Open Access
Add to BibSonomy
Abstract
A physical layer key distribution scheme based on chaotic encryption and signal synchronization is proposed in this paper, which can achieve secure key distribution and enhance the security of an orthogonal frequency division multiplexing based passive optical network (OFDM-PON). The key is embedded into the synchronization header and then encrypted by using chaos. The receiver needs to utilize the correct chaotic parameters to successfully decrypt the synchronization information and extract the key. An experiment is conducted to verify the availability of this method by setting key sequences of various length over different transmission distances. The signals of 35.29 Gb/s are successfully transmitted over 5 km, 15 km and 25 km standard single-mode fiber (SSMF), respectively. It is proved that the proposed scheme is feasible and compatible with the traditional encryption algorithms, and it has almost no effect on the synchronization performance, which can then distribute keys with the sending signals without occupying additional channel resources and enhance the security performance of OFDM-PON simultaneously.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Orthogonal frequency division multiplexing based passive optical network (OFDM-PON) is considered as a powerful technology for fiber access networks because of its robustness against fiber dispersion, high spectrum efficiency, ability to execute multiple services and dynamic bandwidth allocation [1,2]. However, with the rapid development of the Internet, plenty of private data may be leaked when transmitted on the Internet. Since a downlink between optical line terminal (OLT) and optical network unit (ONU) in passive optical network (PON) system adopts broadcast communication, the transmission information is easily to intercept by eavesdropping. Therefore, the security protection technology for optical access networks has been studied extensively in recent years [3].
Chaos has the characteristics of ergodicity, pseudo randomness and initial value sensitivity, which is good for secure communication [4]. Nowadays, as an efficient and fast physical layer data transmission and encryption method, the physical layer chaotic communication has then become one main mean to solve the security problem of optical access network. Until now, the proposed schemes using chaos encryption can be classified according to different type of targets. For the data encryption, the sequence generated by chaotic map was formed the step length and direction of Brownian motion [4]. By using the randomness of Brownian motion, they replaced and interleaved OFDM symbols to improve the security of coherent optical OFDM-PON system. At the bit-level encryption, the DNA coding was introduced into bit sequence encryption [5], which has the advantages of high parallelism, huge storage space and ultra-low power consumption. In addition, the DNA extension code was proposed, this coding rule was expanded from 8 to 384, which improved the complexity and randomness of DNA encryption sequences and made the effect better [6]. After applying odd-even cross-bit DNA encoding and the spiral scrambling, this DNA coding is more mature [7]. Recently, the cellular automata (CA) appeared in a novel DNA encryption scheme, this CA generated a scrambling sequence, expanded the randomness of the system and further increased the security [8]. The extensible works to improve the security were proposed, which used DNA coding to reconstruct chaotic sequences for multi-level encryption, and then introduced RNA codebook to expand the diversity of DNA coding [9,10].
For the transmitted signal, a noise-based encryption and channel / phase estimation method was proposed [11]. The random noise data was replaced by the transmission sequence in the preamble and embedded into a variable number of randomly selected pilot subcarriers to hide the signal recovery information. An encryption algorithm based on chaotic constellation mapping and probabilistic shaping was proposed [12]. This constellation of encrypted data was evenly distributed after chaotic diffusion in the secondary amplitude modulation map to eliminate the statistical characteristics. The probabilistic shaping (PS) was also used to improve the transmission performance of OFDM signal. A new constellation encryption scheme based on PS encrypted chaotic sequences and hash values at bit level was used to enhance transmission performance and security [13]. Interleaved and scaled according to the pre-shared key through time-domain pre-coding, and randomly changed the size of the cyclic prefix according to the key, it would not only lead to the change of the size of the OFDM symbol, but also ICI and ISI interference at illegal party [14]. The improved Lorentz chaotic map has a good feature, a feedback factor was added on the basis of the traditional Lorentz map to make the chaotic perform better [15]. By adding the division of candidate signals on the basis of the traditional SLM, it achieved better PAPR, BER, fiber nonlinear tolerance and security performance [16]. The method adopted the IQ encryption to realize quadrature amplitude modulation (QAM) symbols disturbance, using key sequence to encode real part and imaginary part respectively [17]. It was designed to quantify the characteristics between ONUs as fingerprints for identity authentication, this fingerprint was embedded into the pilot. The signal based on wavelet transform was then used to extract the characteristic matrices, they were classified and identified by a convolutional neural network [18]. The chaotic matrix for encryption was applied, which made the results more random and unpredictable [19,20]. Although these encryption schemes can provide high security, they still lack consideration in protecting all information. The signal synchronization information also needs to be encrypted to prevent any illegal ONU deciphering.
