Key New Fiber Wireless Access Technologies for 5G and Beyond
Gee-Kung Chang, You-Wei Chen, School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA, and Jeff Finkelstein, Cox Communications, Atlanta, GA
IEEE 5G Tech Focus: Volume 3, Number 2, September 2019
To make the performance goals of 5G a reality, New Radio Access Network Architecture based on Fiber-Wireless Integration and Networking (FiWIN) is essential in serving diverse user scenarios, which require high wireless throughput, extremely low-latency, ultra-reliability and seamless multi-device connectivity. This paper reveals some of grand challenges in 5G mobile fronthaul and the enabling radio access technologies as well as the research breakthroughs for heterogeneous mobile data communications.
To date, there is a tremendous traffic growth in mobile applications, e.g., high data rate enhanced mobile broadband (eMBB), virtual reality/augmented reality (VR/AR) wearables, industrial internet of thing (I-IoT), intelligent transport systems (ITS) and vehicle-to-everything (V2X) communication, that motivates researchers to explore research scopes defined by 5G standards and beyond in its system data throughput and network diversity. To deliver the required data seamlessly from a cloud data center to wireless subscribers, a 3G/4G/5G co-existence and heterogeneous radio access in conjunction with a fiber-based mobile fronthaul design is extremely important. The mobile fronthaul architecture gradually migrates form the legacy distributed system to the hoteling and pooling of the base band units (BBUs) system, and currently toward the centralized signal processing in the central units (CUs) and distributed units (DUs). The centralized management enables many beneficial properties for 5G network, including the higher spectral efficiency radio-over-fiber (RoF) transmission, coordinated multi-point (CoMP) data delivery, dense deployment of simplified remote radio units (RRUs) and AI optimized networking, etc. On the other hand, to economically facilitate the small cell densification, using the existing cable infrastructure in a hybrid fiber coaxial (HFC) architecture can save the construction and civil works cost, since there is already 93% broadband access had coverage by cable in the United States . In this paper, we will focus on and highlight some of the key enabling radio access technologies and our recent research achievements for 5G heterogeneous fiber-wireless integration data communications, including 5G new radio (NR), signal formats, all-spectrum heterogeneous network, visible light communication (VLC) based positioning and millimeter wave (MMW) beamforming, as shown in Fig. 1.
Figure 1. Key enabling radio access technologies for 5G heterogeneous wireless data communications.
Due to the rapid increment of wireless data demand, the currently available low frequency wireless spectrum (sub 6GHz) is overcrowded and thus operators can leverage NR (new radio) above 24-GHz MMW frequency bands to offload data from legacy 3G/4G macro cell and relieve the network traffic congestion. 5G-NR has an abundance of spectrum and natively smaller RRU coverage, which results a higher spatial reuse network design and a limited inter-cell interference. In 3GPP Ref.15, the 5G-NR up to 52.6 GHz is standardized, while frequency ranges from 52.6 GHz to 114 GHz are extensively discussed for 5G beyond network. However, it is not a straightforward migration for NRs, some key radio access technologies (RATs) need to be further investigated and discussed in the following content.
3. Signal Waveform Designs for 5G-NR
A debate of multi-carrier or single-carrier data delivery is never dismissed. Family of multi-carrier signal formats, e.g., orthogonal frequency division multiplexing (OFDM) and filter bank multi-carrier (FBMC), have various benefits. By de-multiplexing a high-speed data stream into several narrow-band signals and parallelly transmitting them over orthogonal subcarriers, such manipulation enables a reduced inter-symbol interference (ISI) and a simplified one-tap frequency domain equalizer. Moreover, if the transmitted channel is relatively static and the transmitter knows the channel information, an adaptive power- and bit-loading could be applied among each subcarrier to maximize the channel capacity. However, the main drawback of the multi-carrier signal is its high peak-to-average power ratio (PAPR), which makes the RF power amplifier (PA) design challenging in PA linearity and leading a power back-off, especially for higher RF frequencies of 5G-NRs.
On the other hand, the multipath effect is non-significant for 5G-NR frequency channels, and most application scenarios are mainly relying on the line-of-sight (LoS) transmission, which means the channel response is relatively flat. Therefore, single-carrier signal format or DFT-spread-OFDM (DFT-s-OFDM) with lower PAPR are considered as the promising waveforms. The world record single carrier wireless data delivery of a photonics-aided MMW carrying single-carrier 64-QAM and over 1 Tb/s at D-band was experimentally demonstrated and transmitted over 3.1m wireless distance .
