Open Access
Add to BibSonomy
Abstract
This study proposed a new small star topology network architecture connecting a secondary fiber link between adjacent small cell base stations for disaster tolerance in the access span of a short-reach point-to-multipoint optical communication system. Compared to the edge-side optical node independent architecture, the proposed network architecture exhibits a high robustness and small number of optical components. We compared the conventional and proposed configurations for survivability in the case of a major disaster with a triple-link failure. The proposed method made it easy to add a small star-shaped network with secondary links for significant edge-side optical nodes while maintaining the basic configuration of the conventional passive optical network.
© 2022 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
With the commencement of 5th generation (5G) mobile communication system operation in 2020, wireless communication traffic will continue to increase [1]. Since 5G-terminal electromagnetic waves have high directivity, it is essential to arrange 5G base stations more densely than 4th generation (4G) base stations [2], the range over which the base stations can access to radio terminals being referred to as small cells. Moreover, Supplement 66 to the international telecommunication union telecommunication standardization sector (ITU-T) G-series Recommendations stipulates that for 5G mobile fronthaul architectures the transport architecture should consist of three components—that is, the central unit (CU), distribution unit (DU), and remote unit (RU) [3]. The configuration in which the DUs are centrally located in the access convergence room or small access room is called centralized radio access network (C-RAN), which usually requires many trunk fibers for one-to-one facing fiber connections as the distance between the DUs and RUs can be 10 km or more [4]. The number of trunk fibers can be reduced by applying a passive optical network (PON) consisting of a point-to-multipoint physical topology that can efficiently accommodate long-distance sections using a single optical fiber. When a PON is applied to C-RAN, an optical network unit (ONU) can be placed in the RU, an optical line terminal (OLT) being set in the DU/CU [5,6].
Since the 3GPP standardization organization requires highly reliable communication for 5G mobile networks (NWs) [7], the high reliability of PONs—which are equivalent to mobile front halls—is essential [8]. In Japan, the 2011 Great East Japan earthquake and tsunami caused enormous damage to the optical access system that housed mobile traffic [9]. Due to the optical fiber configuration, the risk of link failure in a major disaster was exceptionally high. Consequently, the ITU-T put together several methods for PON protection [10]—for example, Type A and Type B configurations are partial link protection methods that support trunk fiber failures alone by arranging redundant trunk fibers, but do not support access fiber failures. Moreover, recently reported shared protection methods, which use cost-effective secondary links was configured to support only trunk fiber failures [11,12]. Many other trunk fiber protection methods have been reported [13–16]. The Type C and Type D configurations defined by the ITU-T are overall link protection methods that support trunk fiber and access fiber disconnections by pairing PONs. However, these methods are based on the premise that the remote nodes of the paired PONs are near the same ONU. Since it is necessary to obtain the access status between two PONs after the access fiber has been disconnected, Layer 2 control links between the PONs is indispensable. Consequently, resource management can be complicated, and it may take time for reconnection after access link failure.
The transmission capacity per wireless base station can be expected to reach 100 Gbps in the beyond-5G mobile fronthaul networks [17–19]. Coherent PONs that adopt coherent methods offer excellent capacity per wavelength, making them good candidates for commercialization—for example, the C form-factor pluggable 2 (CFP2) coherent transceiver supports 100 Gbps has already been commercialized and will be increasingly used in the future [20]. Furthermore, since the demand for time particle size (for maximum transmission capacity) decreases and the need for low delays increases, wavelength division multiplexing (WDM)-based PONs are good candidates too [21–24]. The maximum number of coherent WDM-based PONs reported so far is 64 [25]. With coherent WDM-based PONs, the number of uplink local oscillators (LOs) and coherent receivers increases on the OLT side as the number of ONUs increases. As a way of reducing the number of LOs, a configuration in which the light sources can be integrated into a single unit using a comb light source has been proposed [26]. As a way of reducing the number of coherent receivers, a configuration in which a plurality of WDM signals can be collectively received using an uplink WDM signal has also been reported [27–29].