The previous encryption schemes used static pre-shared keys and rarely involved dynamic key management. However, the security of keys is the top priority for the physical layer security of optical access network. According to Shannon's one-time encryption theory, how to dynamically manage keys is a problem worthy of research. Many scholars have made some contributions. For example, the key index embeded in the pilot to distribute, when estimating the channel, they can recover the index and select the key from the locally statically stored key group [21]. Dynamic key generation according to channel state was proposed, its key was generated and distributed periodically by combining random channel phase response and random signal [22]. A new dimension for the generation of random key was executed according to the information of OFDM subchannel [23]. The key distribution scheme was proposed, which measured bit error rate (BER) of a loopback signal, quantified and encoded the results into a consistent key [24]. In addition, the scheme through transferring optical signal with fast fluctuating states of polarization to complete key generation and distribution was also proposed [25]. A pioneering work used the redundancy of training sequence to complete the key distribution, and completed the dynamic key distribution by taking the chaotic sequence embedded in the key as a training sequence was proposed [26]. With the wide application of neural network, there have been some studies to enhance security of time synchronization, encryption process and key management [27–29]. However, these schemes would affect the synchronization performance and require the additional channel resources.
In this paper, we propose a scheme combining chaotic encryption, signal synchronization and key distribution to achieve secure key distribution and enhance the security of OFDM-PON. The key is embedded into the synchronization header sequence (SHS) according to the specific mapping rules. After DNA coding and encryption, it is transmitted with the encrypted data. The receiver firstly creates the same DNA decoding sequence and decrypts the received sequence, then finds out the position of the synchronization header sequence according to the employed algorithm by calculating correlation. Finally the synchronization process is completed and the key is extracted, the system security and the ability to resist attackers can be then improved effectively.
2. Principle
The principle of the proposed scheme is shown in Fig. 1. At the OLT, a pseudo-random binary sequence (PRBS) as the downstream data is input. After serial-to-parallel (S/P) and 16-QAM modulation, the data is converted from a bit sequence to a symbol matrix, and encrypted by the encryption algorithm. We use Brownian motion [4] to encrypt data. After inverse fast Fourier transform (IFFT), the SHS is encoded, and the selected key from key set is embedded in it. The DNA encoding is performed to encrypt the SHS to ensure that the key is secure. Finally, after adding cyclic prefix (CP) and executing parallel-to-serial (P/S), the signal is sent to the ONU.
According to the classical scheme developed by Schmidl and Cox, two identical symbols are required for the synchronization [30]. Therefore, in the SHS, the last symbol (LS) is equal to the first symbol (FS) so that we only need to structure one symbol. To insert M-bit keys in the N-bit symbol, it is necessary to fill the vacancy of the symbol. In fact, more significant than creating rules is how to share them. In our scheme, since the security is mainly based on chaos, we embed the keys by inserting in a regular PRBS, which can be expressed as:
At the ONU, the pre-shared key is used to generate the same chaotic sequences {X} and {Y}, and the decoding sequence {Y’} can be restructured. For every bit, we intercept the following two symbol lengths sequence as FS’ and LS’, operate it with {Y’} because of symmetry and decode the result, then calculate the correlation by every bit until find the SHS, which is as follow:
3. Experiment setup
The experiment setup of the proposed scheme is shown in Fig. 2. A PRBS with the length of 9.6×104 is input into the OLT, after S/P and QAM modulation, the bit stream is converted into 120×200 QAM symbol matrix. The 1024-bit SHS with M-bit key embedded is added at the beginning of each frame. Then we generate the encrypted OFDM signal through using chaos offline by MATLAB.
