You are on page 1of 41

I M T- 2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p
5G Wireless Technology Architecture White Paper















IMT-2020 (5G) Promotion Group was jointly established in February 2013 by three ministries of China (including MIIT, NDRC, and
MOST) based on the original IMT-Advanced Promotion Group. The members include the main operators, vendors, universities, and
research institutes in China. The Promotion Group is the major platform to promote 5G technology research in China and to facilitate
international communication and cooperation.

I M T- 2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p
5G Wireless Technology Architecture White Paper

Over the past three decades, mobile communication

multiple access (PDMA), and multi-user shared

has experienced an extraordinarily high-speed

access (MUSA). Further, massive MIMO, ultra-

development, from simple voice systems to

dense network (UDN) and all-spectrum access (ASA)

mobile broadband multimedia systems. It not only

are identified as the key enabling technologies of 5G.

profoundly changes the life style of everybody, but

Some other technologies, including new multi-carrier

also greatly promotes the advancement of society

waveforms, flexible duplex, advanced modulation

and economy. Mobile internet and Internet of Thing

and coding, device-to-device (D2D) communication,

(IoT) are two major driving forces of future mobile

and full duplex, are also considered to be candidate

communications, providing a broader development

technologies. 5G will be built on the technical

space for the fifth generation mobile communication

framework enabled by novel multiple access, massive

system (5G). Towards year 2020 and beyond, 1000x

MIMO, UDN, ASA, etc., to fully meet the technical

mobile traffic growth, 100 billion connected devices,

requirements for 5G in 2020 and beyond.

and more diverse service requirements bring great

Currently, the global consensus of vision and

challenges to the system design of 5G. Compared

requirements of 5G is basically reached, and 5G

with 4G, 5G needs to support more various scenarios,

concept and technology roadmap are gradually

integrate with other wireless access technologies,

clarified. As the international standardization work

and fully utilize both low-frequency and high-

of 5G will start in the near future, it is urgent to

frequency bands. Furthermore, 5G should be

elaborate globally agreed 5G technical roadmap and

capable of supporting flexible deployment and high-

wireless technology architecture integrating various

efficient operation & maintenance, and significantly

key enabling technologies, to guide the international

improving spectrum efficiency, energy efficiency, and

standardization and industrialization of 5G.

cost efficiency, to realize the sustainable development
of mobile communication networks.
The traditional mobile communication evolution
is based on the revolution of multiple access
technologies. The innovation of 5G wireless
technologies is more diverse. Several novel multiple
access technologies have been proposed, including
sparse code multiple access (SCMA), pattern division


I M T- 2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p
5G Wireless Technology Architecture White Paper

Scenarios and Technical Requirements
Compared with previous generations of mobile

• Low-latency high-reliability scenario mainly

communications, 5G needs to meet extremely

targets special application requirements of

high performance requirements in more diverse

IoT and vertical industries such as Internet

scenarios. Four typical technical scenarios of 5G are

of Vehicles (IOV) and industrial control. The

derived from the main application scenarios, service

ms-level end-to-end latency or/and nearly

requirements, and key challenges of mobile internet

100% reliability need to be guaranteed in this

and IoT. They are seamless wide-area coverage


scenario, high-capacity hot-spot scenario, low-power

• Low-power massive-connection scenario

massive-connection scenario, and low-latency high-

mainly targets sensoring and data collecting

reliability scenario.

use cases, such as environmental monitoring

• Seamless wide-area coverage is the basic

and intelligent agriculture. This scenario

scenario of mobile communications. The

is characterized by small data packets, low

main challenge for 5G is to provide more

power consumption, low cost, and massive

than 100 Mbps user experienced data rate

connections. Specifically, at least 1 million

with guaranteed service continuity anytime

connections per squared kilometer need to be

and anywhere, regardless of static state or


high-speed moving, and coverage center or

To sum up, the major technical challenges of 5G

coverage edge.