The four ONU collaborative star topology architecture connecting secondary links for coherent WDM-based PONs can handle multiple access fiber disconnections while maintaining the structure within a single PON. Of the four ONUs, one ONU is the core node and the other three ONUs are adjacent nodes, all of which are connected in a small star topology using secondary links. Making the ONU node configuration colorless and directionless using optical switches and couplers makes reconnection possible via the core node even if the access fiber is in failure. Recently, there have been reports of optical switches using simple configurations being placed on the ONU side [30]. Our study is an extended version of the work reported in the literature [31], the novelty of which is that we conducted experiments on a four ONU star-type configuration when three access fibers were disconnected simultaneously. Previous studies have only considered double-access link failures.
The remainder of this paper is structured as follows: Section 2 introduces the types of secondary link NWs between ONUs. We describe the route switching based on access fiber disconnections in the secondary link star topology configuration between four ONUs. Section 3 describes the node configuration of the secondary fiber NW between ONUs. Section 4 evaluates the results of survival rate calculations for secondary fiber NWs between ONUs. Section 5 describes the setup of the principle-confirmation experiment. Section 6 describes the experimental results of an 8-ch WDM 100 Gbps dual-polarized quadrature phase shift keying (DP-QPSK) signal emulating access link failure. We demonstrated that the secondary link star topology architecture between four ONUs could be reconnected under conditions of triple-access fiber failure.
2. Theory: types of ONUs collaborative NWs with secondary link
Small cells in 5G networks can be classified into three types based on their area radius. Cells with a cell radius of several meters to 10 meters are called femtocells and can be used in homes and offices. Cells with a cell radius of 10 m to several tens of meters are called picocells, and cells with a cell radius of tens of meters to hundreds of meters are called microcells and can be used outdoors. The target of the small cell is an RU placed in a picocell or microcell.
Figure 1 shows the structure of a standard PON and a PON with three types of secondary link NWs. Figure 1(a) shows a configuration in which the CUs/DUs located in optical aggregation stations and the RUs situated in radio base stations in many microcells are accommodated by PONs. Figure 1(b) shows the configuration of a PON in which two RUs in adjacent small cells are paired, and a secondary link between two ONUs is arranged. Figure 1(c) shows the configuration of a PON in which four RUs in adjacent small cells are linked, and the secondary links between four ONUs are connected in an entire mesh topology. Figure 1(d) shows the configuration of a PON that links four RUs in adjacent small cells and connects the secondary links between four ONUs in a star topology.
Fig. 1. Configuration of a PON for small cells: (a) standard (Type 1), (b) paired ONU (Type 2), (c) secondary link full-mesh topology between four ONUs (Type 3), (d) secondary link star topology between four ONUs (Type 4).
The length of the trunk fiber in C-RAN is 20 km. By arranging the optical coupler corresponding to the remote node at the position of a small cell group’s center of gravity, the access fiber’s length becomes several km, corresponding to the macro-cell radius. The maximum length of the secondary link between four ONUs is several hundred meters, a small NW with secondary link. The power supply can be easily allocated since the optical switch is in the same position as the ONU. Additionally, the secondary link NW does not have to be applied to all ONUs in the standard PON but can be used only for particularly significant RUs.
Here, we describe the route when a link failure occurs in WDM-based PONs to which the 4-ONU secondary link star topology can be applied. In single-fiber bidirectional transmission, different wavelengths can be used for uplinks and downlinks, though here we explain the wavelength and path in just the uplink. The access link can be expressed as km, the secondary link as lm,n, m and n being the number of ONUs to be connected, respectively.
Figure 2(a) shows the access status of each ONU when there is no link. ONU #1 to #4 uses a dedicated access link. Figure 2(b) shows the access status after the access link k4 is disconnected. Adjacent ONUs affected by the broken link use the access link of the core ONU. The uplink signal of ONU #1 can be transmitted over k1 via l4,1.
Fig. 2. Access fiber failure cases: (a) no failure case, (b) single failure case 1, (c) single failure case 2, (d) double failure case 1, (e) double failure case 2, (f) triple failure case.
Figure 2(c) shows the access status after the access link k1 is disconnected. The core ONU affected by the access link disconnection uses the access link of the adjacent ONU. The uplink signal of ONU #4 can be transmitted over k4 via l1,4.