Fig. 2. The experiment setup. AWG: arbitrary waveform generator; LD: laser diode; MZM: Mach-Zehnder modulator; SSMF: standard single-mode fiber; PSC: passive splitter couple; TOA: tunable optical attenuator; PD: photodetector; DSO: digital storage oscilloscope; OLT: optical line terminal; ONU: optical network unit.
The length of IFFT is 256. The format with Hermite symmetric structure is applied to ensure the generation of real values. 120 subcarriers and the other 120 subcarriers carry the encrypted data and the corresponding complex conjugate data, respectively. In order to reduce the inter-symbol interference (ISI), a CP of 1/16 is added in the data sequence. After the signal is loaded to the arbitrary waveform generator (AWG, Keysight M8196A) with the sampling rate of 20 GSa/s, the digital-to-analog conversion (DAC) is completed. The optical source is a tunable laser diode (LD) whose wavelength is set to 1552 nm. A Mach-Zehnder modulator (MZM) with a bandwidth of 10 GHz is used to modulate the encrypted signal onto the optical carrier. In order to explore the influence of transmission distance on the experiment, the optical signal is transmitted over the 5-km, 15-km and 25-km standard single-mode fiber (SSMF), respectively. At the ONU, the received optical power is controlled by the tunable optical attenuator (TOA), then a photodetector (PD) with a bandwidth of 10 GHz is used to detect the optical signal. The signal is recorded using a digital storage oscilloscope (DSO, Keysight Z594A) with a sampling rate of 40 GSa/s as an analog-to-digital converter (ADC). Finally, MATLAB conducts signal demodulation and decryption offline.
4. Results and discussions
The accuracy of data decryption and key extraction depends on the correct signal synchronization, and the signal synchronization depends on the similarity of the FS and the LS. Because the chaotic sequence is unique and the structure of two symbols is the same, even if there are some data segments with the similar structure, they will be inconsistent after chaotic decryption. The SHS can only be decrypted in the accurate location using the correct initial key, so the problem of peak platform in [30] can be ameliorated in this scheme. The synchronization performs well. Figure 3(a) shows the synchronization under different key numbers of legal ONUs, the peak appears higher compared with the flatness of other parts. Therefore, the location of SHS can be found, and then the keys can be extracted correctly and the data can be decrypted. With the increasing of the length of embedded key sequence, the peak value decreases slightly, but it isn’t almost impact on the synchronization performance. This cost can be ignored for the increasing of the number of transmission keys. Moreover, the key transmission makes use of the redundancy of the synchronization head, which isn’t affect the data segment. Under the normal circumstances, the synchronization head behaves like noise. Even if any illegal ONU eavesdrops the signal, it is difficult to find out the location of SHS and complete the synchronization signal. Figure 3(b) shows the synchronization of illegal ONU. The security of the signal is enhanced.
Fig. 3. The synchronization (a) the extracted keys when 0-bit, 64-bit, 128-bit and 256-bit keys are embedded of legal ONUs; and (b) illegal ONU.
The BER results are shown in Fig. 4. Figure 4(a) is the comparison under the effect of different key numbers on BER. Because the synchronization header in the traditional scheme is a known sequence, the content is meaningless. However, after carrying the key information, it is important that the synchronization header is restored correctly. For the legal ONUs, when M = 0, no key is embedded in SHS as an initial signal. After the key is embedded, these BER curves are basically consistent, and at a sufficiently high received optical power, the key can be extracted without error. For any illegal ONU, the BER is always about 0.5, which indicates that illegal ONU cannot obtain useful information. Figure 4(b) is the comparison under the effect of different transmission distance on BER. Due to the limited number of experiments, the accidental situations may occur, but in fact, the BER will change due to the change of transmission distance only when the received optical power is relatively low.