Beside the aforementioned signal formats, non-orthogonal multiple access (NOMA)  technology is widely discussed in the standard body for further increasing the available subscriber number. Via allocating the multiple access signature in the power domain (PD), distinct users can share the same resource element (i.e., time or frequency) and be separated at the receiver-side. Therefore, with additional PD resource, a 3D resource allocation is achieved with a finer granularity than 3G/4G network. However, conventional successive interference cancellation would inevitably induce latency and error propagation issues for users in the worse channel condition. A deep neural network (DNN) empowered bi-directional PD-NOMA in optical access network was demonstrated, and it can individually identify each users’ multiple access signature and decode all the users data simultaneously .
4. All-spectrum Heterogeneous Radio Access Technologies
Given that abundant NR spectra are introduced in 5G, a compatible mobile fronthaul is needed to get benefits from 3G/4G or HFC networks and facilitate 5G implementation. As the delivering RF carrier becomes higher, the 5G cell will become smaller, and thus a huge amount of wired fronthaul may be distributed for supporting the wideband radio data delivery. In the case that wired fronthaul is presently unavailable or too expensive in capital expenditure, e.g., in urban area or in disaster area, a heterogeneous fiber-wireless integration network can offer an alternative approach to achieve a Gbps-class ubiquitous-connectivity. However, the weather condition is one of the issues for the wireless MMW relay. To overcome such bottleneck, a weather-proof design by MMW/ free space optics (FSO) architecture was introduced in a mobile fronthaul  through the mutually complementary channel response. A hybrid transmission with adaptive diversity combining technique (ADCT) was demonstrated with 4-Gb/s OFDM to build a high-reliability wireless connection under fog, rain, and turbulence conditions.
Without considering the poor weather conditions, even in a sunny day, the blocking is another fundamental issue for LoS 5G-NR transmission, which causes wireless service interruption. With the help of the dense deployment of the 5G small cells, it is natural to assemble multiple cells for conducting a multi-connectivity transmission of 5G-NRs. A reconfigurable multiple-input multiple-output (MIMO) MMW mobile fronthaul  was demonstrated with enhanced path diversity, non-LoS tolerance, and infrastructure utilization as compared to the conventional single-input single-output system. The results reveal that a multi-point multi-section mobile fronthaul with dual RRUs connection can greatly extend its wireless transmission distance by 22% and enhance its tolerance of antenna misalignment by 76%.
Although the mutli-connectivity design could enhance the network reliability, the inter-cell interference (ICI) may degrade the received performance without an appropriate and reasonable control mechanism. A CoMP technology is needed to synchronize adjacent cells in the time and frequency domain. A centralized management of CU and DU could deliver the control signal through a low frequency microcell to provide frequency handover and synchronization of small cells. Therefore, the mobility and connectivity in the 5G-NR networks can be enhanced and provide the diverse services on the fly through the small cells. However, in an urban area, the available space for the small cell deployments would be limited and irregular, and thus it complicates the time discrepancy among RRUs, which increases the difficulty of synchronization. A complex-valued artificial neural network (ANN) was firstly proposed to address this issue, and it enhances the centralized network design for the MIMO 5G-NR mobile fronthaul . Three time increment/improvement of time discrepancy tolerance was achieved as compared to the traditional MIMO scheme.
5. Positioning and MMW Beamforming
Since the 5G-NR poses to go to higher RF carrier frequency, the propagation loss would generally increase. Thus, a phase array antenna with large array size must be employed in both the transmitter and receiver sides for compensating the path loss, leading to eventually a “pencil-like” beam-width property and a higher spatial resolution. Because of the narrow beam-width, a new resource for wireless data communication is introduced, which is the spatial domain as shown in Fig. 2. In contrast to the widely discussed power domain NOMA, the spatial dimension in 5G and beyond can be a fully orthogonal wireless resource aided by the advanced IC design technology of spatial selectivity , and each signal with relative reception angle can be decoded individually. The fast-developing technologies of MMW positioning combined with beamforming enables operators to manipulate the signal transmission in spatial selectivity for further enhancing the system throughput and minimizing the inter-cell interference as well as it can greatly increase the network flexibility via reducing the size of the resource granularity in a 4D resource allocation scheme.