include 0.1 ~ 1 Gbps user experienced data rate,

• High-capacity hot-spot scenario mainly

tens of Gbps peak data rate, tens of Tbps/km2 traffic

targets local indoor and outdoor hot-spot

volume density, 1 million connections per squared

areas where ultra-high data rates should

kilometer, ms-level end-to-end latency, and 100 times

be provided to users and ultra-high traffic

energy efficiency improvement and cost reduction

volume density needs to be handled. The

per bit.

main challenges include 1 Gbps user
experienced data rate, tens of Gbps peak data
rate, and tens of Tbps/km 2 traffic volume


a unified 3 . a high-frequency new air low-frequency and high-frequency new AIs. evolution is complementary. The data rate. interface is needed to achieve ultra-high data rates As the de-facto unified standard of 4G.I M T. The new air interface has two branches: scenarios. As a result. 5G requires a brand new and high-mobility scenario. and latency can be improved to partly newly designed with the introduction of advanced meet the technical demands of 5G. In order frequency spectrum resources. a suitable wireless technology roadmap technologies to meet all the requirements of 5G. The dense network. to meet the demands of user experiences and massive which includes a new air interface (AI) and 4G connections in wide-coverage and high-mobility evolution AI. while some other low-frequency new air interface can efficiently advanced technologies including all-spectrum fulfill the performance requirements including user access and several novel multiple access schemes experienced data rate. connection density. and new waveforms. novel multiple With the constraint of the current 4G technical access. connection The low-frequency new air interface will be capability. 4G evolution is not and energy efficiency in most scenarios such as wide- capable of meeting the extremely high performance area coverage scenario. more simplified signaling procedure.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper 5G Wireless Technology Roadmap In order to meet the technical requirements of 5G compatibility and make full use of various advanced scenarios. technology trends. cannot be employed. the air interface of 4G frequency new air interface working below 6GHz evolution should adopt enhanced technologies based and a high-frequency new air interface working on LTE/LTE-Advanced framework in the legacy above 6GHz. requirements in future. A low-frequency new air interface of 5G is required the wireless technology roadmap can be derived. With consideration of 5G requirements. system capacity. the performance of some enhanced structure. and smooth network evolution. Although different air interface to break the restriction of backward scenarios have very different requirements. spectrum. massive-connection scenario. LTE/LTE- and system capacity by utilizing abundant high- Advanced has been deployed worldwide. 4G should be further enhanced. The is needed to guide the development of 5G standards new air interface is the main body of 5G. and technologies. technologies such as massive MIMO. The shorter frame framework. is limited. different frequency bands. Moreover. while 4G and industry. such as massive MIMO and ultra- more flexible duplex modes will be supported. 5G should include a low- With backward compatibility. latency. According to global to improve user experience continuously and ensure spectrum planning and propagation characteristics on smooth evolution.

where the operation and maintenance. the low-frequency air including waveform. With parameters technical components and parameters can be flexibly adjustment. modulation. and simplify channel condition is good.I M T. and basic data transmission. In such cases. while the candidate high-frequency bands in a wide range. interface offers seamless coverage. control plane should be optimized. it can adapt to the characteristics of high- configured to adapt to different scenarios. the air interface technical solution. interface. high-frequency air interface is supplementary and a unified technical solution should be designed to provides high data rates in hotspot areas when the facilitate standardization. Figure 1 5G wireless technology roadmap 4 . and MIMO. needs to be integrated with the low-frequency air Correspondingly. Since there are a number of functions. reduce cost.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper technical solution needs to be designed. In the high-frequency new air interface. the high-frequency air to consider the impacts of high-frequency channel interface is unfit to stand-alone deployment. coding. and characteristics and radio frequency (RF) components. it is necessary Due to the poor coverage. frequency channels and RF components.

