Open Access
Add to BibSonomy
Abstract
We propose a novel security-enhanced power division multiplexing (SPDM) optical non-orthogonal multiple access scheme in conjunction with seven-core optical fiber in this paper. This scheme could improve the security of data transmission at the physical layer and the split ratio of the access network, ensuring more users can be served at the same time. Additionally, multiple signals can be superimposed in the digital domain, leading to a significant improvement in spectral efficiency. We have further experimentally demonstrated the transmission of 47.25 Gb/s SPDM orthogonal frequency division multiplexing (OFDM) signals in a 2 km seven-core fiber system. The experimental results confirm that our scheme can increase the number of access users by 14 times without influencing the privacy of different users. It is worth mentioning the signal encryption method based on amino acids combine with extended zigzag is proposed for the first time as we know. Meanwhile, the key space reaches 10182, indicating that the data transmission process can be effectively protected from the attack of stealers. The proposed security-enhanced power division multiplexing space division multiplexing passive optical network (SPDM-SDM-PON) support multi-threading and multi-functions, showing a great potential to be applied in the future telecommunication systems.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
With the growing number of service types with various traffic patterns, such as enterprise/campus connectivity, video surveillance and monitoring, and virtual reality/augmented reality, the network traffic has exploded. Owning to the ‘smart’ technology, access networks with large bandwidth and capacity could meet such requirements [1,2]. Passive optical networks (PON) have attracted attention as an efficient solution for the “last mile” problem due to their low costs, long transmission distances, multiple supportive services, and large capacities. All optical network units (ONUs) under the optical line terminal (OLT) of time division multiplexing (TDM)-PON could utilize the bandwidths more effectively. Upstream and downstream data transmission occupy a certain wavelength with only one transceiver in the OLT, thus it is expected to be more cost effective. However, redundant bytes are needed to distinguish different ONUs, causing several problems involving insufficient bandwidth and poor capacity expansion [3,4]. Besides, wavelength division multiplexing (WDM)-PON supplies multiple available wavelengths and each wavelength independently corresponds to an ONU. The privacy of different users can be guaranteed. However, the number of wavelengths limits the number of simultaneous accesses. And high price of optical components limits substantial investments of WDM-PON [5,6]. In addition, optical code division multiple access (OCDMA)-PON is proposed, which adopts broadcast mode in the downstream direction. To maintain high confidentiality and low delay, different optical address codes are allocated to each user in the same wavelength channel in the upstream direction. As implicitly mentioned above, OCDMA-PON can achieve high security, large capacity, tunability and high rate, performing better than the other PONs. However, how to correctly arrange the address code is a complex process, delaying the development speed of OCDMA-PON [7].
Optical orthogonal frequency division multiplexing (OFDM) is a parallel transmission technology in which high-speed serial data is decomposed into multiple low-speed parallel. These data can be dynamically allocated to multiple sub-carriers with orthogonal spectrum, leading to an improvement of the spectrum efficiency and flexibility of bandwidth resource [8,9,10]. OFDM shows a potential to meet the requirements of transparent service transmission, independent protocol and wireless convergent access, which is regarded as one of the strongest competitors for the optical access network solution. OFDM based on intensity modulation/direct detection (IM/DD) [11,12] has been widely studied in access networks in order to improve the spectrum efficiency and lower the implementation cost [13,14]. However, normal OFDM allocates different carriers to different users independently, thus the quality of services is uneven. While some users are enjoying an eximious experience, others are complaining about the telecom carriers. This reflects the extremely uneven distribution of resources across the network, poor fairness is the common failings of orthogonal multiple access (OMA). For OMA, each user occupies orthogonal resources relatively independently, and there is no inter-user interference. As a result, the complexity and cost of the receiver are relatively low. However, OMA has certain technical defects: the number of users carried on each orthogonal resource is limited, which reduces the spectral efficiency of the system and the number of users served. In this term, OMA can not reach the theoretical multi-user capacity limit; OMA requires authorized transmission, which consumes high transmission delay and signaling overhead. Non-orthogonal multiple access (NOMA) [15,16] allows the signals of multiple users to be superimposed on the same time/frequency/code domain resources, which can improve spectral efficiency and system capacity, improve user fairness, increase the number of device connections, reduce transmission delay and signaling overhead, and has relatively loose channel feedback. What’s more, NOMA can be better compatible with OMA. Conclusively, in order to make up for the shortcomings of the OMA scheme, NOMA is considered to be a promising new type of multiple access technology.
