demodulation of Data2. The increase in the number of bits will eventually affect the system performance. Therefore, when PDR is low, the BER is still at a very high level even though the constellation of low-power level signals is very clear. With the increase of PDR, the BER of high-power signals decreases continuously, and the influence of the error propagation effect on low-power signals is weakened, leading to a BER decrease of the low-power signals. However, with the further increase of PDR, the power allocated by the low-power signal becomes less and less. It’s hard for lower-power levels to demodulate correctly. Fig. 8 also shows the constellation distribution of the low-power signal when the PDR is too high. The constellation point distribution is very fuzzy, which leads to the deterioration of transmission performance. The PDR corresponding to the lowest BER of the low-power signal is the optimal PDR we need. It can be seen that the optimal PDR of the three schemes is 4. The main fact that affects the optimal PDR is the structure of constellation. Different constellation structure of the three schemes determines their optimal PDR respectively.
Fig. 8. BER curves of the different PDRs at the received optical power of −19 dBm. (a) 2D-NOMA, (b) 3D-NOMA1, and (c) 3D-NOMA2
On the basis of optimized PDR, we compared the transmission performance of different NOMA schemes. Figure 9 illustrates the BER curves of three NOMA schemes with different power levels under different received optical power when PDR = 4. As can be seen from Fig. 9, with the increase of the received optical power, SNR continuously increases, and the BER of signals with different power shows a decreasing trend. Due to more power being allocated to the high-power signal, the performance of the high-power signal in the three schemes is significantly better than that of the low-power signal. The lower-power level signals of the three schemes are the first to appear error codes at the higher received optical power. At the hard-decision forward error correction (HD-FEC) BER limit of 3.8 × 10−3, the receiver sensitivity of the high-power signal of the three schemes is 1,1.5 dB and 0.8 dB higher than that of the low-power signal, respectively. We also inserted the constellation diagram of the two power-level signals with an optical power of -18 dBm. In Fig. 9(b), the constellation diagram of the low-level signals is eight constellation points. This is due to that in the process of SIC, the high-power signal is demodulated first. Then the low power signal is obtained by subtracting the high-power signal from the signal at the receiver. Since the four cases are simultaneously demodulated, there are a total of eight constellation points for all the constellation points of the four cases, which leading to the appearance of eight constellation points.
Fig. 9. BER curves of the different power levels at the PDR = 4 (a) 2D-NOMA, (b) 3D-NOMA1, and (c) 3D-NOMA2
In order to verify the superiority of the proposed 3D-NOMA scheme, we compare the high-power, low-power, and average BER performance of all signals between 3D-NOMA and 2D-NOMA. Fig. 10(a–c) shows the BER curves of performance comparison of higher-power level, lower-power level and the average BER of all signals. It can be seen from Fig. 10 that the performance of 3D-NOMA1 and 3D-NOMA2 outperformed the traditional 2D-NOMA scheme, no matter the high-power signal, low-power signal or the average BER of all signals. As shown in Fig. 10(a), for high-power signals, at the HD-FEC limit of ∼ 3.8 × 10−3, received optical power of 2D-NOMA is -18.7 dBm, while that of 3D-NOMA1 and 3D-NOMA2 is -19.4 dBm and -19.7 dBm. In other words, the sensitivity gains of 0.7 dB and 1 dB were obtained by the two 3D-NOMA schemes, respectively. Similarly, 3D-NOMA1 and 3D-NOMA2 can gain 0.3 dB and 1 dB for low-power signals, respectively. Regarding average BER performance, the receiver sensitivity of the 3D-NOMA1 signal and 3D-NOMA2 signal is increased by 0.45 dB and 1 dB. It can be attributed to the fact that 3D-NOMA can have a larger MED for the same transmission power and a lower PAPR than 2D-NOMA.
Fig. 10. Performance comparison of different power (a) BER of high-power, (b) BER of lower power, and (c) average BER of the NOMA
Further, we compared the transmission efficiency of 2D-OFDM, 3D-OFDM, and the proposed 3D-NOMA. Among them, 2D-OFDM adopts square 16QAM, while 3D-OFDM adopts Chen's 3D constellation and the 3D constellation formed by tetrahedron, which has the same constellation point distribution as 3D-NOMA. Fig. 11 presents the BER curves of 2D-OFDM, 3D-OFDM and 3D-NOMA. The transmission performance of 2D-OFDM is the worst among all schemes. For 3D-OFDM with Chen's constellation and 3D-NOMA1 with the same constellation point distribution, their BER curves are close. When the BER is 3.8 × 10−3, 3D-OFDM with Chen's constellation has a gain of about 0.2 dB compared with 3D-NOMA1.Similarly, the receiver sensitivity of 3D-OFDM with tetrahedron is increased by 0.1 dB compared with 3D-NOMA2. The main reason for the difference between 3D-OFDM and 3D-NOMA is in the different judgment methods at the receiving end. For 3D-OFDM, the minimum distance between the signal constellation points and the standard constellation point is directly used for the judgment at the receiver. However, due to the different power levels, 3D-NOMA uses SIC to demodulate two channel signals step by step. SIC has error propagation effect, and the slightly poor error performance of low-power signals leads to the increase of average BER, which ultimately results in the error performance worse than 3D-OFDM. SIC algorithm can perform two-channel estimation, which can recover signal damage more accurately and improve the BER of the system to a certain extent. The results indicate that the proposed 3D-NOMA scheme has no obvious deterioration in performance compared with 3D-OFDM. What's more, by the combination of orthogonal multiple access in the time and frequency domain as well as the NOMA in the additional power domain, 3D-NOMA can achieve higher spectral efficiency and increase the number of access users, which indicates 3D-NOMA access has a very broad application promise in future optical access system.