such as filter bank multicarrier 5 . and TDD is more suitable for high- main modules. However. Each module should cover time and frequency resources flexibly to the common technical contents as many as possible better adapt to non-uniform dynamic service to reach a good balance between flexibility and distribution. diverse performance requirements. it is possible modules are described as follows. Based on such services. and channel environments. and their enhancements. duplex. of standardization and industrialization. by flexibly configuring technical components and and other advanced technologies. flexibly configured to support the application an optimized technical solution can be obtained of massive MIMO. They are frame structure. Design Principle parameters of 5G frame structure can be The technical framework of 5G air interfaces should flexibly configured for diverse scenarios and be unified. The bandwidth. TDD. the technical framework of support new duplex technologies such as flexible duplex and full duplex. 5G should scenarios. and protocol. Full duplex is a potential duplex simplicity. For each scenario. frequency air interface. uplink-downlink configuration. According to the technical and receiving at the same time and frequency requirements of each scenario. modulation & coding. can be optimized and flexibly configured to form the • Waveform: In addition to traditional OFDM corresponding technical solution. and may 2. the technical modules resource.. flexible. and configurable.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper 5G Air Interface Technical Framework 1. flexible waveform. subcarrier spacing. it is preferred cyclic prefix (CP). In addition.I M T. for each scenario. to adopt new waveforms based on optimized • Frame structure & channelization: The filtering. etc. transmission time interval to design a specially optimized technical solution (TTI). from the perspective can be adapted to different frequency bands. The have a unified technical framework for the new reference signals and control channels can be air interface and 4G evolution. Technical Framework Based on the main functions of mobile communication systems. Different modules can interconnect technology which can support transmitting and work together. novel multiple access. The main technical and single-carrier waveforms. Lowfrequency air interface will adopt FDD and 5G air interfaces can be divided into the following TDD. • Duplex: 5G will support traditional FDD. multiple access. duplex can allocate uplink and downlink antenna. parameters.

2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Figure 2 Flexible and configurable AI technical framework of 5G (FBMC). waveforms can improve spectrum utilization Moreover.I M T. multiple waveforms can coexist. Therefore. access. and MUSA. filtered-OFDM (F-OFDM). • Modulation & coding: 5G needs to support various services such as high data rate. and peak-toaverage power ratio (PAPR). connections and improve spectrum efficiency With very low out-of-band leakage. efficiently utilize fragmented significantly reduced by grant-free contention spectrum. the air interface latency can be efficiency. 6 further improve link spectrum efficiency. suitable waveforms can be selected to adapt to the requirements of different scenarios. low latency. low data rate. For high-data-rate services. and new constellation mapping. polar codes. those via multi-user superposition transmission. and high reliability. overhead. small packet. which can support more universal filtered multicarrier (UFMC). M-ary low density parity check (LDPC). but also some novel multiple access schemes including SCMA. and coexist with other waveforms. and PDMA. Different waveforms have different outof-band leakages. and faster-than Nyquist (FTN) can • Multiple access: 5G will support not only traditional OFDMA. compared with traditional binary Turbo codes and QAM. In some cases. For low-data-rate and .

As a Based on the air interface technical framework result. service requirements to realize “forward compatibility”. 7 . The low-frequency new air interface can be widely link adaption. the coding schemes with WLAN. As a be a number of wireless backhaul links in result. beamforming. The used in various scenarios including seamless wide-area protocol is capable of being configured coverage. For high-reliability services. Massive MIMO and parameters can be selected to form a technical can also be used in high-frequency bands solution for specific performance requirements. which mechanism to meet low power consumption can enable advanced multi-user MIMO to requirements of IoT services. to overcome large path loss by adaptive available frequency bands. For low- interface. low-power massive- and optimized.I M T. In addition. Consideration on 5G Low-Frequency New AI needs to support various advanced scheduling. the free access protocol can be employed to “error floor” effect of decoding algorithms simplify signaling interaction procedure needs to be avoided. The adaptive hybrid automatic increased by network coding. channel capable of dealing with the future new scenarios and estimation. For the application of massive and implementation cost. channel information feedback. the access latency will be reduced dense networks. equipment capabilities. To reduce the signaling overhead fast encoding and decoding algorithms are of massive small-packet services. and latency services. The technical framework is MIMO. multi-user scheduling mechanism. the grant- preferred. and multi-connection. the reference signals. and efficiently supporting radio to approach Shannon capacity in the cases access technologies including the new air of short code length and low SNR. support more users in spatial domain. the system spectrum efficiency will be of 5G. 4G evolution air interface. and receiving algorithms need to be improved • Protocol: The air interface protocol of 5G 3. Each base station can be in latency and reliability. there may between base stations and terminals. polar codes and low- flexibly to meet the requirements of different code-rate convolutional codes can be used scenarios. In addition. an optimized combination of technologies improved by several times. the system capacity can be greatly. repeat request (HARQ) protocol of 5G • Antenna: Massive MIMO will be utilized should match different service requirements in 5G systems. high-capacity hot-spot.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper small-packet services. 5G equipped with more than one hundred should support more efficient energy-saving antennas and dozens of antenna ports.