Fig. 4. The BER vs. the received optical power. (a) BER curves of the extracted keys when 0-bit (Initial), 64-bit, 128-bit and 256-bit keys are embedded. (b) BER results under different transmission distance.
In terms of data encryption, we use the classic Brownian Motion Encryption [4]. The initial signal only encrypts the data. By comparing the BER curves of two legal ONUs and an illegal ONU under different received optical power, as shown in Fig. 5, the BER curves are very close at the same transmission distance, which means that the embedded key number has very little effect on BER performance. When the received optical power is −15 dBm, the received constellation of illegal ONU is shown in the Fig. 5(a). When the received optical power are −19 dBm and −15 dBm, the constellation diagrams of the encrypted signal with 256-bit keys are shown in the Fig. 5(b) and Fig. 5(c), respectively. According to Fig. 4(a), the key can be extracted without error when the BER of the transmitted data is lower than the FEC limit, and if the channel is not in a particularly bad condition, the key can still be extracted well. For different transmission distance, we can conclude that as the transmission distance decreases, the signal loss and the requirements for channel condition also decrease. However, it is only different at higher received optical power, and it is almost not reflected at lower received optical power. For any illegal ONU, although it can receive the correct constellation by eavesdropping, whether the key is inserted or not, and whether the transmission distance is long or short, the BER is always above −0.5, which means that any illegal ONU is unable to decrypt the correct information.
Fig. 5. The BER curves under different key numbers or different transmission distances. Insert results of (a) The received constellation of illegal ONU, (b) The received constellation of legal ONU in −19 dBm, (c) The received constellation of legal ONU in −15 dBm.
Ideally, for the comprehensive performance of key distribution, when the scheme has the characteristics of high number of carrying keys, no occupation of additional channel resources and low computational complexity, there is no doubt that it is a potential scheme. We assume that the number of the transmitted keys is N, and the chaotic sequences are also generated. The comparison analysis to several existing typical key distribution schemes is listed in Table 1. The others stand for operations other than addition and multiplication such as mod, abs and extract, and each operation will increase N. Compared with the methods of [21] and [24], the proposed scheme does not need to consume additional channel resources and use the redundancy of channels to transmit additional information. Compared with the strategy of [26], the proposed scheme has not any impact on synchronization, and even better protection of synchronization information. Although the computational complexity of this scheme is improved compared with other schemes, it becomes serious when dealing with large numbers of keys for all schemes.
Table 1. The Comparison Analysis
5. Conclusion
To achieve secure key distribution and enhance the security of OFDM-PON, we united chaotic encryption, signal synchronization and key distribution in this scheme and analyzed it comprehensively. By taking advantage of the redundancy of synchronization head, without occupying additional channel resources, the secure distribution of key and the hiding of synchronization information were realized. The key and the synchronization information were hidden in noise, and the security was enhanced by the employed encryption algorithm, which can deal with exhaustion attacks. The experimental results indicated that the number of embedded keys has no significant effect on data transmission, and the speed of operation mainly depends on the length of data segment. The extracted key can be used as the specific parameters required for the next transmission, which increased the difficulty of cracking from illegal ONU and improved the security of this system. In a long-distance transmission, we need to pay attention to the requirements of channel conditions. Moreover, because of the disorder of chaotic sequence, this scheme reduced the risk of the peak platform occurring and improved the stability of synchronization. As an efficient and dynamic key update mechanism, this scheme has good robustness to channel noise, and can be applied to most data encryption schemes in secure OFDM-PON.
Funding
National Natural Science Foundation of China (62071088); National Key Research and Development Program of China (2018YFB1801302); Project for Innovation Team of Guangdong University (2018KCXTD033).