Figure 2(d) shows the access status after the simultaneous disconnection of access links k1 and k4. Adjacent ONUs and the core ONUs affected by the access link disconnection can use the access links of the unaffected adjacent ONUs. The uplink signal of ONU #1 can be transmitted over k3 via l1,3, and the uplink signal of ONU #4 can be transmitted over k3 via l4,1, and l1,3.
Figure 2(e) shows the access status after the access link k1 and secondary link l1,4 are simultaneously disconnected. The core ONU affected by the access link disconnection and the secondary link disconnection can use the access link of the adjacent ONU that is unaffected. The uplink signal of ONU #1 can be transmitted over k3 via l1,3.
Figure 2(f) shows the access status after the simultaneous disconnection of access links k1, k2, and k4. The core ONU is affected by the access link failure, and the two adjacent ONUs can use the access link of the unaffected adjacent ONU. The uplink signal of ONU #1 can be transmitted over k3 via l1,3, ONU #2 can be transmitted over k3 via l2,1 and l1,3, and ONU #4 can be transmitted over k3 via l4,1 and l1,3.
3. Methods: configuration of secondary fiber NW between ONUs
Figure 3 shows the four types (Types 1–4) of secondary fiber NW protection architectures. A protection switch consists of feeder fibers, a coupler (CPL), and an optical switch. Table 1 summarizes three types of secondary link NWs. From the cost of view, optical couplers are passive and relatively inexpensive devices. It is essential to select what kind of optical switch to reduce the cost impact. Mechanical non-latch optical switches have already been commercialized on 1 × 2 and 1 × 4 switch scales and can realize protection configurations with relatively low power consumption and cost. Since the specific cost fluctuates greatly depending on the price of the optical switch, the number of optical switches and optical couplers is limited. The switch configuration in which the core ONU and the adjacent ONU are linked comprises a passive optical device capable of branching and coupling and an active optical device capable of switching. Figure 4 summarizes the passive optical branching, coupling devices, and active optical switches with and without wavelength dependence. Optical devices without wavelength dependence are optical couplers and optical switches, and optical devices with wavelength dependence include array waveguide grating and wavelength selection switches (WSSs). The ONU collaborative switch—comprising just a wavelength-independent optical coupler and an optical switch—can realize a colorless and directionless protection function. The colorless function allows one to change the ONU’s wavelength without physically changing the connection. The directionless function allows the output access fiber to be freely changed after link failures.
Fig. 3. Secondary fiber NW architectures: (a) standard (Type 1), (b) paired ONU (Type 2), (c) secondary link full-mesh topology between four ONUs (Type 3), (d) secondary link star topology between four ONUs (Type 4).
Table 1. Comparison of three types of secondary link NWs
Before a disaster occurs, the first backup path can be determined to reduce the recovery time for protection. If the route is not restored from the disconnection after switching paths, the route is automatically changed sequentially. To reduce the cost of the protection switch associated with the number of ports and to save the CPL branch loss associated with an increase in the number of switch ports, one can use as few port switches as possible—for example, 1×2 and 1×4—for the deployed protection switch in each ONU. Distributing optical devices with simple configurations is an essential disaster resistance factor.
Here, we assume that all eight ONUs use the link simultaneously. NW protection considers one group of protection devices consisting of a set of four ONUs—a Type 1 NW is a standard PON structure without secondary fibers and 1×N optical switch; Type 2 NWs can only be connected to adjacent ONUs using 1×2 optical switches; a Type 3 NW can be connected to all ONUs using 1×4 optical switches, each ONU being able to transmit to three of the four ONUs in one hop; a Type 4 NW has a core ONU connected to three adjacent ONUs, and three of the four ONUs are connected to the core ONU. The advantage of the Type 4 NW is that all ONUs can be connected in two hops.
Figure 5 shows the total number of switching ports and the maximum CPL insertion loss. The Type 3 NW includes the cost of optical switches, secondary fibers, and insertion loss at the CPL compared to the other NW types. The Type 4 NW also exhibits a substantial insertion loss, but its installation cost is low.