macro cells can be further enhanced and the C/U split In seamless wide-area coverage scenario. and advanced of massive MIMO. and multi- between access and backhaul can improve frequency 8 . massive MIMO can without serious interferences. For the frame structure. and low-latency high-reliability. channel estimation. In order to reduce interferences 10 users can be supported. and the same waveform design adaptive small cell clustering. Furthermore. high-capacity hot-spot scenario.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper connections. Higher order modulation spectrum efficiency. and frequency resource coordination can new waveform technology like F-OFDM can be used be utilized. the traditional cell boundaries to support the coexistence with technical solutions can be broken. with the needs to be supported. The access and backhaul. so that multiple cells can co-work more for other scenarios. The OFDM waveform in densely deployed networks. TDD can technologies including massive MIMO. Adaptive frequency allocation reference signals. by increasing cell can achieve 100 Mbps user experienced data rate density. structure can be taken into account. To provide a flexible backhaul mode improved by introducing novel multiple access for small cells. For duplex technologies. The parameters of frame structure can be modified the frame length can be divided exactly by 1 ms and according to the channel and service characteristics of the subcarrier spacing can be a multiple of 15 kHz. novel multiple use channel reciprocity to obtain better performance access. This user pairing mechanism should be newly designed technical solution can be optimally designed based on for the application of massive MIMO and novel the unified 5G technical framework by integrating key multiple access. flexible duplex or full- be adopted in base stations to improve the system duplex may be introduced. spectrum efficiency and bandwidth. The transmission. it is data rates and traffic volume density to some extent.I M T. The number of antennas can be and coding schemes can be adopted to improve the up to 128. multi-cell cooperative can be utilized for both uplink and downlink. And the compatibility with LTE frame needs to be further optimized in the following aspects. the low-frequency new air interface In high-capacity hot-spot scenario. below 6 GHz. technologies including can still be adopted. it should be considered to jointly design technologies based on superposition coding. the low- through increasing bandwidth and improving frequency new air interface can meet the requirements of spectrum efficiency. the control plane of modulation & coding. and the parallel transmission of more than spectrum efficiency. in order to realize the efficient good channel propagation at the frequency bands control and management for small cells and users. The necessary to increase the subcarrier spacing and technical solution should be generally consistent with the shorten the frame length to better support wider solution of seamless wide-area coverage scenario and bandwidth. For example. new multi-carrier waveforms. Furthermore. The connection capability and efficiently to serve one subscriber to achieve user centric spectrum efficiency of OFDMA systems can be cell virtualization. In some environments For multi-antenna technology.