The typical NOMA schemes mainly include: power-domain non-orthogonal multiple access (PD-NOMA), sparse code multiple access (SCMA), pattern division multiple access, (PDMA), multi-user shared access (MUSA) and interleave-division multiple access (IDMA), etc. This paper focuses on PD-NOMA. PD-NOMA could superimpose the signals of multiple users based on the same time-frequency resources and distinguish the signals in the power domain according to the successive interference cancellation (SIC) [17,18] multi-user detection algorithm. The power distribution ratio (PDR) can also be optimized based on the quality of service (QoS) of each user. The multi-carrier NOMA system can be regarded as a hybrid multiple access system, originating from the combination of OMA and NOMA technologies. In such system, users could be allocated to different groups which are associated with the corresponding orthogonal resource blocks. The users in the same group share the time and frequency resources with the implementation of NOMA scheme. It’s demonstrated that the PD-NOMA based on OFDM [19,20] will provide higher spectrum efficiency, better throughput, more equipment accesses and low latency. Generally, the independent OFDM signals are combined with certain power ratios using superposition coding in the digital domain. This enables that different signals can be transmitted in the same time slot and spatial channel. By combining with space division multiplexing (SDM), the optical splitting ratio of OFDM-based PD-NOMA can be manifestly increased. NOMA-carrierless amplitude phase modulation (CAP) and multicore fiber (MCF) have been combined in Ref. [21] to achieve a total aggregated traffic of 490 Gb/s. The transmission of NOMA combined with multi-band CAP signals through a 2 km MCF with seven spatial channels has been demonstrated in Ref. [22]. The results show transmission penalty can be neglected with different total aggregated traffics ranging from 350 Gb/s to 630 Gb/s. MCF allows multiple spatial channels to be multiplexed in parallel, playing an important role for future requirements for high channel count optical connections. It facilitates the combination of fiber optics and photonic integrated circuits (PICs), overall integration can not only save hardware cost but also reduce volume. Moreover, large core-count cladding pumped multi-core erbium doped fiber amplifiers are compatible with equivalent MCFs and amplify several lower core-count MCFs. Therefore, we use MCF for experimental verification in this paper. It’s worth mentioning that considering the data of several users are superimposed on the same resource block during transmission, there is a potential risk of data leakage for PD-NOMA. Also, PON is vulnerable to eavesdropping and interference attacks [23]. Therefore, the security problem of PD-NOMA-PON needs to be solved urgently. Currently, few articles address the safety of PD-NOMA-PON.
In this work, we propose and experimentally verify dual-channel quadrature phase shift keying (QPSK) OFDM-based security-enhanced power division multiplexing (SPDM) -PON in a seven-core fiber transmission system. In our proposal, the signals of two users which are modulated by QPSK are superimposed in the power domain, and then can be transmitted in the same core. Simultaneously, we use seven-core fiber in our scheme. Traditional optical communication adopts single-core optical fiber, each core transmits signals of one user, so compared with the traditional program, the number of PON access users can be increased by 14 times, improving the throughput of PON downstream effectively. To evaluate the potency, we compare the performances of our method with 16 QAM-based OFDM-PON, because they have almost the same spectral efficiency. SPDM could perform amino acid encryption on the transmitted signals at bit level and extended zigzag encryption at symbol level, facilitating the information security effectively. We successfully demonstrate the transmission of 47.25 Gb/s encrypted optical SPDM OFDM signals over 2 km seven-core fiber with enhanced security of physical layer.
