Open Access
Add to BibSonomy
Abstract
The number of base stations (BSs) for the fifth generation (5G) wireless network is substantially increased, as each coverage is greatly reduced. Therefore, both the miniaturization and the densification of BSs suffer from the challenges of electrical power supply and deployment cost. Here, we present an optically powered 5G fronthaul network, in support of the co-propagation of spatial-division-multiplexing (SDM) energy light and wavelength-division-multiplexing (WDM) 5G new radio (NR) signals over the weakly-coupled multicore fiber (WC-MCF). When the 60-W energy light at 1064.8-nm is equally distributed among the outer six cores, and the 9-Gbit/s 5G NR WDM signals are transmitted over the central core of 1-km WC seven-core fiber (WC-7CF), we can collect total 11.9-W electrical power at the remote node, for the purpose of optically powered small cells. Meanwhile, the error-vector magnitude (EVM) values of 1.5-Gbit/s 5G NR 64-level quadrature amplitude modulation orthogonal frequency division multiplexing (64QAM-OFDM) signals at the central frequency of 3.5 GHz fluctuate within a range of 0.3%∼0.39%, under a received electrical power of −25 dBm, for all six-wavelength channels. Six optically powered small cells are equipped with the characteristics of centralized management and flexible access-rate.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
The fifth-generation (5G) mobile network is intended to build a world of the Internet of everything [1], with typical characteristics of high-speed access, low latency, and massive connection [2]. In particular, the mobile access-rate is expected to reach 1-Gbit/s. In order to fulfill such a goal, an advanced radio access network (RAN) together with a cost-effective fiber optical fronthaul link is indispensable. Unlike conventional RAN where the baseband units (BBUs) and the radio units are placed together at the base station (BS), the centralized RAN (C-RAN) architecture migrates the BBUs to the central office (CO) with highly centralized and shared equipment, supporting the 5G wireless network and beyond [3]. Although the BBU is divided into the centralized unit (CU) and the distributed unit (DU) for the 5G application, the 5G C-RAN still has the characteristics of centralization and cooperation. The passive optical network (PON) is one of the most economical techniques to provide a broadband optical connection between the BBU pool and the BSs. Especially, the wavelength division multiplexing-PON (WDM-PON) is the outstanding candidate for the high-capacity low-latency 5G fronthaul network, in comparison with the time division multiplexing counterpart [4]. Different wavelength channels of WDM-PON can be assigned with different protocols, variable access-rates, and flexible modulation formats, for the ease of providing a high-speed fronthaul network, together with a substantial reduction of the fiber resource and the scheduled delay.
Furthermore, one of the prominent features of the 5G wireless network is the deployment of small cell, in order to provide excellent cellular coverage under the hot-spot outdoor environment [5,6]. In addition, massive small cells are vital to overcoming the explosive upsurge of mobile traffic cost-effectively for the telecom carrier. Generally, the type of small cells can be divided into femto-cell, pico-cell, and micro-cell, whose coverage range from tens of meters to a few hundred meters. Meanwhile, under the 5G scenario, the remote antenna unit (RAU) is suitable for the hot-spot indoor environment, whose power consumption is around 1-W [7]. Thus, the RAUs are useful for extending the indoor coverage area of the small cellular network [8]. In order to realize a wide coverage and high-speed access of the 5G C-RAN, small cells must cooperate with macro cells to form a heterogeneous network. Since the number of small cells is more than that of macro cells, the deployment and the management of electrical power supply become a challenging for the 5G BSs. Current schemes of electrical power supply for the BSs include the direct electrical power supply, the electrical power transfer supply, and the solar power supply. The scheme of direct electrical power supply can realize the power delivery from the power supply bureau to each small cell. However, in face of the high density of small cell deployment, the electrical power supply of small cells can be obtained from other units such as estate, instead of the direct connection with the power supply bureau. Such scheme is referred to the electrical power transfer, suffering from several disadvantages of low power quality, poor stability, and potential waste of resources. Therefore, both the BS construction cost and the implementation complexity of the fronthaul network become a headache, when copper wires are employed to feed the electrical power, as shown in Fig. 1. Moreover, as for the ultra-remote BSs where supplying electrical power is difficult, the solar power supply system, including solar panels, controllers, batteries, and inverters, becomes an alternative choice. However, the performance of the solar power supply system is constrained by the weather condition. Therefore, it is highly desirable to explore an efficient and centralized-management solution for the electrical power supply of 5G small cells.