Figure 2. Resource allocation for 5G and beyond; 3D/4D resource allocation enables by NOMA and MMW beamforming.
As we reported before, a heterogeneous VLC/ MMW system can benefit from the low-cost LED indoor illumination system, which provides accurate receiver location and helps to enable a fast, reconfigurable MMW beamforming link as shown in Fig. 1. A 3D VLC positioning via two-layer machine learning technique with the world record of positioning resolution of <1cm in a unit volume of 0.9x1x0.4 m3 was demonstrated for indoor applications . In addition, the spatial sparsity MMW together with narrow beam characteristic provides a greater spatial resolution, which facilitates the positioning estimation via time/angle of arrival (T/AoA) information in outdoor environments.
Given the improved positioning accuracy, it can enable a high-speed data transmission allowing operators to provide new network services. In the receiver-end, a user element with beamforming receiver is necessary; it should be transparent to signal waveforms, possess widely and continuously steering angles and fast beam-tracking capability. To meet the above requirement, a 2 GHz autonomous passive beamforming array with full field-of-view (±90o) reception angles was firstly introduced in 5G fixed-wireless access network . 9-Gb/s data rate was achieved after 25-km fiber transmission with the world record of beam-tracking time of <3 ms.
The driving force for realizing 5G mobile data networks is based on the diverse user scenarios, and the killer application of the first phase of 5G is high-speed and high-reliable and low-latency wireless data delivery. To fulfill such demanding requirements, emerging fiber-wireless integrated access technologies of 5G-NR, NOMA, MMW and FSO hybrid wireless transmission, coordinated MMW relay, VLC positioning, and MMW beamforming was demonstrated with proof-of-concept experiments and reviewed in this paper. The next opportunities for 5G era and beyond may align with seamless migration from legacy 3G/4G infrastructure to provide advanced 5G application and toward AI/ML-based fiber-wireless access to implement optimized and cognitive networking. In FiWIN research center, we tackle the real-world problems and address urgently demanding service requirements via combining the advantages of optical and wireless access technologies for heterogeneous mobile data communications.
- CableLabs, Cable Broadband Technology Gigabit Evolution, Fall, 2016.
- Li, J. Yu, L. Zhao, K. Wang, W. Zhou, and J. Xiao, “1-Tb/s photonics aided vector millimeter-wave signal wireless delivery at D-band,” in Proc. OFC, San Diego, CA, USA, 2018, paper Th4D.1
- Lu, M. Xu, L. Cheng, J. Wang, and G.-K. Chang, "Power division non-orthogonal multiple access (NOMA) in flexible optical access with synchronized downlink/ asynchronous uplink," J. Lightw. Technol., vol. 35, no. 19, pp. 4145-4152, 2017.
- Zhou, S. Shen, Y.-W. Chen, J. He and G.-K. Chang, “Efficient Power-Division NOMA for Intelligent Optical Access Network Enabled by Deep Learning,” in Proc. IEEE SUM, 2019, paper WB3.3.
- Zhang, J. Wang, Y. Xu, M. Xu, F. Lu, L. Cheng, J. Yu, and G.-K. Chang, “Fiber–wireless integrated mobile backhaul network based on a hybrid millimeter-wave and free-space-optics architecture with an adaptive diversity combining technique,” Opt. Lett., vol. 41, no. 9, pp. 1909-1912, 2016.
- Cheng, F. Lu, J. Wang, M. Xu, S. Shen, and G.-K. Chang, “Millimeter-Wave Radio Bundling for Reliable Transmission in Multi-Section Fiber-Wireless Mobile Fronthaul,” in Proc. OFC, Los Angeles, CA, USA, 2017, paper W1C.2.
- Liu, P.-C. Peng, C.-W. Hsu, J. He, H. Tian, and G.-K. Chang, “An Artificial Neural Network MIMO Demultiplexer for Small-Cell MM-Wave RoF Coordinated Multi-Point Transmission System,” in Proc. ECOC, Rome, Italy, 2018, pp. 1-3.
- -Y. Huang, T. Chi, F. Wang, T.-W. Li, and H. Wang, “A 23-to-30GHz hybrid beamforming MIMO receiver array with closed-loop multistage front-end beamformers for full-FoV dynamic and autonomous unknown signal tracking and blocker rejection,” in Proc. ISSCC, San Francisco, CA, USA, 2018, pp. 68-70.