2. Principles
2.1 Power domain NOMA
Figure 1 shows the architecture of the NOMA access network. As shown in the figure, the flexible allocation of electrical domain power ameliorates the overall performance of the PON. On the OLT side, ${s_i}(t)$ represents the signal to be sent to the corresponding user, so the sum of all the transmitted signals can be written as $x(t) = \sum\nolimits_{i = 1}^n {\sqrt {{p_i}} } {s_i}(t)$, where ${p_i}$ denotes the power occupied by the ${i^{th}}$ signal in the digital domain. The powers of electrical signals for different users are not equal because they are connected with different communication links. But the total power is ${p_{total}} = \sum\nolimits_{i = 1}^n {{p_i} = 1}$. When transmitted through the optical fiber, the received signal on the ONU side can be expressed as $y(t) = \sum\nolimits_{i = 1}^n {{h_i}(t)\sqrt {{p_i}} {s_i}(t) + n(t)}$, where ${h_i}(t)$ is the corresponding channel response of the ${i^{th}}$ signal and $n(t)$ is the additive noise. In order to describe our SPDM solution deeply with universal applicability, we try to adopt two users and one eavesdropper to illustrate SPDM-PON in detail.
2.2 SPDM-PON
As illustrated by Fig. 2, serial-to-parallel (S/P) conversion is performed for the two-channel signals to obtain $N \times 2M$ matrix on the OLT side.$N$ represents the number of symbols carried by each subcarrier in the OFDM transmission, and M represents the number of carriers. Before QPSK mapping, the acid encryption process is used to address the hidden security risks that emerge in the physical layer. The types of amino acids are inseparable from deoxyribonucleic acid (DNA). DNA is mainly composed of four parts, namely adenine (A), guanine (G), thymine (T) and cytosine (C). In the DNA, A and T are paired while C and G are paired. There are 8 types of DNA coding rules as shown in Table 1. Two adjacent bits in the same row could form a DNA-like base, generating a $N \times M$ matrix. Here, the chaotic sequence ${R^\chi }$ determines the bit coding rules of each row.
Table 1. DNA Encoding and Decoding Rules
As shown in Fig. 3, DNA is transcribed to synthesize messenger ribonucleic acid (mRNA), the amino acid is composed of three codons, which can be transported by transfer ribonucleic acid (tRNA). In the similitude of DNA, RNA also contains four types of bases, namely A, uracil(U), C and G, where A and U (or C and G) are paired. Figure 4 depicts the production process of amino acid. After the translation, we can get $N \times \frac{M}{3}$ amino acid matrix, where each amino acid has a specific number correspondence to the codon. And the coded amino acid matrix could be scrambled. In this process, ${R^\alpha }$ and ${R^\beta }$ control the exchange positions of rows and columns, respectively.
With the scrambling process completed, the coded amino acids could be transformed into bases inversely, which then are decomposed into bit sequences according to the coding rules that are mainly decided by ${R^\delta }$ (see Fig. 5). Finally, we could obtain a completely scrambled $N \times 2M$ matrix with bit-level encryption, which could further be transformed into a $N \times M$ complex matrix after QPSK mapping. As we know, joint encryption contributes to the enhancement of system’s anti-attack ability in exponential index. In addition to the first-level encryption, we also adopt the extended zigzag-type second-level encryption in the article. In the standard zigzag transformation, the sweep order is from the upper left corner to the lower right corner. For a $4 \times 4$ matrix, the position of each element (e.g., 1, 2, 15, and 16) will not change after countless iterations with the adoption of standard zigzag. The position certainty of the element is determined by the fatal flaw of this transformation. In this paper, we scan the elements from the lower right corner to the upper left corner. Lower left corner to the upper right corner or the upper right corner to the lower left corner can also change the position of each element. The standard zigzag transform is $4 \times 4$, we extend it to $12 \times 12$ in this paper. As illustrated in Fig. 6, the encryption scheme uses three transformation rules, which improves the efficiency of scrambling. Changing rule is controlled by ${R^\varepsilon }$ and the number of scrambling is controlled by ${R^\phi }$. For such extended zigzag transformation, all the elements can be scrambled with a limited number of scrambling.
Fig. 5. Schematic diagram of the effect of amino acid encryption scrambling: (a) the matrix is N rows after amino acid transformation, (b) the matrix is column after amino acid transformation, (c) fully scrambled amino acid matrix.
Fig. 6. Extended zigzag transformation process: (a) Rule 1 of zigzag transformation, (b) Rule 2 of zigzag transformation, (c) Rule 3 of zigzag transformation.