With the reduced electrical power consumption of a single small cell, the power-over-fiber (PWoF) technique is promising to realize the electrical power supply for a large number of RAUs [9]. During past few years, optically powered RAU, based on the standard single-mode fiber (SSMF), the multi-mode fiber (MMF), the microstructure optical fiber (MOF), the double- clad fiber (DCF), the multi-core fiber (MCF) and the plastic optical fiber (POF), have been experimentally demonstrated [10–19]. As for our previously proposed power-over-SSMF scheme, the co-transmission of 10-W energy light at 1064.8-nm and 5G new radio (NR) 64-level quadrature amplitude modulation orthogonal frequency division multiplexing (64QAM-OFDM) signal at 1550-nm with a data-rate of 1.5-Gbit/s has been reported over the 1-km SSMF [10]. Although the optical power transmission efficiency (OPTE), which is defined as the ratio of the collected optical power to the launched optical power, is 71.8%, the parameter requirement of wavelength division multiplexer, in terms of wavelength isolation and high-power damage threshold, is particularly high in order to suppress the crosstalk between the signal and the energy light. Furthermore, a 2-W energy light at 1480 nm and the 5G NR signal at 1532 nm are co-transmitted over the 10-km SSMF [11]. Such PWOF scheme is examined with various modulation formats, including 256-QAM, 64-QAM, 16-QAM and QPSK. However, the optical-to-electrical conversion efficiencies of photovoltaic converters at 1480 nm is only 26%, leading to the reduction of the collected electrical power from the energy light [11]. The MMF has been also proposed to simultaneously transmit 9.7-W energy light at 1550-nm and 64QAM-OFDM signal at 1310-nm [12]. However, the data rate of radio frequency (RF) signal is limited to 54-Mbit/s, due to the large modal dispersion arising in the MMF. In addition, the mode coupling between the energy light and the optical RF signal causes a performance penalty of the RF signal, with the growing power of the energy light [12]. Please note that, the optical-to-electrical conversion efficiencies of high-power photovoltaic converters (HPPC) at shorter wavelengths are generally higher than those at longer wavelengths. Meanwhile, the transmission loss of energy light at the shorter wavelength is higher than at the longer wavelengths. Thus, high-power energy light at a short wavelength is highly desired for the short-reach PWOF scheme. A MOF has been proposed to co-transmit the 54-Mbit/s 64QAM-OFDM signal at 1550-nm with the 40-W energy light at 976-nm [13]. The lack of a suitable optical coupler to efficiently separate the RF signal from the energy light results in the poor transmission performance of the received signal and the energy light [13]. The power-over-DCF scheme can deliver energy light at 808-nm with a power of either 60-W or 150-W through the large-core 105-µm or 125-µm inner cladding region, while the 64QAM-OFDM signal at 1550-nm at the 8-µm or 9-µm core region [14–16]. Although the 54-Mbit/s RF signal maintains good transmission performance, the transmission loss of the energy light is high [14–16]. In particular, efficiently extracting the optical power from the inner cladding is inconvenient [14–16]. Recently, an optically powered C-RAN network based on the MCF has been demonstrated. A 1.26-W energy light at 1480-nm co-transmits with the RF signal centered at 25.5-GHz over the 10-km weakly coupled seven-core fiber (WC-7CF) [17]. However, it is challenging to provide sufficient electrical power to drive an RAU, when the collected power of energy light is only 133-mW [17]. For other schemes based on WC-MCF, the total optical power arising in the WC-MCF is generally smaller than 1-W, owing to the crosstalk and fiber nonlinearity [18]. The 10-m step-index POF (SI-POF) is also used for the co-transmission of the 4-Gbit/s 16-level pulse amplitude modulation (PAM-16) signal at 650 nm and the 20 mW energy light at 405 nm [19]. Although the PAM-16 signal has better transmission performance, the OPTE is only 20.6% [19] and the additional hardware for the conversion from PAM-16 signal to the 5G NR signal is necessary.