technologies such as advanced idle states need to be improved.I M T. modulation. to achieve "zero" F-OFDM and FBMC) can be used to lower out-of-band waiting time for uplink data packet scheduling. low power. Non-contiguous and should be improved. To meet the requirement of high reliability. In addition. In connected state. In terms of waveforms. Another solution is to adopt dynamic. and signaling process in different contents. PDMA. contention access can simplify the signaling process. reduce the power consumption. and shorten turn-on time. and MUSA can enable grant-free transmission. Multiple access technologies including SCMA. avoiding resource allocation process. should be decreased through minimizing forwarding In low-power massive-connection scenario. functions to access network and localizing service multiple access. such as SCMA. services featured with small data packet. In addition. Moreover. The coding. the IoT nodes. To reduce interferences. idle state. and the network forwarding delay design. the impacts of wireless In low-latency high-reliability scenario. and cost of terminals can be reduced. By supporting grant-free CP and CP sharing between multiple symbols can be transmission. the retransmission performance can reduce user access delay.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper utilization efficiency. By narrow- access network structure and support dynamic MESH band system design. clustered subbands can be configured independently. to achieve the networking modes including centralized. Since the shorter TTI may cause and MUSA. coding and space/time/frequency diversity can be utilized. one solution is to shorten bands efficiently. and the transmission reliability of single link require large channel bandwidth. longer paging intervals can make terminals technologies such as enhanced coordinated multiple sleep longer to reduce the power consumption of points (CoMP) and dynamic MESH can strengthen the terminals. single-hop and/or multi-hop direct capability can be improved. In terms of protocol. the air interface delay should be MESH networks should be taken into account in system reduced significantly. can support massive connections via multi- larger CP overhead. the new waveforms supporting non- user superposition transmission. the new efficiently-filtered waveform technologies (e.g. the fragmented low-frequency bands or a few OFDM frame length needs to be shortened and the compatibility subcarriers can be used for this scenario. To improve data energy-saving mechanisms for both connected and transmission reliability. and requirement of latency. cooperation between base stations and between terminals 9 . the to further reduce end-to-end latency. to utilize non-contiguous and fragmented the network forwarding delay. The novel with the solution for seamless wide-area coverage scenario multiple access technologies. distributed. and the power consumption communications among base stations or/and terminals. the system coverage and connection communication link. PDMA should be considered.. they can simplify the signaling process and utilized. the number of retransmission per unit time should be massive connections and burst transmission do not maximized. and to decouple technical solutions the transmission path through sinking some core network between different subband. To reduce the air interface delay. In be improved via enhanced HARQ mechanism.

Consideration on 4G Evolution AI frequency wireless signals. where OFDM scheduling and feedback mechanisms. for control plane functions. With the space multiplexing to support more users and increase guarantee of smooth evolution. the low-frequency air interface in support of C/U split. However. To support extra-large bandwidth. By using rich high-frequency interface of 4G evolution should be able to cooperate spectrum resource. In addition. Considering the requirements of high data rates and connection capability. the wireless self-backhaul of base tightly with the new air interface. traffic the subcarrier spacing can be increased over 10 times. Consideration on 5G High-Frequency New AI The high-frequency new air interface utilizes extra-large bandwidth to fulfill the requirement of extremely high transmission rate in high-capacity hot-spot scenario. Because of small coverage and strong directivity of high- 5. it can partly meet the system capacity. The 3D channel is still an important candidate. The air interface and the frame length can be shortened significantly in latency can be reduced by decreasing the OFDM comparison with LTE. and connection density. To meet the requirements and system capacity. To deal with the poor coverage. the narrow-band scheme can decoding and require less storage should be selected. through adaptive beamforming and tracking. and use MIMO. and serve users via stations can be realized through a jointly designed access advanced technologies like dual connectivity. such as data rate. channel codes that support fast of IoT applications. volume density. It should be further Massive MIMO can compensate the large path loss optimized in several aspects such as frame structure. The low-frequency air interface should be responsible 4. single carrier information can be utilized to achieve more accurate is also a potential waveform if the challenges caused beamforming. the high-frequency air interface needs to be integrated with 10 . improve coverage. and backhaul links. requirements of 5G. which can support more users and more by high-frequency RF components and propagation data streams. The same waveform design can symbols in each TTI and introducing the optimized be adopted in both uplink and downlink. the air speed backhaul links. while the high-frequency air interface can be used to achieve high data rates. Dual connectivity and dynamic load balancing can be realized in user planes of the two air interfaces. TDD is preferred to enhanced interference cancellation algorithms can be support high-frequency communications and massive employed to improve the system spectrum efficiency MIMO. LTE-Advanced technical framework. and multiple access technologies. and The high-frequency new air interface requires high- reduce power consumption and cost. latency. increase connection capability. Multi-user superposition transmission and characteristics are considered.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper to further enhance the data transmission reliability. dense deployment can be The air interface of 4G evolution will be based on LTE/ considered to achieve very high traffic volume density.I M T.