The chaotic sequences $X,Y,Z,U,V,W$ are generated by the 6D-hyperchaotic mapping, of which model can be expressed as
Fig. 7. Block flow diagram of ${R^\alpha },{R^\beta },{R^\chi },{R^\delta },{R^\varepsilon }$ and ${R^\phi }$.
With the completion of two-level encryption, inverse fast Fourier transform (IFFT) is performed. At last, ${x_1}$ and ${x_2}$ are obtained after OFDM modulation process, where cyclic prefix (CP) and parallel-to-serial (P/S) conversion are performed. The powers of the corresponding signals are ${p_1}$ and ${p_2}$ (${p_1} \ne {p_2}$, ${p_2} = \sqrt {1 - {p_1}}$), respectively. The PDR in the electrical domain can be calculated by $PDR = {{{p_1}} / {{p_2}}}$. With the superimposition of these two signals, we could obtain
of which time-domain forms can be expressed as: and the received frequency domain signal can be expressed as: where ${H_1}$ and ${H_2}$ are the channel responses, while N is the additive noise in the frequency domain.In the receiving end, the SIC algorithm is used to sequentially demodulate the entangled signals, based on the power value. The signal ${x_1}$ with the strongest power is demodulated firstly while all other signals ${x_2} \cdots {x_n}$ are treated as noise. Only when the decryption for the signal is completed with the correct key, the ONU1 can obtain the accurate information. In this work, we take advantage of two signals, where ${x_2}$ is regarded as the noise. In order to demodulate ${x_2}$, we also need to re-modulate the demodulated signal to get ${\hat{x}_1}$ which is the encrypted signal. Here, the channel response ${h_1}$ can be obtained by comparing the training sequence. Through subtracting ${\hat{x}_1} \times {h_1}$ from the received signal y, the second signal can be deduced. The flow chart of demodulation is shown in Fig. 8.
2.3 Encryption model analysis
The initial value of the 6D chaos model $a,b,c,d,e,f,g,x,y,z,u,v,w$ are set to 10, 55, 8/3, 0.6, −3, 1, 3,0.1, 0.6, 0.2, 0.02, 1 and 0.5, respectively. Select 0.001 as the simulation step size. Under the circumstances, the system has three positive Lyapunov exponents (LE), which further proves it is a hyperchaotic system. Table 2 shows the comparison of the LE between the 6D chaos model and traditional low-dimensional chaotic systems.
Table 2. LE comparison of different chaotic systems
The phase diagram of this model is depicted in Fig. 9 and the corresponded time-domain waveform diagram is shown in Fig. 10. As we can see, the chaotic sequence generated by the 6D fractional hyperchaotic model shows no periodicity.
Fig. 9. Phase diagram of the 6D fractional hyperchaotic model: (a) $x - y$ plane, (b) $x - y - z$ plane, (c) $x - z$ plane, (d) $y - u$ plane, (e) $x - v$ plane, (f) $z - w$ plane, (g) $u - v$ plane, (h)$v - w$ plane.
Fig. 10. The time-domain waveform diagram: (a) state diagram of x changing with time t, (b) state diagram of y changing with time t, (c) state diagram of z changing with time t, (d) state diagram of u changing with time t, (e) state diagram of v changing with time t, (f) state diagram of w changing with time t.
For the verify the key space of this chaotic system, we transform the initial value of $(x,y,z,u,v,w) $ from (0.10000000000001, 0.60000000000001, 0.20000000000001, 0.02000000000001, 1.0000000000 0001, 0.50000000000001) to (0.10000000000002, 0.60000000000002, 0.20000000000002, 0.02000000000002, 1.0000000000 0002, 0.50000000000002) respectively. Then extract the 14th digit for comparison. n represents the ${n^{th}}$ number of the chaotic sequence, (1, 2, 3, 4, 5, 6) on the dimension corresponds to 6 dimensions respectively, and $x(n)$ represents the value of the 14th digit after the decimal point. Figure 11 shows the values of points from 902 to 990. There are two color lines in each dimension, taking dimension x as an example, when $x = 0.10000000000001$, the value of the 14th digit after the decimal point on the 901 to 910 position of the obtained chaotic sequence is (6, 0, 6, 8, 9, 3, 7, 0, 5,1), when $x = 0.10000000000002$, they are (0, 4, 0, 2, 2, 6, 0, 3, 9, 4) respectively, as shown in the figure, the values of the two color lines at the same position are significantly different, which proves that even if the initial value changes slightly $1 \times {10^{ - 14}}$, the generated chaotic sequences are different. Therefore, the chaotic system is highly sensitive to the initial value, and the parameter values are similar. As a result, the key space of this paper is at least ${10^{182}}$.