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. S. Y. Jung, C. H. Kim, S. M. Jung, and S. K. Han, “Optical pulse division multiplexing-based OBI reduction for single wavelength uplink multiple access in IM/DD OFDMA-PON,” Opt. Express 24(25), 29198–29208 (2016). [CrossRef]
2. F. Carvalho and A. Cartaxo, “Multisignal OFDM-based hybrid optical wireless WDM LR-PON with colorless ONU,” IEEE Photonics Technol. Lett. 27(11), 1193–1196 (2015). [CrossRef]
3. A. Teixeira, A. Vieira, J. Andrade, A. Quinta, M. Lima, R. Nogueira, P. Andre, and G. Tosi Beleffi, “Security issues in optical networks physical layer,” 2008 10th Anniversary International Conference on Transparent Optical Networks, 123–126(2008).
4. W. Zhang, C. Zhang, C. Chen, H. Zhang, and K. Qiu, “Brownian motion encryption for physical-layer security improvement in CO-OFDM-PON,” IEEE Photonics Technol. Lett. 29(12), 1023–1026 (2017). [CrossRef]
5. C. Zhang, W. Zhang, C. Chen, X. He, and K. Qiu, “Physical-enhanced secure strategy for OFDMA-PON using chaos and deoxyribonucleic acid encoding,” J. Lightwave Technol. 36(9), 1706–1712 (2018). [CrossRef]
6. T. W. Wu, C. F. Zhang, H. Huan, Z. Zhang, H. H. Wei, H. P. Wen, and K. Qiu, “Security improvement for OFDM-PON via DNA Extension Code and Chaotic Systems,” IEEE Access 8, 75119–75126 (2020). [CrossRef]
7. Y. Xiao, Y. Chen, C. Long, J. Shi, J. Ma, and J. He, “A Novel Hybrid Secure Method Based on DNA Encoding Encryption and Spiral Scramblieng in chaotic OFDM-PON,” IEEE Photonics J. 12(3), 1–15 (2020). [CrossRef]
8. 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]
9. M. Cui, C. Zhang, Y. Chen, Z. Zhang, T. Wu, and H. Wen, “Multilayer dynamic encryption for security OFDM-PON using DNA-reconstructed chaotic sequences under cryptanalysis,” IEEE Access 9, 18052–18060 (2021). [CrossRef]
10. M. Cui, Y. Chen, C. Zhang, X. Liang, T. Wu, S. Liu, H. Wen, and K. Qiu, “Chaotic RNA and DNA for security OFDM-WDM-PON and dynamic key agreement,” Opt. Express 29(16), 25552–25569 (2021). [CrossRef]
11. W. Jin, C. Zhang, and W. Yuan, “Physically secured orthogonal frequency division multiplexing–passive optical network employing noise-based encryption and signal recovery process,” Opt. Eng. 55(2), 026103 (2016). [CrossRef]
12. Z. Zhang, Y. F. Luo, C. F. Zhang, X. S. Liang, M. W. Cui, and K. Qiu, “Constellation shaping chaotic encryption scheme with controllable statistical distribution for OFDM-PON,” J. Lightwave Technol. 40(1), 14–23 (2022). [CrossRef]
13. 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]
14. L. Samara, A. O. Alabbasi, A. Gouissem, R. Hamila, and N. Al-Dhahir, “A novel OFDM waveform with enhanced physical layer security,” IEEE Commun. Lett. 25(2), 387–391 (2021). [CrossRef]
15. J. Shen, B. Liu, Y. Mao, R. Ullah, J. Ren, J. Zhao, and S. Chen, “Enhancing the reliability and security of OFDM-PON using modified Lorenz chaos based on the linear properties of FFT,” J. Lightwave Technol. 39(13), 4294–4299 (2021). [CrossRef]