4. Conclusion
This paper proposes a performance improvement of NOMA with a 3D constellation and 2D-IFFT modulator. Multiplexing in the power domain can increase the number of access users and improve spectrum efficiency. Two kinds of superposition of constellations with regular tetrahedron as the basis and square in three-dimensional space as the basis are proposed. Compared with the QPSK, the MED of 3D constellation points increased by 15.48%. 3D-NOMA also achieves a 2 dB PAPR reduction. To verify the performance of the proposed 3D-NOMA, we successfully demonstrated a 12.17 Gb/s 3D-NOMA transmission over 25 km SMF. The optimal PDR of the different schemes is investigated. We compare the transmission performance of the three schemes based on obtaining the optimal PDR. The results show that the performance of 3D-NOMA signals at both power levels is better than that of 2D-NOMA signals. When the BER is 3.8 × 10−3, 3D-NOMA2 can achieve receiver sensitivity gains of 1 dB of the high-power level signal and the low-power level signal, respectively. 0.7 dB and 0.3 dB gain of high-power and low-power levels can also be achieved for 3D-NOMA1 when compared with 2D-NOMA. Furthermore, for the same constellation distribution, the performance of the proposed 3D-NOMA is not noticeably different from that of the 3D-OFDM. Given these benefits, we anticipate a great deal of success for 3D-NOMA in the near and distant futures of PON.
Funding
National Key Research and Development Program of China (2020YFB1805801); National Natural Science Foundation of China (62205151, 62225503, 61835005, 62275127, 62171227, U2001601, 62035018, 61935005, 61935011, 61975084); Jiangsu Provincial Key Research and Development Program (BE2022079, BE2022055-2); The Natural Science Foundation of the Jiangsu Higher Education Institutions of China (22KJB510031); The Startup Foundation for Introducing Talent of NUIST.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. F. J. Effenberger, “Next generation PON: lessons learned from G-PON and GE-PON,” in 2009 35th European Conference on Optical Communication, (2009), 1–3.
2. X. Xu, B. Liu, X. Y. Wu, L. J. Zhang, Y. Y. Mao, J. X. Ren, Y. Zhang, L. Jiang, and X. J. Xin, “A robust probabilistic shaping PON based on symbol-level labeling and rhombus-shaped modulation,” Opt. Express 26(20), 26576–26589 (2018). [CrossRef]
3. K. L. Hall, “High-speed TDMA techniques,” in Opt. Fiber Commun. Conf.2000, Paper ThV4.
4. J. Lee, K. Choi, J. Moon, and C. Lee, “Seamless Upgrades From a TDM-PON With a Video Overlay to a WDM-PON,” J. Lightwave Technol. 27(15), 3116–3122 (2009). [CrossRef]
5. J. Armstrong, “OFDM for optical communications,” J. Lightwave Technol. 27(3), 189–204 (2009). [CrossRef]
6. I. N. Cano, X. Escayola, P. Schindler, M. Santos, V. Polo, J. Leuthold, I. Tomkos, and J. Prat, “Experimental Demonstration of a Statistical OFDM-PON With Multiband ONUs and Elastic Bandwidth Allocation [Invited],” J. Opt. Commun. Netw. 7(1), A73–A79 (2015). [CrossRef]
7. X. F. Hu, P. Cao, L. Zhang, L. P. Jiang, and Y. K. Su, “Energy-efficient optical network units for OFDM PON based on time-domain interleaved OFDM technique,” Opt. Express 22(11), 13043–13049 (2014). [CrossRef]
8. S. G. Kang, “An OFDM with 3-D signal mapper and 2-D IDFT modulator,” IEEE Commun. Lett. 12(12), 871–873 (2008). [CrossRef]
9. S. Cho and S. K. Park, “Improved 16-ary constellation mapping for three-dimensional OFDM systems,” Electron. Lett. 48(9), 530–532 (2012). [CrossRef]
10. Y. T. Wu, C. C. He, Q. W. Zhang, Y. Z. Sun, and T. Wang, “Low-complexity recombined SLM scheme for PAPR reduction in IM/DD optical OFDM systems,” Opt. Express 26(24), 32237–32247 (2018). [CrossRef]
11. X. J. Liang, W. Li, W. D. Ma, and K. Wang, “A simple peak-to-average power ratio reduction scheme for all optical orthogonal frequency division multiplexing systems with intensity modulation and direct detection,” Opt. Express 17(18), 15614–15622 (2009). [CrossRef]
12. W. O. Popoola, Z. Ghassemlooy, and B. G. Stewart, “Pilot-assisted PAPR reduction technique for optical OFDM communication systems,” J. Lightwave Technol. 32(7), 1374–1382 (2014). [CrossRef]
13. D. Guel and J. Palicot, “Analysis and comparison of clipping techniques for OFDM peak-to-average power ratio reduction,” in Proc. International Conference on Digital Signal Processing, Jul. 2009, pp. 1–6.