In Table 3, it can be seen that the key space of our model is superior to the traditional and the recent proposed encryption model [24,25,26].
Table 3. The Key Space Comparison
Figure 12(a) and (b) show the diagrams of auto-correlation and cross-correlation function diagrams of the chaotic sequence X generated by the same and different initial values, respectively. When there is no lag-off, the diagram of correlation function shows a spike whose value is nearly 1 while the auto-correlation coefficient is close to 0 at other points. Besides, the cross-correlation function has no shock pulse and jitters around 0. The proof of strong auto-correlation and weak cross-correlation demonstrates that the chaotic sequence displays good randomness.
Fig. 12. Auto-correlation and cross-correlation of $X$: (a) auto-correlation function of the chaotic sequence X generated by the same initial values, (b) cross-correlation of the chaotic sequence X generated by different initial values.
3. Experimental setup and results
The experimental setup of the proposed SPDM-SDM-PON has been shown in Fig. 13. Figure 14 is the photo of the experimental setup. The system we propose superimposes two users in one fiber core through NOMA. Using the space division multiplexing dimension of the seven-core fiber, the number of PON access users can be increased by 14 times. It is compatible with other DSP technologies in the existing PON architecture and can support the expansion of the current 10/50G-PON and future 100G-PON. The target system is the current transmission standards. On the OLT side, the encrypted power domain multiplexed signal is generated offline through MATLAB. The subcarriers of the OFDM signals of the two users ${x_1}$ and ${x_2}$ in each core are both 216 and the IFFT transform points are 1024, respectively. M is 216 and N is 120. The length of CP is 1/4 IFFT transformation points. The total bit rate is tantamount to the expression of subcarrier number × 2 (bits/symbol) × 2 (channels) × AWG sampling rate × number of cores / (IFFT size + CP size) = 216 × 2 × 2 × 10 × 7 / (1024 + 256) = 47.25 Gb/s. Due to the 7% FEC overhead, the transmission net data rate of the signal is 47.25 / 1.07 = 44.16 Gb/s. The symbol rate = bit rate/${\log _2}4$. The bandwidth of both the QPSK OFDM and 16QAM OFDM signal is 10 × 216/1024 = 2.11 GHz. The 2.11 GHz bandwidth are occupied by 216 subcarriers, and the subcarrier spacing is 2.11 GHz/216 = 9.77 MHz. Digital-to-analog conversion is conducted by an arbitrary waveform generator (AWG, TekAWG70002A) with a sampling rate of 25 GSa/s. After linear amplification by an electrical amplifier (EA), the analog signals are fed into the intensity modulator. An external cavity laser (ECL) with the output power of 14.5 dBm and the linewidth smaller than 100 kHz is adopted to supply the continuous wave light. The launch power of transmitter is 2dBm. Before being divided into 7 channels and decorrelated by the delay line, the modulated optical signals are amplified by the erbium-doped fiber amplifier (EDFA) with 5.5 dB noise figure. These signals are coupled by a fan-in device and transmit in a 2 km weakly coupled seven-core optical fiber. After measurement, it is found that the average insertion loss of each core in the fiber is about 1.5 dB, and the crosstalk between the adjacent cores is about −50 dB. The seven-core fiber is demultiplexed by a fan-out device. For the ONU, a variable optical attenuator (VOA) is utilized to keep the received power suitable for the bit error rate (BER) measurement. The received optical signals are detected by a photodiode (PD) with 3 dB bandwidth of 40 GHz, which are then analyzed by a 50 GSa/s mixed-signal oscilloscope (MSO, TekMSO 73304DX). Finally, the offline digital signal processing is applied to recover the transmitted signals.
