Alan Gatherer, CTO, Baseband System on Chip, Huawei and Anthony Soong, Chief Scientist, US Wireless Research and Standards, Huawei
Published: 13 Jun 2017
CTN Issue: June 2017
We thought it was about time for an update on the 5G standard progress. 5G is getting close to prime time which will occur perhaps as early as 2019 with the US leading the charge with fixed wireless to the home being a big initial application for 5G. In this article we give a brief summary of the features that have settled in to the 5G standard and also present a few demo and trial results. All the 5G nerds out there are welcome to write in and tell us what we missed.
Alan Gatherer, Editor-in-Chief
5G Standardization and Demonstration Update
CTO, Baseband System on Chip, Huawei
Chief Scientist, US Wireless Research and Standards, Huawei
General Progress and Timeline
The overall performance requirements of 5G are now quite settled and is shown in the usual spider diagram format in Figure 1 along with the performance of LTE evolution and IMT-Advance for comparison. It can be seen that LTE evolution tracks the performance of 5G-NR pretty well except in the areas of network efficiency and mobility and this reflects a desire to see most of the features of 5G as soon as possible in an evolved LTE form. Indeed 3GPP has decided that both LTE evolution and NR are 5G technologies and can be closely integrated. We will touch on this later.
The timeframe for standardization is shown in Figure 2. Note that we have recently passed the completion of the New Radio Study Item (NR SI) and are heading towards some "freezes" in the standard that will allow the Release 15 of the standard to occur in the second half of 2018. This should allow products to appear in limited trials in 2019 and 2020. Release 16 should then be in a stage 3 freeze by the end of 2019.
The Air Interface Improvements
New Radio is the name given to the part of the 5G standard that redefines the air interface for the mmWave bands. There is also a sub 6GHz band air interface that will be more heavily used in the early phases of 5G productization.
In order to have backward compatibility with 4G and also because it is a good idea generally, the 5G air interface continues to use OFDM in the down link and DFT spread OFDM in the uplink. OFDM in the uplink is now also supported. Additionally FDM and TDM multiplexing of different numerologies are allowed in a single frame so that a single user can get a very flexible allocation of resources in both time and frequency as shown in Figure 3 . Tight coupling of LTE AND NR are obtained by allowing 5G NR and LTE to coexist; whether that is in a FDM and/or TDM multiplexing is under consideration the standard.
This allows an efficient combination of the 3 main services for of 5G, enhanced Mobile Broadband (eMBB), Ultra Reliable Low latency (URLLC) and Massive Machine Type (MMTC) communications.
New additions to the Forward Error Correction options are Low Density Parity Check (LDPC) codes and Polar Codes. LDPC codes were added because they are seen as having lower latency decoding implementations than turbo codes and hence were good for low latency communications. They also have a lower error floor than Turbo codes. Polar Codes are a good solution for short block lengths and were therefore added to improve the control channel.
Multi Input Multi Output (MIMO) antenna arrays are a critical technology to achieve many of the gains in 5G but can have high hardware complexity. So the standard allows for a variety of options from no MIMO through beamforming and into massive MIMO and multiuser MIMO as shown in Figure 4. The Channel State Information Reference Signal (CSI-RS) and SRS are supported for the estimation of the channel. CSI reporting uses an LTE like and an enhanced version that allows explicit feedback and/or codebook-based feedback with higher spatial resolution are under discussion in the 3GPP.
Significant gain is achieved with even modest amounts of MIMO as shown in Figure 5. Note that the standards are discussing a probing-based link adaptation which allow for better estimation of the interference seen by the signal. Additionally, full duplex is also supported in the standard, by allowing the base station to assign the same radio resource for simultaneous uplink and downlink transmission. Site division duplex, which allows separate downlink and uplink “cell” association on an unpaired spectrum, provides additional advance duplexing.
Network Function Improvements
Outside of the air interface, the 5G standard body has also spent a lot of effort to bring the cellular system into a more virtualized, cloud like environment, supporting data connectivity and services enabling deployments using techniques such as e.g. Network Function Virtualization and Software Defined Networking. Beside the traditional reference point view of the system, 3GPP will also define a system service based architecture (Figure 6). The intension is for Network Functions (NFs) in the Control Plane to exhibit their functionality via service-based interfaces, so that the NF services can be flexibly used by other authorized NFs. This will allow the system designer, to flexibly combine network functions/sub-function into network slices that will provide a customized network to support services. This functionality is very important because it allows flexible application of resources to the services, such as eMBB, URLLC and MMTC, legacy LTE support as well as new vertical services, with the needed isolation. The standard is currently discussing the mechanisms for the life cycle management of the slices. On the RAN side, a new node, gNB, supports the NR as well as connectivity to Next Generation Core (NGC). 3GPP is investigating the logical functional split of the gNB RAN functions into two logical nodes, centralized unit (CU) and distributed unit (DU). A new interface between the CU and DU, F1, is, also, under current discussion. This will allow for a scalable cost effective RAN solutions with service agility to support the myriad 5G applications of the future.
Multiple Companies have demonstrated trial systems but for this brief summary we will stick to the ones we know about and summarize the Huawei results so far. We started our 5G trial process in the Chinese city of Chengdu in cooperation with NTT DoCoMo. As can be seen in Figure 7, 24 test UEs had a good time sitting under their umbrellas while communicating with one basestation over 100MHz channel in the sub 6GHz band. The demonstration successfully showed 0.5-2dB of gain for polar codes compared to turbo codes as well as 24 layers of MIMO giving a 3.6Gbps Peak Rate. From there we moved onto a more sophisticated demonstration in Tokyo with both stationary and moving UEs showing more than 11 Gbps average cell rate over a 200MHz band and less than 0.5ms of user plane delay. We have since worked with Deutsche Telekom on a mmW demonstration of 70Gbps in a 2GHz band with up to 24Gbps per user using advanced acquisition and bean tracking algorithms. With Vodafone we demonstrated 5Gbps over 1GHz BW at 300m distance and 3Gbps at 2km distance as well as inter-site handover.
5G specification is on track to allow productization to begin in 2018. The standard has defined an air interface that can achieve significant performance improvements as well as new network architectures that allow agile deployment and coexistence of different wireless services as well as coexistence with existing LTE. Trails have shown the feasibility of the performance goals and it now remains to be seen how rapidly the cost can be driven down and the new services deployed. The focus is now shifting from standardization to hardware implementation as well as new business opportunities for this technology. It is also clear that 5G will coexist with LTE much more closely than previous standard have coexisted with legacy.
Statements and opinions given in a work published by the IEEE or the IEEE Communications Society are the expressions of the author(s). Responsibility for the content of published articles rests upon the authors(s), not IEEE nor the IEEE Communications Society.