14. R. Chandrasekhar, M. Kamaraju, M. V. S. Sairam, and G. T. Rao, “PAPR reduction using combination of precoding with Mu-Law companding technique for MIMO-OFDM systems,” 2015 International Conference on Communications and Signal Processing (ICCSP), 2015, pp. 0479–0483.
15. Z. Shen, X. Yang, H. He, and W. Hu, “Secure transmission of optical DFT-S-OFDM data encrypted by digital chaos,” IEEE Photonics J. 8(3), 1–9 (2016). [CrossRef]
16. B. S. Krongo and D. L. Jones, “PAPR reduction in OFDM via active constellation extension,” IEEE Trans. Broadcast. 49(3), 258–268 (2003). [CrossRef]
17. L. Chen, T. Xu, W. Jiang, J. Zhang, and Q. Yang, “Integrating SLM Algorithm into Neural Network for Reducing the PAPR of OFDM System,” in Asia Communications and Photonics Conference2021. [CrossRef]
18. X. M. Song, B. Liu, Y. Bai, S. D. Chen, J. X. Ren, Y. Y. Mao, H. X. Zhang, X. Y. Wu, Y. F. Wu, and X. J. Xin, “SCMA-OFDM PON based on chaotic-SLM-PTS algorithms with degraded PAPR for improving network security,” Opt. Lett. 47(20), 5293–5296 (2022). [CrossRef]
19. L. Deng, X. L. Wang, C. Zhou, M. Tang, S. N. Fu, M. M. Zhang, P. P. Shum, and D. M. Liu, “Experimental Demonstration of a 16.27 Gb/s 2-D Coherent Optical OFDM System With 3-D Signal Mapper and 2-D IFFT Modulator,” J. Lightwave Technol. 34(4), 1177–1183 (2016). [CrossRef]
20. A. A. Khansa, Y. Yin, G. Gui, and H. Sari, “Power-domain NOMA or NOMA-2000?” in Proc. 25th Asia-Pacific Conf. Commun., 2019, pp. 336–341.
21. B. Lin, Z. Ghassemlooy, X. Tang, Y. Li, and M. Zhang, “Experimental demonstration of an NOMA-PON with single carrier transmission,” Opt. Commun. 396, 66–70 (2017). [CrossRef]
22. F. Lu, M. Xu, L. Cheng, J. Wang, and G. K. Chang, “Power-division non-orthogonal multiple access (NOMA) in flexible optical access with synchronized downlink/asynchronous uplink,” J. Lightwave Technol. 35(19), 4145–4152 (2017). [CrossRef]
23. Y. B. Wan, B. Liu, J. X. Ren, R. Ullah, Y. Y. Mao, S. Y. Zhu, S. D. Chen, X. Y. Wu, F. Wang, T. T. Sun, Y. F. Wu, and L. L. Zhao, “Performance-Enhanced Optical Non-Orthogonal Multiple Access Enabled by Orthogonal Chirp Division Multiplexing,” J. Lightwave Technol. 40(16), 5440–5449 (2022). [CrossRef]
24. S. D. Chen, B. Liu, J. X. Ren, Y. Y. Mao, X. Y. Wu, R. Ullah, X. M. Song, Y. Bai, Y. B. Wan, Y. F. Wu, L. L. Zhao, and T. T. Sun, “Power Sparse Code Division Non-Orthogonal Multiple Access Scheme for Next-Generation Flexible Optical Access,” J. Lightwave Technol. 40(22), 7236–7245 (2022). [CrossRef]
25. Z. Chen and S. G. Kang, “Probability of symbol error of OFDM system with 3-dimensional signal constellations,” Proc. Int. Symp. on Consumer Electronics, Kyoto, Japan, May 2009, pp. 442–446
26. J. G. Proakis and M. Salehi, Digital Communications, McGraw Hill, New York, NY, USA2008, pp. 160–289.
27. H. Li, Z. Huang, Y. Xiao, S. Zhan, and Y. Ji, “Solution for error propagation in a NOMA-based VLC network: Symmetric superposition coding,” Opt. Express 25(24), 29856–29863 (2017). [CrossRef]