16. Y. Wan, B. Liu, J. Ren, R. Ullah, Y. Mao, S. Chen, H. Zhang, S. Han, X. Song, T. Nan, and F. Tian, “PAPR-degraded secure OFDM-WDM-PON based on chaotic set-partitioned SLM,” IEEE Photonics Technol. Lett. 33(24), 1387–1390 (2021). [CrossRef]
17. W. Zhang, C. F. Zhang, W. Jin, C. Chen, N. Jiang, and K. Qiu, “Chaos coding-based QAM IQ-encryption for improved security in OFDMA-PON,” IEEE Photonics Technol. Lett. 26(19), 1964–1967 (2014). [CrossRef]
18. S. Li, M. Cheng, Y. Chen, C. Fan, L. Deng, M. Zhang, S. Fu, M. Tang, P. Shum, and D. Liu, “Enhancing the physical layer security of OFDM-PONs with hardware fingerprint authentication: a machine learning approach,” J. Lightwave Technol. 38(12), 3238–3245 (2020). [CrossRef]
19. Z. Hu and C. Chan, “A real-valued chaotic orthogonal matrix transform-based encryption for OFDM-PON,” IEEE Photonics Technol. Lett. 30(16), 1455–1458 (2018). [CrossRef]
20. Z. Hu, P. Song, and C.-K. Chan, “Chaotic non-orthogonal matrix-based encryption for secure OFDM-PONs,” IEEE Photonics Technol. Lett. 33(20), 1127–1130 (2021). [CrossRef]
21. W. Zhang, C. F. 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]
22. Y. Wu, Y. Yu, Y. Hu, Y. Sun, T. Wang, and Q. Zhang, “Channel-based dynamic key generation for physical layer security in OFDM-PON systems,” IEEE Photonics J. 13(2), 1–9 (2021). [CrossRef]
23. H. M. Furqan, J. M. Hamamreh, and H. Arslan, “New physical layer key generation dimensions: subcarrier indices/positions-based key generation,” IEEE Commun. Lett. 25(1), 59–63 (2021). [CrossRef]
24. X. Wang, J. Zhang, Y. Li, Y. Zhao, and X. Yang, “Secure key distribution system based on optical channel physical features,” IEEE Photonics J. 11(6), 1–11 (2019). [CrossRef]
25. L. Zhang, X. Huang, W. Hu, and X. Yang, “Point to multi-point physical-layer key generation and distribution in passive optical networks,” Opt. Lett. 46(13), 3223–3226 (2021). [CrossRef]
26. S. Li, M. Cheng, L. Deng, S. Fu, M. Zhang, M. Tang, P. Shum, and D. Liu, “Secure key distribution strategy in OFDM-PON by utilizing the redundancy of training symbol and digital chaos technique,” IEEE Photonics J. 10(2), 1–8 (2018). [CrossRef]
27. M. Bi, X. Zhuo, X. Fu, X. Yang, and W. Hu, “Cellular neural network encryption scheme for time synchronization and CPAs resistance in OFDM-PON,” IEEE Access 7, 57129–57137 (2019). [CrossRef]
28. Y. Zhou, M. Bi, X. Zhuo, Y. Lv, X. Yang, and W. Hu, “Physical layer dynamic key encryption in OFDM-PON system based on cellular neural network,” IEEE Photonics J. 13(2), 1–14 (2021). [CrossRef]
29. Y. Xiao, B. Wang, Z. Wang, Z. Zhou, and J. He, “One-time pad scheme based on polar code and OFDM for MMW-RoF system at W-band,” Opt. Express 30(3), 4412–4423 (2022). [CrossRef]
30. T. M. Schmidl and D. C. Cox, “Robust frequency and timing synchronization for OFDM,” IEEE Trans. Commun. 45(12), 1613–1621 (1997). [CrossRef]
31. N. J. Corron, R. M. Cooper, and J. N. Blakely, “Entropy rates of low-significance bits sampled from chaotic physical systems,” Phys. D 332, 34–40 (2016). [CrossRef]