Fig. 13. Experimental setup of the QPSK OFDM-based dual-channel SPDM-SDM-PON. (DSP: digital signal processing; AWG: arbitrary waveform generator; EA: electrical amplifier; MZM: Mach-Zehnder modulator; EDFA: erbium-doped fiber amplifier; VOA: variable optical attenuator; PD: photodiode; MSO: mixed signal oscilloscope).
The PDRs of ${x_1}$ and ${x_2}$ could affect the performance of signals significantly. Figure 15 shows the BER line chart from core 1 to core 7 under different PDR conditions at the received optical power (ROP) of −13 dBm. It can be seen that the performance of ${x_1}$ continues to improve with the increase of PDR, but the BER of ${x_2}$ shows a trend of first decreasing and then increasing. The ${x_2}$ channel performs best in 7 core fibers when PDR = 4, indicating that the best channel performance has nothing to do with the fiber core.
Fig. 15. BER of ${x_1}$ and ${x_2}$ under different PDRs at the ROP of −13 dBm for seven-core fiber: (a)core 1, (b)core 2, (c) core 3, (d) core 4, (e) core 5, (f) core 6, (g) core 7.
In this paper, the PDR is fixed at 4. As shown in Fig. 16, the performances of OFDM signal and SPDM under the same conditions are compared because they have the same spectrum efficiency. It is observed that the ${x_1}$ of SPDM performs best while the performance of OFDM lies between ${x_1}$ and ${x_2}$. ${x_2}\_{x_1}$ represents the ${x_1}$ signal obtained by the user of ${x_2}$ without the correct key. In this case, the BER is about 0.5. User 2 cannot get any useful information, and the information security of user 1 is guaranteed. The information of user 2 is transmitted to user 1 as noise, leading to a significant improvement of user 2’s security. Take the performance of core-4 as an example for analysis, 16 QAM-OFDM signals reach hard-decision forward error correction (HD-FEC) limit of 3.8 × 10−3 when received optical power (ROP) of −14.0 dBm. However, ${x_1}$ can achieve a receiver sensitivity improvement of 0.2 dB compared to 16QAM-OFDM signals, of which the corresponding ROP of ${x_2}$ is −10.7 dBm. The signal demodulated by ${x_1}$ can be re-modulated to obtain ${x_2}$. Due to the channel interference, ${x_2}$ shows higher BER than ${x_1}$ under the same ROP condition. Additionally, the performance of ${x_2}$ can be adjusted at any time according to the requirements users, improving the flexibility of the access network greatly.
Fig. 16. Measured BER curves of two users for seven-core: (a)core 1, (b)core 2, (c) core 3, (d) core 4, (e) core 5, (f) core 6, (g) core 7.
The performances of $SPDM - {x_1}$ and $SPDM - {x_2}$ in the 7 cores are shown in Fig. 17. It can be seen that the performance difference between the two SPDM signals in the 7 cores is not obvious. At the HD-FEC limit, for ${x_1}$, the difference between the best-performing core and the worst core is only 0.6 dB, and the difference for ${x_2}$ is 0.9 dB, which also shows the good isolation and stability of the seven-core fiber from the side. In addition, power budget defines the amount of optical power available for successful transmitting signal over a distance of optical fiber. From Fig. 17, ${x_2}$ reaches HD-FEC limit of 3.8 × 10−3 when received optical power is −11.3 dBm. The modulated optical signal transmit power is 2 dBm, so the power budget of this link is $2 - ( - 11.3) = 13.3\textrm{ }dB$.
4. Conclusion
In this paper, a novel SPDM-SDM-PON is proposed and the corresponding transmission performances are demonstrated experimentally. The 6D hyperchaotic system is used to generate a chaotic sequence to perform amino acid and extended zigzag encryption on the transmitted signals. The key space can reach 10182. The BER maintains at around 0.5 with the change of parameters of chaotic system. The encryption method is applied to the digital domain power division multiplexing in the non-orthogonal multiple access, where each user can obtain the transmitted signals through the SIC algorithm. The performance of the 47.25 Gb/s SPDM signal verified in a 2 km seven-core optical fiber. At the HD-FEC limit, the output difference between the best and the worst cores is 0.85 dB. The performance of ${x_1}$ is better than that of the 16 QAM signal and ${x_2}$. The spectral efficiency of ${x_1}$ and ${x_2}$ equals to that of 16 QAM. Multiple signals are superimposed in the digital power domain, which improves the spectrum efficiency and the number of user accesses effectively. PDR can also be adjusted in real time based on the demands of users to improve the utilization efficiency of the network. In view of the above advantages, the proposed scheme deserves more extensive attentions and researches.