4. Connection survivability calculation model and results
This section describes the simulation model and results regarding the connection survivability associated with the occurrence of link failures. The definition of a routing problem for a collection of static traffic should use 1:N protection and sequential switching, the heuristics for solving the routing problem being described below.
Given: NW topology G (V, E) where V denotes a set of nodes and E a set of single-fiber links connecting two nodes in V; an ordered set of frequency-slot units (FSUs) F = f(${f_1},{f_2}, \cdots {f_{|V |}}$) for each link; traffic demand set D = d (${d_1},{d_2}, \cdots {d_{|V |}}$), where d is determined by the source, and destination nodes, and the requested capacity c is expressed in 100 Gb/s; the required FSU count as a function of the requested capacity f(c) is 1.
Objective: Find a backup path for each demand that is identified by the route to be used in each link on the route.
Constraints:
- • All FSUs used on each link between a source node and a destination node must use the same wavelength.
- • Working path w and its backup path b must be a link- and node-disjoint.
- • The FSU cannot share between wi and wj; wi and bj; or wj and bi, ${b_i}$ and ${b_j}$.
- (1) For each node pair (i, j), we calculate the working paths (WPs) and respective backup paths (BPs). Then i is the ONU number from 0 to 7, and j is the OLT number 8. As a deciding method of BP, the BP candidates are sorted in ascending order by hop number, and then the OLT node numbers that pass through them are sorted in ascending order if they are BP candidates with the same hop number.
- (2) We select the top WP/BP pair amongst the sorted results.
- (3) For each WP and BP, one FSU is assigned as the FSU of each WP source node number.
- (4) We simultaneously cut any links (${l_i}$ for a single link failure, ${l_i}\; \textrm{and}\; {l_j}$ for double-link failures, ${\; }{l_i}$, ${\; }{l_j}$, and ${\; }{l_k}$ for triple-link failures), in the NW and then create demand (${D_i}$ for a single link failure, ${D_i}$ and ${\; }{D_j}$ for double-link failures, ${\; }{D_i}$, ${\; }{D_j},{\; \textrm{and}\; }{D_k}$ for triple-link failures) that comprise demands affected by the cuts at links, respectively, and switch ${D_i}$, ${\; }{D_j},{\; \textrm{and}\; }{D_k}$ to the backup path.
- (5) When ${D_i}$, ${D_j}$, or ${D_k}$ working and backup paths are simultaneously cut, we search for remaining NW resources and calculate a detour route. The switches are sequentially changed in order from the smallest ONU number node to search for a communicable link.
Figure 6 show the average CS in the overall PON for the only protection and protection with sequential switching in the case of double- and triple-link failures. The protection with sequential switching in Type 3 and Type 4 NWs achieves a CS of 100% and 80%, respectively, in the case of triple-link failures. When comparing Type 2 and Type 4 NWs with the same number of additional elements, the Type 4 NW increases the CS by 19.4% by adding 1.5 times the amount of hardware elements.
5. Experimental setup and results
Figure 7 shows the experimental setup of secondary fiber NWs (Types 1, 2, and 4) for the uplink/downlink of 8λ×100 Gbps/λ DP-QPSK signal with 50 GHz channel spacing using a protection switch. The Type 3 NW was beyond the scope of this experiment because it was not the configuration proposed in this paper. Since there is no interaction—such as reflection— between uplinks and downlinks in bidirectional communications, by assigning different wavelengths to them the uplinks and downlinks can be evaluated in other experimental systems.
Table 2 and Table 3 show the upstream and downstream optical frequency assignments for the ONUs. To explain the concept behind protection using sequential switching in the case of the access-span link failure, we emulated single-, double-, and triple-link failures and 1:N protection using sequential switching for standard (Type 1), paired ONU (Type 2), and secondary link star topology between four ONUs (Type 4), as shown in Fig. 3. Here, the trunk fiber length of the single-mode fiber had a reach of 50 km, fiber loss of 0.18 dB/km, and chromatic dispersion of 17 ps/nm/km. We assumed the length to be equivalent to that of a long reach PON, which was sufficiently longer than the standard 20 km. The insertion loss of the 1×2 and 1×4 switch was 0.6 dB.