- He, C.-W. Hsu, Q. Zhou, M. Tang, S. Fu, D. Liu, L. Deng, and G.-K. Chang, in Proc. OFC, San Diego, CA, USA, 2019, paper Th3I.2.
- Tang, M.-Y. Huang, Y.-W. Chen, P.-C. Peng, H. Wang, and G.-K. Chang, “A 4-channel Beamformer for 9-Gb/s MMW 5G Fixed wireless Access over 25-km SMF with Bit-loading OFDM” in Proc. OFC, San Diego, CA, USA, 2019, paper W3J.7.
Gee-Kung Chang is the Georgia Research Alliance and Byers Eminent Scholar Chair Professor in Optical Networking at the School of Electrical and computer Engineering of Georgia Institute of Technology. He is currently the Director of the NSF supported I/UCRC for Fiber-Wireless Integration and Networking (FiWIN). He received his B.S. degree in Physics from the National Tsing Hua University in Taiwan and his PhD degree from University of California, Riverside. Dr. Chang devoted 23 years of services in R&D at the Bell Labs, Bellcore, and Telcordia Technologies. Prior to joining Georgia Tech, he served as Vice President and Chief Technology Strategist of OpNext, Inc. Dr. Chang has co-authored more than 600 papers on peer-reviewed professional journals and international conferences. He is a Fellow of both IEEE and OSA.
You-Wei Chen was born in 1988. He received his Ph.D. degree from the Institute of Photonics Technologies, National Tsing Hua University, Hsinchu, Taiwan, in 2016. He is now a postdoctoral researcher and a lab manager at the School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, US. His research interests include but not limited to orthogonal frequency division multiplexing (OFDM), wavelength division multiplexing (WDM), passive optical networks (PON), 5G fiber-wireless integration networks and millimeter-wave (MMW) radio-over-fiber systems.
Jeff Finkelstein is the Executive Director of Advanced Technology at Cox Communications in Atlanta, Georgia. He has been a key member of the engineering organization at Cox since 2002 and led the team responsible for the deployment of DOCSIS technologies, from DOCSIS 1.0 to DOCSIS 3.0. He was the initial innovator of advanced technologies including Pro-Active Network Maintenance, Active Queue Management, DOCSIS 3.1, DOCSIS 4.0, and the Flexible MAC Architecture. His current responsibilities include defining the future cable network vision at Cox. Jeff has over 50 patents issued or pending.
Editor: Rod Waterhouse
Rod Waterhouse received his BEng, MS, and PhD in Electrical Engineering from the University of Queensland, Australia, in 1987, 1989 and 1994, respectively. In 1994 he joined RMIT University as a lecturer, became a Senior Lecturer in 1997 and an Associate Professor in 2002. From 2001 – 2003 Dr Waterhouse was with the venture-backed Dorsal Networks which was later acquired by Corvis Corporation. In 2004 he co-founded Pharad, an antenna and high performance RF-over-fiber technologies company of which he is the Chief Technology Officer. From 2003 he was also appointed as a Senior Fellow within the Department of Electrical and Electronic Engineering at the University of Melbourne. Dr Waterhouse’s research interests include antennas, electromagnetics and microwave photonics engineering. He has over 290 publications in these fields, including 2 books and 4 book chapters. Dr Waterhouse received the grade of IEEE Fellow for his work on printed antenna and microwave photonic technologies. Dr Waterhouse was an Associate Editor for IEEE AP Transactions during 2003 – 2009 and he is the Member at Large for IEEE Int. Topical Meeting on Microwave Photonics since 2016. Dr Waterhouse has been on the IEEE Photonics Society Fellow Evaluation Committee since 2013 and was a Representative of the IEEE Photonics Society for the National Photonics Initiative in 2015. In 2000, IEEE Third Millennium Medal for ‘Outstanding Achievements and Contributions’.
Subscribe to Tech Focus
Join our IEEE Future Networks Technical Community and receive IEEE Future NetworksTech Focus delivered to your email.
Article Contributions Welcome
Rod Waterhouse, Editor-in-Chief
Mithun Mukherjee, Managing Editor
Imran Shafique Ansari
Zhi Ning Chen