Funding
National Key Research and Development Program of China (2018YFB1800901); National Natural Science Foundation of China (61720106015, 61727817, 61775098, 61822507, 61835005, 61875248, 61935005, 61935011, 61975084, 62035018, U2001601); Open Fund of IPOC (BUPT); Opened Fund of the State Key Laboratory of Integrated Optoelectronics (IOSKL2020KF17); Jiangsu team of innovation and entrepreneurship; 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. D. Veen and V. Houtsma, “Strategies for economical next-generation 50G and 100G passive optical networks [Invited],” J. Opt. Commun. Netw. 12(1), A95–A103 (2020). [CrossRef]
2. K. Zhong, X. Zhou, J. Huo, C. Yu, C. Lu, and A. Lau, “Digital signal processing for short-reach optical communications: A Review of Current Technologies and Future Trends,” J. Lightwave Technol. 36(2), 377–400 (2018). [CrossRef]
3. V. Houtsma and D. Veen, “A study of options for high-speed TDM-PON beyond 10G,” J. Lightwave Technol. 35(4), 1059–1066 (2017). [CrossRef]
4. T. Kanai, M. Fujiwara, R. Igarashi, N. Iiyama, R. Koma, J. Kani, and T. Yoshida, “Symmetric 10 Gbit/s 40-km reach DSP-based TDM-PON with a power budget over 50 dB,” Opt. Express 29(11), 17499–17509 (2021). [CrossRef]
5. G. Chang, A. Chowdhury, Z. Jia, H. Chien, M. Huang, J. Yu, and G. Ellinas, “Key Technologies of WDM-PON for Future Converged Optical Broadband Access Networks [Invited],” J. Opt. Commun. Netw. 1(4), C35–C50 (2009). [CrossRef]
6. C. Wagner, M. Eiselt, M. Lawin, S. Zou, K. Grobe, J. Olmos, and I. Monroy, “Impairment Analysis of WDM-PON Based on Low-Cost Tunable Lasers,” J. Lightwave Technol. 34(22), 5300–5307 (2016). [CrossRef]
7. H. Beyranvand, S. Nezamalhosseini, J. Salehi, and F. Marvasti, “Performance analysis of equal-energy two-level OCDMA system using generalized optical orthogonal codes,” J. Lightwave Technol. 31(10), 1573–1584 (2013). [CrossRef]
8. A. Skidin, O. Sidelnikov, and M. Fedoruk, “Statistical properties of nonlinear distortion of a polarization-multiplexed OFDM signal in long-haul fiber links,” Opt. Lett. 45(19), 5550–5553 (2020). [CrossRef]
9. H. Hu and S. Radic, “Hybrid OFDM receiver assisted by a variable frequency comb,” Opt. Express 28(4), 5658–5668 (2020). [CrossRef]
10. S. Jansen, I. Morita, T. W. Schenk, N. Takeda, and H. Tanaka, “Coherent optical 25.8-Gb/s OFDM transmission over 4160-km SSMF,” J. Lightwave Technol. 26(1), 6–15 (2008). [CrossRef]
11. B. Baeuerle, C. Hoessbacher, W. Heni, Y. Fedoryshyn, U. Koch, A. Josten, D. Elder, L. Dalton, and J. Leuthold, “100 GBd IM/DD transmission over 14 km SMF in the C-band enabled by a plasmonic SSB MZM,” Opt. Express 28(6), 8601–8608 (2020). [CrossRef]
12. T. Wiegart, F. Ros, M. Yankov, F. Steiner, S. Gaiarin, and R. Wesel, “Probabilistically shaped 4-PAM for short-reach IM/DD links with a peak power constraint,” J. Lightwave Technol. 39(2), 400–405 (2021). [CrossRef]
13. N. André, H. Louchet, V. Filsinger, E. Hansen, and A. Richter, “OFDM and PAM comparison using a high baudrate low resolution IM/DD interface for 400G Ethernet access,” Opt. Express 24(11), 11926–11931 (2016). [CrossRef]
14. S. Jung, S. Jung, and S. Han, “I/Q Channel Separated Baseband OFDM Optical Transmission Using Orthogonal Polarizations in IM/DD System,” J. Lightwave Technol. 32(13), 2392–2398 (2014). [CrossRef]
15. Z. Feng, M. Tang, X. Guan, C. C. Chan, Q. Wu, R. Wang, R. Lin, S. Fu, L. Deng, and D. Liu, “Digital domain power division multiplexing ddo-ofdm transmission with successive interference cancellation,” in 2016 Conference on Lasers and Electro-Optics (CLEO), (2016), 1–2.