and suppress multi-user interferences.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper 5G Wireless Key Technologies 1. Thus. exceeds the number of terminal antennas. As a result. etc. Figure 3 Deployment scenarios of massive MIMO The major research areas of massive MIMO include: (2) Transmission and detection technologies (1) Deployment scenario and channel modeling The performance gain of massive MIMO systems is The potential deployment scenarios of massive mainly guaranteed by the orthogonality between the MIMO includes macro coverage. improve received power. field measurements According to the large-scale random matrix theory. The MIMO has been widely used in 4G systems. the base station with large number of antennas can form extreme sharp beams with high spatial resolution and beamforming gain. to be mutually orthogonal. On the other hand. In the above antennas is a major direction of MIMO enhancement. the random which can reflect the propagation characteristics in channel vectors between a base station and users tend 3D domain. indoor and outdoor hotspots number of antennas. which can provide flexible spatial multiplexing capability. In practical deployments. for effective suppression of co-channel 11 . are needed to model the distribution and correlation when the number of base station antennas greatly of a serial of parameters in massive MIMO channel. Massive MIMO the possible deployment scenarios as well. channel vectors of the users provided by the large heterogeneous network. much more users can be served on the same time-frequency resources. increasing the number of in hotspots and wireless backhaul. high-rise coverage. the significant array gain is able to effectively improve the signal-to-interference-plus-noise ratio (SINR) of individual users. typical deployment scenarios. In addition. In current frequency bands can be utilized in the macro order to meet the requirements of high data rates coverage. leading to the asymptotic zero interference among users. Under real channel conditions with and wireless backhaul. the much higher system capacity and spectrum efficiency can be achieved. while high-frequency bands can be adopted and system capacity of 5G. the massive non-ideal factors caused by equipment and propagation MIMO system with distributed antenna is one of environment.I M T.

transceiver mobility solution modules with high efficiency. the algorithms of cell. Factors that Kronecker product-based CSI measurement and have impact on the performance and efficiency of feedback. high reliability. and modularization. mobility should be taken into account in the massive In addition. Beam tracking and beam broadening for massive MIMO will be related with the antenna techniques. In this case. the hybrid digital- based CSI acquisition. power efficiency and are able to improve the transmission reliability and data system performance. product-based separate horizontal and vertical precoding (5) Multi-user scheduling and resource algorithm. as gain on coverage of the traffic channel. high- (4) Coverage enhancement techniques and high accuracy supervision and calibration. inner-outer dual-ring beam sweeping downlink transmission and uplink reception need to can be used to cope with the wide coverage problems be designed carefully. equipment cost. need to be studied. the precoding & beamforming algorithm MIMO systems. hybrid digital-analog beamforming. in into full-digital array and hybrid digital-analog addition to the schemes such as traditional codebook- array. the requirements of of transmission and detection algorithms are relevant reliability and high-data-rate transmission under high to the scale of antenna arrays and the number of users.I M T. However. or staged management beamforming will reduce computational complexity of Due to the high spatial resolution and much more large-scale antenna systems effectively.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper interference among users and cells. design are very important for the application of (6) Large-scale active antenna array MIMO technology. traffic CSI measurement. utilized in the high-frequency bands. 12 . feedback and reference signal load balancing and resource management. Moreover. low cost. which are less dependent on CSI acquisition. It is expected that the Kronecker rate by utilizing the array gain efficiently. Increasing the number of antennas will offer huge miniaturization. To better compromise between Large-scale active antenna array can be classified the overhead and the accuracy CSI acquisition. considerable performance and feedback gain can be obtained by multi-user scheduling. and pre-experienced massive MIMO in practical environments include feedback. compression sensing. array structure. the large-scale active antenna array structure. will cause some challenges for the broadcast channel These key factors will have great impact on the which requires uniform coverage for all the users in the practical deployment of massive MIMO. spatial freedom of the wireless access network (3) Channel state information (CSI) measurement offered by massive MIMO. it well as high-accuracy supervision and calibration. new CSI mechanisms analog structure will have great potential when including staged CSI measurement and feedback. The computational complexity of the narrow beams. based implicit feedback and channel reciprocity- power consumption and cost. Considering the implementation complexity.