Table 2. Wavelength assignment of uplink transmission
Table 3. Wavelength assignment of downlink transmission
A protection switch was set to the working path of the same route as the standard PON before a disaster occurred After link failure, the protection switch operated based on protection using sequential switching. A WDM signal was generated by the optical transmitter (Tx)— including the laser diode (LD)-array, CPL, IQ modulator, and real-time digital signal processor (DSP)—before being launched into the trunk fiber. At the receiver (Rx) side of each ONU, the signal power was set using a variable optical attenuator, mixed with the LO light(s), and coherently detected. The detected signal was then regenerated and the bit error ratio (BER) was measured using a real-time DSP. For emulating the independent signal generation from different transmitters, the bundled modulation signal was divided using a WSS and delayed by an additional optical fiber in the uplink, the other setups of the uplink and downlink remaining the same. The WSS is not used in the practical system. The fiber link failure in the experiment was emulated by opening the fiber patch cord to generate Fresnel reflection at the interface with air. When a single link failed, the end face of the access fiber of ONU 8 was open. When two links failed, the end faces of the access fibers of ONU 7 and 8 were empty. With three-link breaks, the end faces of the access fibers of ONU 6, 7, and 8 were open.
Figures 8, 9, and 10 show the measured BERs of all ONU uplinks in the case of the before failure, after single, double, and triple-link failures, and after sequential switching of the uplink. Figures 8, 9, and 10 show the cases of the standard PON, paired ONU, and four ONUs star topology. Figure 11 shows the measured BERs of all ONU downlinks in the case of triple-link failures. Figures 11(a), 11(b), and 11(c) show the cases of the standard PON, paired ONU, and four ONUs star topology. The standard PON cannot connect with the same number of ONUs as the number of access link failures. Moreover, the paired ONU topology cannot be connected if two or more links fail. However, the four ONUs star topology can be connected even if triple-link failure occurs.
Fig. 8. Measured BERs of the standard PON (a) no-link failure, (b) single link failure, (c) double-link failure, (d) triple-link failure.
Fig. 9. Measured BERs of paired ONUs topology (a) no-link failure, (b) single link failure, (c) double-link failure, (d) triple-link failure.
Fig. 10. Measured BERs of four ONUs star topology (a) no-link failure, (b) single link failure, (c) double-link failure, (d) triple-link failure.
Fig. 11. Measured BERs of triple-link failure case (a) the standard PON, (b) paired ONU topology, (c) four ONUs star topology.
All ONUs received power equalized in the case of no-link failure. Here, we associate a 7% hard decision forward error correction (HD-FEC) limit (3.8×10−3) with achieving error-free operation. The power penalties—that is, the receiver sensitivity difference at the HD-FEC limit—between inside and outside channels were less than 1 dB due to the differences in crosstalk. There was no penalty due to the influence of linear crosstalk due to the power difference between ONUs after protection with sequential switching.
6. Conclusions
We proposed two-ONU topology, four-ONU mesh topology, and four-ONU star topology NWs between ONUs consisting of disaster resistance secondary links as access links for point-to-multipoint WDM-based PON systems. Specially, the four-ONU star topology showed that the number of additional parts for a standard PON was small, and the survivability ratio for triple-link failure was high in the event of a major disaster. When an 8λ×100 Gbps coherent DP-QPSK signal was applied to a four-ONU star topology NW, we confirmed that there was almost no receive power penalty associated with self-healing for a specific three-link failure. In this study, four ONUs formed one sub-link network because the number of optical switch ports and branch loss in the protection switch increase as the network scale increases. Currently, there are no reports of a complete access link protection architecture within a single PON for a mobile fronthaul system. In future work, we aim to build a PON system with robust access and trunk spans by using the protection method for access links proposed here and the protection for trunk spans that has been studied elsewhere.
Funding
Japan Society for the Promotion of Science (JP22K04105).
Acknowledgments
We thank Y. Azuma and T. Ishikawa of the Kagawa University for their support in the calculations and experiments. We would like to thank Editage (www.editage.com) for English language editing.