16. K. Suzuoki, D. Hisano, K. Mishina, K. Maruta, and A. Maruta, “Demonstration of asynchronous successive interference canceller for uplink NOMA-PON,” in 2011 Optical Fiber Communication Conference (OFC), (2021), paper W6A.40.
17. X. Huang, F. Yang, C. Pan, and J. Song, “Flexible NOMA-based NOHO-OFDM scheme for visible light communication with iterative interference cancellation,” Opt. Express 29(4), 5645–5657 (2021). [CrossRef]
18. Z. Feng, M. Tang, X. Guan, C. Chan, Q. Wu, X. Chen, R. Wang, R. Lin, S. Fu, L. Deng, and D. Liu, “Spectrally overlaid DDO-OFDM transmission enabled by optical power division multiplexing,” in 15th International Conference on Optical Communications and Networks (ICOCN), (2016), 1–3.
19. Q. Wu, Z. Feng, M. Tang, X. Li, M. Luo, H. Zhou, S. Fu, and D. Liu, “Digital domain power division multiplexed dual polarization coherent optical ofdm transmission,” Sci. Rep. 8(1), 15827 (2018). [CrossRef]
20. S. Swain, S. Bhashyam, R. Koilpillai, and D. Venkitesh, “Digital power division multiplexed DD-OFDM using fundamental mode transmission in few-mode fiber,” Opt. Express 28(12), 17809–17819 (2020). [CrossRef]
21. S. Sarmiento, J. Mendinueta, J. Altabás, S. Spadaro, S. Shinada, H. Furukawa, J. Olmos, J. Lázaro, and N. Wada, “Split-enabled 490 Gb/s optical interconnect with direct detection NOMA-CAP and 7-core multi-core fibre,” in 2019 European Conference on Optical Communication (ECOC), (2019).
22. S. Sarmiento, J. Mendinueta, J. Altabás, S. Spadaro, S. Shinada, H. Furukawa, J. Olmos, J. Lázaro, and N. Wada, “Split-enabled 350-630 Gb/s optical interconnect with direct detection NOMA-CAP and 7-core multi-core fiber,” Opt. Commun. 463, 125321 (2020). [CrossRef]
23. X. Song, B. Liu, H. Zhang, R. Ullah, Y. Mao, J. Ren, S. Chen, J. Zhang, J. Zhao, S. Han, X. Liu, D. Zhao, and X. Xin, “Security-enhanced OFDM-PON with two-level coordinated encryption strategy at the bit-level and symbol-level,” Opt. Express 28(23), 35061–35073 (2020). [CrossRef]
24. W. Zhang, C. Zhang, C. Chen, and K. Qiu, “Experimental demonstration of security-enhanced OFDMA-PON using chaotic constellation transformation and pilot-aided secure key agreement,” J. Lightwave Technol. 35(9), 1524–1530 (2017). [CrossRef]
25. Z. Wang, Y. Xiao, S. Wang, Y. Yan, B. Wang, Y. Chen, Z. Zhou, J. He, and L. Yang, “Probabilistic shaping based constellation encryption for physical layer security in OFDM RoF system,” Opt. Express 29(12), 17890–17901 (2021). [CrossRef]
26. L. Liu, X. Tang, X. Jiang, Z. Xu, F. Li, Z. Li, H. Huang, P. Ni, L. Chen, L. Xi, and X. Zhang, “Physical layer encryption scheme based on cellular automata and DNA encoding by hyper-chaos in a CO-OFDM system,” Opt. Express 29(12), 18976–18987 (2021). [CrossRef]