and provides flexible dense urban areas. and apartment. The continuing in the heterogeneous and dense networks. To massive MIMO will surely play an important role in meet the requirements of typical scenarios and the 5G systems. a n d network include office. sites acquisition. the joint access and feedback. and realize easy network with easy deployment. transmission resource. open-air gathering. subway. transmission reliability. are important research areas in ultra- density in year 2020 and beyond. Ultra-dense network is a major technology to interference management and suppression. easy maintenance. Ultra-Dense Network user experience. With increase of cell density brings many new challenges the breakthrough of key techniques and further such as interference. and delightful 2. interference management and coordination schemes stadium. technical challenges.I M T. High spectral dense network. dense residential areas. This results in hundreds of times capacity improvement Figure 4 Key Technologies in UDN 13 . massive MIMO improves system in hot-spot areas. development of RF components and antennas. and cell meet the requirements of ultra-high traffic volume virtualization. mobility. campus. and deployment cost.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper In general. Typical scenarios of ultra-dense s p e c t r u m e ff i c i e n c y. u s e r e x p e r i e n c e . reuse factor can be achieved in ultra-dense network via densely deployed wireless equipment.

the small cells on the network which can support high rank MIMO and second stratum are connected to the base stations via reduce the interference. and cluster-based multi- resource allocation and dynamic resource sharing cell frequency resource coordination. joint efficiency. multi-cell coherent cooperative transmission optimization of backhaul and access links. multi-hop. which facilitates a plug- network topology and backhaul strategy to match the and-play way of network deployment. flexibly adjusting together in this architecture. there would be more overlap between cells. utilizing available network resource (including both Wired and wireless backhaul links are combined wired and wireless resource). The radio system capacity. leading There are two research topics in wireless self- to severe interferences. the same radio access strategy technology and frequency bands can be shared Ultra-dense network can effectively improve the between backhaul and access links. dynamic route selection. multi-route backhaul. and the small cells on efficient and economical solution to improve the the third stratum are connected to small cells on the backhaul capability of UDN. as the cell density increases. flexible backhaul is expected to meet the interference management and resource coordination end-to-end quality-of-service requirement with lower for backhaul and access links. The first resulting in improved throughput. and flexibly Multi-hop and multi-route backhaul technique deals allocating the network resource in both backhaul and with route selection.Flexible backhaul is an one-hop wireless transmission. Through flexibly second stratum via one-hop wireless transmission. resource can be partitioned in time or frequency. wireless exploited to achieve transmission and reception of self-backhaul and flexible backhaul. independent data streams in spatial sub-channels. The strategy of interference backhaul: 1) joint optimization of backhaul and management and suppression includes adaptive access links. 2) enhanced backhaul link.I M T. However. adaptive based on centralized control. In enhanced Joint design of access and backhaul links contains backhaul link aspect. network resource and service payload. establishment and management access link to improve the end-to-end transmission of multiple routes. the broadcast and multiple several key techniques. provide significant capacity gains.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper (1) Joint design of access and feedback the overall utilization of resources. base thus offering more spatial degrees of freedom and stations are categorized by stratum indices. (2) Interference management and suppression With wireless self-backhaul. In joint clustering. such as hierarchical access nature of wireless communication can be backhaul. By turning on/ between backhaul and access links can improve off small cells at subframe level. In the hierarchical backhaul architecture. adaptive clustering 14 . which is expected to deployment and operating cost. Multiple nodes stratum includes macro cell layer and small cell layer or terminals can cooperate to form a virtual MIMO connected with wired backhaul.