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. T. Nakamura, “5G evolution and 6G,” Proc. IEEE Symposium VLSI Technology, pp. 1–5, Jun. 2020.
2. X. Ge, H. Cheng, M. Guizani, and T. Han, “5G wireless backhaul networks: challenges and research advances,” IEEE Networks 28(6), 6–11 (2014). [CrossRef]
3. ITU-T G Suppl. 66, “5G wireless fronthaul requirements in a passive optical network context,” Oct. 2018 [Online].
4. A. Oliva, J. A. Hernandez, D. Larrabeti, and A. Azcorra, “An overview of the CPRI specification and its application to C-RAN-based LTE scenarios,” IEEE Commun. Mag. 54(2), 152–159 (2016). [CrossRef]
5. J. Kani, J. Terada, K. Suzuki, and A. Otaka, “Solutions for Future Mobile Fronthaul and Access-Network Convergence,” J. Lightwave Technol. 35, 1 (2016). [CrossRef]
6. J. S. Wey and J. Zhang, “Passive optical networks for 5G transport: technology and standards,” J. Lightwave Technol. 37(12), 2830–2837 (2019). [CrossRef]
7. A. Ghosh, A. Maeder, M. Baker, and D. Chandramouli, “5G evolution: A view on 5G cellular technology beyond 3GPP release 15,” IEEE Access 7, 127639–127651 (2019). [CrossRef]
8. G. Mountaser, T. Mahmoodi, and O. Simeone, “Reliable and low-latency fronthaul for tactile Internet applications,” IEEE J. Select. Areas Commun. 36(11), 2455–2463 (2018). [CrossRef]
9. M. Kobayashi, “Experience of infrastructure damage caused by the great east Japan earthquake and countermeasures against future disasters,” IEEE Commun. Mag. 52(3), 23–29 (2014). [CrossRef]
10. ITU-T G Suppl. 51, “Passive optical network protection considerations,” June 2017 [Online].
11. T. Kodama, T. Goto, T. Nakagawa, and R. Matsumoto, “Bypass/backup-link switchable coherent point-to-multipoint configured WDM-PON system with shared protection,” Opt. Express 30(7), 10502–10512 (2022). [CrossRef]
12. T. Kodama, T. Nakagawa, and R. Matsumoto, “Any-Double-Link Failure Tolerant Bypass/Backup Switchable WDM-PON Employing Path-Pair Shared Protection and Bidirectional Wavelength Pre-assignment,” Proc. Optical Fiber Communication Conference (OFC), Online, USA, Mar. 2022, W3G.5.
13. F. J. Effenberger, “PON Resilience,” J. Opt. Commun. Netw. 7(3), A547–A552 (2015). [CrossRef]
14. K. Honda, H. Nakamura, K. Sone, G. Nakagawa, Y. Hirose, T. Hoshida, and J. Terada, “Wavelength-shifted protection for WDM-PON with AMCC scheme for 5G mobile fronthaul,” Proc. Optical Fiber Communication Conference (OFC), San Diego, USA, Mar. 2019, W3J.6.
15. S. Zhang, W. Ji, X. Li, K. Huang, and Z. Yan, “Efficient and Reliable Protection Mechanism in Long-Reach PON,” Journal of Optical Communication and Networking 8(1), 23–32 (2016). [CrossRef]
16. S. Kaneko, T. Yoshida, S. Kimura, and N. Yoshimoto, “Reliable λ-Tuning OLT-Protection Method Based on Backup-Wavelength Pre-assignment and Discovery Process for Resilient WDM/TDM-PONs,” J. Lightwave Technol. 33(8), 1617–1622 (2015). [CrossRef]
17. T. Kodama, R. Matsumoto, and N. Suzuki, “Demonstration of data-rate and power-budget adaptive 100 Gb/s/λ-based coherent PON downlink transmission,” Proc. Optical Fiber Communication Conference (OFC), Los Angeles, USA, Mar. 2017, Th2A.22.