In this work, we experimentally demonstrate the optically powered 5G fronthaul network in support of the co-propagation of spatial-division-multiplexing (SDM) energy light and WDM 5G NR signals over the WC-7CF. When the 60-W energy light at 1064.8-nm is equally distributed among the outer six cores, and the 9-Gbit/s 5R NR WDM signals at 1550-nm are transmitted over the central core of 1-km WC-7CF, we are able to provide six optically powered small cells with the centralized management, the high-speed access, and the low-latency.
2. Experimental setup
As shown in Fig. 2, through simultaneous transmission of the optically carried RF signals and the energy light over a single optical fiber, each small cell with the help of PWoF technique can function independently, leading to a cost-saving and centralize-management implementation. The high-power energy light of the PWoF technique can be centralized at the BBU pool, which is defined as the PWoF pool. Although the power delivery efficiency of PWoF is lower than that of traditional copper wires, PWoF is flexible to satisfy the requirement of each small cell, owing to the PWoF pool centralized at the BBU pool. What’s more, lightning damage is avoided, due to the excellent insulation inherent in the PWoF technique.
Figure 3 shows the experimental setup for the co-transmission of both the 60-W energy light and 9-Gbit/s 5G NR signals over 1-km WC-7CF. Six distributed-feedback laser diodes (DFB-LDs) with each output of 10 dBm are used to generate six continuous-wave (CW) optical carriers, which are subsequently divided into two groups. The channel spacing among six optical carriers is 100 GHz. Meanwhile, according to the 5G NR standard, the 1.5-Gbit/s 64QAM-OFDM electrical signal with a bandwidth of 100 MHz is generated by a vector signal generator (VSG, R&S SMBV100B) at the central frequency of 3.5 GHz. Then, the 1.5-Gbit/s 5G NR electrical signal with an electrical power of 0 dBm, is modulated onto the odd and even groups of six optical carriers, respectively, by two LiNbO3-based electro-optic Mach-Zehnder modulators (MZM, EOSpace AX-0MSS-10-LV) with 3 dB bandwidth of 10 GHz, after the wavelength-division-multiplexing. The MZM, which is biased at the quadrature point of its transfer function, is used to realize the double-sideband (DSB) modulation. Meanwhile, six single-mode high-power laser diodes (SM-HPLDs) with a linewidth of 0.134 nm at 1064.8-nm are used for the SDM-enabled energy light delivery. The output power fluctuation of the HPLD is ±0.5% and its power conversion efficiency is around 30%. As shown in Fig. 4, the six-wavelength 5G NR signals combined by an optical coupler are transmitted over the central core of WC-7CF, which is named as the core-1. Meanwhile, the 60-W energy light spatially transmits over the outer six cores of WC-7CF. Next, the optical carried 5G NR WDM signals and the six-channel energy light are spatially division multiplexed and co-transmit over the 1-km step-index WC-7CF (YOFC, MCF 7-42) with a cladding diameter of 150 µm, enabled by two self-fabricated fused-taper fan-in and fan-out (FIFO) devices with an average insertion loss of 1.3 dB at 1064.8-nm [20]. The cross-section of the WC-7CF is shown in the inset (a) of Fig. 3. The inter-core crosstalk values of the used WC-7CF are -70 dB/km at 1550 nm and -110 dB/km at 1064.8 nm, respectively. The FIFO devices are specially fabricated with a high-power damage threshold, which can realize the linear transmission of 10-W energy light, as shown in the inset (b) of Fig. 3. The power of the six-channel energy light is monitored by an optical power meter (OPM, Laser-Point A-40-D25-HPB) and is converted to electrical power by six HPPCs with optical-to-electrical conversion efficiencies of 35% at 1064.8-nm. Meanwhile, after the wavelength division demultiplexing, each optically carried 5G NR signal is converted to an electrical signal by a photodiode (PD) with a 3 dB bandwidth of 5 GHz, respectively. Meanwhile, both the constellation and the error-vector magnitude (EVM) of individually received 5G NR signals are evaluated by a vector spectrum analyzer (VSA, R&S FSW-K70). Furthermore, a variable optical attenuator (OATT) is used to adjust the power of the 5G NR optical signal, in order to characterize the receiver sensitivity. The received electrical power of a VSA is equal to the product of the square of the optical-to electrical conversion current of PD and the 50Ω matched impedance. Finally, the six-channel converted electrical power and six-wavelength 5G NR electrical signals can be assigned to each small cell with a mapping of one-channel energy light and one-wavelength 5G NR signal.
Fig. 3. Experimental setup for the co-propagation of SDM energy light and 5G NR WDM signals over the 1-km WC-7CF.
3. Results and discussions
We simultaneously transmit the 60-W energy light over the outer six cores of WC-7CF, and the collected optical power of each core is shown in Fig. 5. The six-channel collected optical powers are 4.5-W, 5.1-W, 6.5-W, 6.8-W, 5.7-W, and 5-W, after the 1-km WC-7CF transmission, respectively. The difference of the collected optical power among six cores is mainly due to the core-dependent loss (CDL) arising from the FIFO device. As a result, the average OPTE of the outer six cores is 56%. Meanwhile, the six-channel electrical powers are 1.6-W, 1.8-W, 2.3-W, 2.4-W, 2-W, and 1.8-W, respectively. In addition, since the electrical power consumption of a single small cell is around 1-W [21], six optically powered small cells can be realized. Next, by comparing the optical spectrum of the forward and backward energy light, we identify that, the central wavelength of the energy light is the same, indicating that both the SBS and SRS effects do not occur for the power-over-WC-7CF scheme [10]. Next, we estimate the temporal stability of the high-power energy light delivery, when the 60-W energy light is distributed over six cores of WC-7CF. As shown in Fig. 6, the collected power of six-channel energy light is monitored over 6 hours, with a measurement interval of 10 minutes. The fluctuation of the six-channel collected optical powers are less than 0.2%, indicating the stable function of optically powered small cells.
Figure 7 exhibits the measured EVM values of six-wavelength 5G NR 64QAM-OFDM signals, under the variable received electrical power (REP) with the co-transmission of the 60-W energy light over the 1-km WC-7CF. As a comparison, the EVM values of back-to-back (B2B) transmission are 0.33%, 0.34%, 0.3%, 0.34%, 0.35%, and 0.38%, respectively, when the REP of each wavelength is −25 dBm. Under the same REP, the EVM values of the received six-wavelength 64QAM-OFDM signals with the co-transmission of the 60-W energy light over the 1-km WC-7CF, are 0.33%, 0.33%, 0.3%, 0.35%, 0.36%, and 0.39%, respectively. All EVM values are below the EVM threshold of 9% defined by the 5G NR standard [11]. Since the EVM penalty between the co-propagation of the energy light and the B2B transmission is all smaller than 0.01%, the co-transmission of the 60-W energy light over the 1-km WC-7CF does not bring considerable side effects to the transmission performance of the optically carried 5G NR WDM signals. When the REP is increased, the EVM values of six-wavelength 5G NR 64QAM-OFDM signals become smaller, and the EVM values are almost constant in case the REP is more than -30 dBm. In addition, both the fiber nonlinearity and the inter-core crosstalk have a negligible impact on their transmission performance. Figure 8 presents the measured EVM temporal variation with a duration of 6 hours and a measurement interval of 10 mins. The EVM temporal variation is less than 0.01%, indicating the high-quality transmission of 9-Gbit/s 5G NR WDM signals even under the co-transmission of the 60-W energy light.
Fig. 7. Measured EVMs under different received electrical power with 60-W energy light co-transmission over 1-km WC-7CF.
Overall, the choice of fiber, photonic component, fronthaul reach, and the HPPC is vital for the performance of various PWoF schemes, as shown in Table 1 [10–19]. As the large-core fiber will restrict the data rate due to the occurrence of modal dispersion, it is challenging to realize high-speed 5G access. Especially, when the MMF, the MOF, and the DCF are chosen as the fronthaul link media, the data-rate of RF signals is inferior to that of the power-over-SSMF system. As for our previous power-over-SSMF scheme, it is unable to further enhance the maximum power of the energy light delivery over the SSMF, due to the stimulated Brillouin scattering (SBS) and stimulated Raman scattering (SRS) effects. Consequently, it is not enough to secure the electrical power supply for several small cells. In addition, the use of a wavelength division multiplexer with more than 38-dB wavelength isolation and a high-power damage threshold is costly for the 5G fronthaul network. Alternately, the SDM technique arising in the WC-MCF can integrate multiple SSMFs within the same cladding, which has the potential to greatly improve the capacity of the RF signal transmission and the energy light delivery. Thus, the realization of point-to-multiple fronthaul between the BBU pool and several small cells is possible, when the WC-MCF is expected to replace the SSMF as the 5G fronthaul link medium. Furthermore, the OPTE is jointly determined by the fiber loss at the operation wavelength and the feeding fiber length of the energy light. No matter what fiber is used, the OPTE of high-power energy light is substantially degraded, due to the occurrence of the SBS and the SRS effects after the long-distance fiber optical transmission. However, the longer 5G fronthaul links will obviously have a higher loss, when the used fiber type is fixed. Thus, by taking into account both the scenario of 5G fronthaul and the OPTE, we believe the reach of PWoF is around 1-km, for the ease to cover several small cells. Finally, we also notice that a substantial increase of optical-to-electrical conversion efficiency of the HPPC occurs from the theory to the practice, based on the joint optimization of material, structure, and temperature stability [22]. Meanwhile, optical-to-electrical conversion efficiency of the HPPC is a vital role to evaluate the transmission efficiency of the energy light.