18. M. S. Faruk, X. Li, and S. J. Savory, “Experimental demonstration of 100/200-Gb/s/λ PON downlink transmission using simplified coherent recievers,” Proc. Optical Fiber Communication Conference (OFC), San Diego, USA, Mar. 2022, Th3E.5.
19. Y. Zhu, L. Yi, B. Yang, X. Huang, J. S. Wey, Z. Ma, and W. Hu, “Comparative study of cost-effective coherent and direct detection schemes for 100 Gb/s/λ PON,” J. Opt. Commun. Netw. 12(9), D36–D47 (2020). [CrossRef]
20. Y. Loussouarn, E. Pincemin, M. Pan, G. Miller, A. Gibbemeyer, and B. Mikkelsen, “Multi-rate multi-format CFP/CFP2 digital coherent interfaces for data center interconnects, metro, and long-haul optical communications,” J. Lightwave Technol. 37(2), 538–547 (2019). [CrossRef]
21. N. Suzuki, H. Miura, K. Matsuda, R. Matsumoto, and K. Motoshima, “100 Gb/s to 1 Tb/s based coherent passive optical network technology,” IEEE/OSA Journal of Lightwave Technology 36(8), 1485–1491 (2018). [CrossRef]
22. M. Yoshida, T. Kan, K. Kasai, T. Hirooka, K. Iwatsuki, and M. Nakazawa, “10 channel WDM 80Gbit/s/ch, 256 QAM bi-directional coherent transmission for a high capacity next-generation mobile fronthaul,” J. Lightwave Technol. 39(5), 1289–1295 (2021). [CrossRef]
23. D. V. Veen, “Transceiver technologies for next-generation PON,” Proc. Optical Fiber Communication Conference (OFC), San Diego, USA, Mar. 2020, W1E.2.
24. J. S. Wey, Y. Luo, and T. Pfeiffer, “5G wireless transport in a PON context: An overview,” IEEE Comm. Stand. Mag. 4(1), 50–56 (2020). [CrossRef]
25. S. Smolorz, E. Gottwald, H. Rohde, D. Smith, and And A. Poustie, “Demonstration of a coherent UDWDM-PON with real-time processing,” Proc. Optical Fiber Communication Conference (OFC), Los Angeles, USA, Mar. 2011, PDPD4.
26. H. Zhang, M. Xu, Z. Jia, and A. Campos, “Frequency comb and injection locking based mutual protections in coherent optical access network,” Proc. Optical Fiber Communication Conference (OFC), San Diego, USA, Mar. 2022, M1I.5.
27. J. Zhang, Z. Jia, H. Zhang, M. Xu, J. Zhu, and L. Alberto, “Rate-flexible single-wavelength TFDM 100G coherent PON based on digital subcarrier multiplexing technology,” Proc. Optical Fiber Communication Conference (OFC), San Diego, USA, Mar. 2020, W1E5.
28. D. Welch, A. Napoli, J. Back, W. Sande, J. Pedro, F. Masoud, C. Fludger, T. Duthel, H. Sun, S. J. Hand, T.-K. Chiang, A. Chase, A. Mathur, T. A. Eriksson, M. Plantare, M. Olson, S. Voll, and K.-T. Wu, “Point-to-multipoint optical networks using coherent digital subcarriers,” J. Lightwave Technol. 39(16), 5232–5247 (2021). [CrossRef]
29. T. Kodama, K. Shimada, R. Matsumoto, T. Inoue, S. Namiki, and M. Jinno, “Frequency-packed multiband-coherent transceiver with symbol rate-adaptive Nyquist WDM signals,” IEEE Photonics Technol. Lett. 33(21), 1205–1208 (2021). [CrossRef]
30. N. Parkin and A. Rafel, “Novel low cost PON protection via harvested power,” Proc. OFC, San Diego, CA, USA, Mar. 2020, Th2A.35.
31. T. Kodama, Y. Azuma, and T. Ishikawa, “Colorless and directionless coherent WDM-PON architecture with extended star topology using a self-healing for bidirectional link protection,” Proc. Opto Electronics 8and Communications Conference (OECC), T1-3.6, Taipei, Taiwan, Nov. 2020.