Table 1. Comparison of existing PWoF experiments for optically powered RAUs.a
4. Conclusion
We have experimentally demonstrated the co-transmission of the 60-W energy light and optically carried 9-Gbit/s 5G NR WDM signals over the 1-km WC-7CF. The 60-W energy light is distributed equally among six cores of the WC-7CF, while six-wavelength optically carried 1.5-Gbit/s 5G NR signals transmit over the central core of the WC-7CF. After the successful management of the fronthaul link, a record OPTE arising in the WC-7CF link has been obtained. In addition, the collected power fluctuation among six-channel energy light is smaller than 0.2% over 6 hours, while the EVM fluctuations among six-wavelength 5G NR 64QAM-OFDM signals are less than 0.01%. Through the simultaneous transmission of multi-channel energy light and 5G NR signal, the independent operation of six small cells can be realized by using the WC-7CF based fronthaul link. Given that the centralized management becomes more prominent in the 5G C-RAN network, the optically powered 5G fronthaul is promising to revolutionize the next-generation power supply of 5G small cells.
Funding
National Key Research and Development Program of China (2018YFB1801001); National Natural Science Foundation of China (62025502); Guangdong Introducing Innovative and Entrepreneurial Teams of “The Pearl River Talent Recruitment Program” (2019ZT08X340); Research and Development Plan in Key Areas of Guangdong Province (2018B010114002).
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. Boccardi, R. W. Heath, A. Lozano, T. L. Marzetta, and P. Popovski, “Five disruptive technology directions for 5G,” IEEE Commun. Mag. 52(2), 74–80 (2014). [CrossRef]
2. C. Ranaweera, C. Lim, A. Nirmalathas, C. Jayasundara, and E. Wong, “Cost-optimal placement and backhauling of small-cell networks,” J. Lightwave Technol. 33(18), 3850–3857 (2015). [CrossRef]
3. F. Musumeci, C. Bellanzon, N. Carapellese, M. Tornatore, A. Pattavina, and S. Gosselin, “Optimal BBU placement for 5G C-RAN deployment over WDM aggregation networks,” J. Lightwave Technol. 34(8), 1963–1970 (2016). [CrossRef]
4. J. J. Zhou, L. Gan, C. Chen, S. N. Fu, M. Tang, Q. Yang, and D. M. Liu, “8×10 Gb/s downstream PAM-4 transmission for cost-effective coherent WDM-PON application,” J. Lightwave Technol. 39(9), 2837–2846 (2021). [CrossRef]
5. A. Ghosh, N. Mangalvedhe, R. Ratasuk, B. Mondal, M. Cudak, E. Visotsky, A. Thomas, G. Andrews, P. Xia, H. S. Jo, S. Dhillon, and D. Novlan, “Heterogeneous cellular networks: from theory to practice,” IEEE Commun. Mag. 50(6), 54–64 (2012). [CrossRef]
6. I. Ashraf, F. Boccardi, and L. Ho, “Sleep mode techniques for small cell deployments,” IEEE Commun. Mag. 49(8), 72–79 (2011). [CrossRef]
7. J. D. López Cardona, P. C. Lallana, R. Altuna, A. Fresno-Hernandez, X. Barreiro, and C. Vazquez, “Optically Feeding 1.75 W With 100 m MMF in Efficient C-RAN Front-Hauls With Sleep Modes,” J. Lightwave Technol. 39(24), 7948–7955 (2021). [CrossRef]
8. T. Alade and Q. Z. Ahmed, “Performance Comparison of Small Cell and Distributed Antenna Systems for In-Building Mobile Communications,” in IEEE 89th Vehicular Technology Conference (VTC2019-Spring) (2019), paper 1-6.
9. J. D. L. Cardona, R. Altuna, D. S. Montero, and C. Vazquez, “Power over fiber in C-RAN with low power sleep mode remote nodes using SMF,” J. Lightwave Technol. 39(15), 4951–4957 (2021). [CrossRef]
10. H. L. Yang, D. Peng, Y. W. Qin, J. P. Li, M. Xiang, O. Xu, and S. N. Fu, “10-W energy light co-transmission with optically carried 5G NR signal over standard single-mode fiber,” Opt. Letters 46(20), 5116–5119 (2021). [CrossRef]
11. F. M. A. Al-Zubaidi, J. D. Lopez-Cardona, D. S. Montero, and C. Vazquez, “Optically Powered Radio-Over-Fiber Systems in Support of 5G Cellular Networks and IoT,” J. Lightwave Technol. 39(13), 4262–4269 (2021). [CrossRef]
12. H. Kuboki and M. Matsuura, “Optically powered radio-over-fiber system based on center- and offset-launching techniques using a conventional multimode fiber,” Opt. Lett. 43(5), 1067–1070 (2018). [CrossRef]
13. J. L. Li, A. Y. Zhang, G. Y. Zhou, J. T. Liu, C. M. Xia, and Z. Y. Hou, “A large-core microstructure optical fiber for co-transmission of signal and power,” J. Lightwave Technol. 39(13), 4511–4516 (2021). [CrossRef]
14. M. Matsuura, H. Furugori, and J. Sato, “60 W power-over-fiber feed using double-clad fibers for radio-over-fiber systems with optically powered remote antenna units,” Opt. Lett. 40(23), 5598–5601 (2015). [CrossRef]
15. M. Matsuura, N. Tajima, H. Nomoto, and D. Kamiyama, “150-W power-over-fiber using double-clad fibers,” J. Lightwave Technol. 38(2), 401–408 (2020). [CrossRef]
16. M. Matsuura, H. Nomoto, H. Mamiya, T. Higuchi, D. Masson, and S. Fafard, “Over 40-W electric power and optical data transmission using an optical power,” IEEE Trans. Power Electron. 36(4), 4532–4539 (2021). [CrossRef]
17. J. D. L. Cardona, S. Rommel, E. Grivas, D. S. Montero, M. Dubov, D. Kritharidis, I. T. Monroy, and C. Vazquez, “Power-over-fiber in a 10 km long multicore fiber link within a 5G fronthaul scenario,” Opt. Lett. 46(21), 5348–5351 (2021). [CrossRef]
18. C. Vazquez, J. D. L. Cardona, P. C. Lallana, D. S. Montero, F. M. A. Al-Zubaidi, S. P. Prieto, and I. P. Garcilopez, “Multicore fiber scenarios supporting power over fiber in radio over fiber systems,” IEEE Access 7, 158409–158418 (2019). [CrossRef]
19. F. M. A. Al-Zubaidi, D. S. Montero, and C. Vázquez, “SI-POF Supporting Power-Over-Fiber in Multi-Gbit/s Transmission for In-Home Networks,” J. Lightwave Technol. 39(1), 1–3 (2021). [CrossRef]
20. Y. F. Kang, X. C. Guo, L. Gan, L. Shen, C. Yang, R. Zhang, L. Shen, W. J. Tong, S. N. Fu, M. Tang, and D. M. Liu, “Broadband low-loss fan-in/fan-out devices for multicore fibers,” in Asia Communication and Photonics Conference (ACP) (2019), paper T1A.4.
21. D. L. Perez, X. L. Chu, and I. Guvenc, “On the expanded region of picocells in heterogeneous networks,” IEEE J. Sel. Top. Signal Process. 6(3), 281–294 (2012). [CrossRef]
22. S. Kouro, J. I. Leon, D. Vinnikov, and L. G. Franquelo, “Grid-connected photovoltaic systems: an overview of recent research and emerging PV converter technology,” IEEE Ind. Electron M. 9(1), 47–61 (2015). [CrossRef]