2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper can form dynamic cell clusters and terminate cells and transmission. equipment 15 . Soft sector technology can provide that are not serving users or in need for extra capacity. Cluster-based multicell frequency resource coordination can improve cell edge performance by optimal partitioning of frequency resource between dense clusters. For soft sector technology. which can improve edge user experienced rates. Virtual layer technology is a solution that relies on virtual layer and real layer networks. (3) Cell virtualization Cell virtualization includes user centric virtual cell. Driven by the coverage and services. User Figure 5 Virtual layer technology centric virtual cell has the goal of edgeless network.I M T. while data transmission is carried on the real layer. a unified management platform between soft sectors thus reducing the interference to neighboring cells. multiple soft sectors are formed by means Figure 6 Soft sector of multiple beams generated by the centralized entity. e. virtual layer technology. paging and mobility management. the virtual layer is responsible for broadcasting.g. It can significantly reduce the cost of site rental. regardless the location of the end user. Multi-cell coherent cooperative transmission enables It is a light solution with easy deployment and easy joint transmission over appropriate set of neighboring maintenance. cells. Wherein. resulting in system throughput enhancement. virtual cell keeps updating with the movement of users and maintains a good link quality and QoS/QoE with each terminal. utilizing the same frequency bands in each cluster and different frequency bands in different clusters. terminals can combine the signals from multiple cells without interference. and soft sector technology. and reduce the operation complexity. thus the user experience can be delightful. and real sectors. Therefore. No cell reselection or handover is needed for the mobile users within the same virtual layer.

To fully exploit the application in cellular communication is just in a advantages of different frequencies. between high-frequency bands and traditional low- for example. and RF for 5G is from 6 GHz to 100 GHz. Nowadays. new been widely recognized that the high-frequency range air interface design. ultra-high data rates. High-frequency signals are susceptible supports hybrid network integrating low-frequency to environmental effects such as obstruction. the research on all-spectrum access variation. it has frequency channel measurement and modeling. and high-frequency bands to simultaneously meet the refraction and air absorption in mobility. networking technology. the ultra-high bands below 6 GHz and high-frequency (HF) bands capacity and ultra-high data rates required by 5G can be above 6 GHz. it is mainly focuses on key technologies involving high- necessary to conduct comprehensive research on high- frequency mobile communications. Considering the big differences obviously different from conventional cellular channels. strong propagation path loss. As a result. Hence. fast channel frequency bands. Figure7 Key Technologies of HF Communication 16 . makes high-frequency channels have some properties and ultra-high capacity. low diffraction ability. and the latter High-frequency communication has been implemented is the supplementary bands that are used to achieve in military communications and WLAN. where the former is the core bands of achieved. reflection.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper 3. 5G that are used for seamless coverage. which can provide components. which requirements of seamless coverage.I M T. all-spectrum access beginning stage. All-Spectrum Access abundant available spectrum to obtain more than 10 All-spectrum access involves low-frequency (LF) Gbps user transmission rates. but its high data rates in hotpot areas. etc.

symmetric and asymmetric bands. The main candidate bands for bands. containing authorized and training and tracking for larger coverage. small coverage. the large traffic volume and high data rates in widely accepted path loss models: close-in reference and some scenarios like hotspots can be achieved with low floating intercept. 5G focused by the industry include 6 GHz. especially for transmission and outdoor hotspots to provide high data rates. the high-frequency Compared with low and medium frequencies. Furthermore. the propagation characteristics of high frequency signals. and high-frequency adaptive on channel transmission loss can be obtained by two backhaul. The parameters ultra-dense networking. 45 GHz. high-frequency channel modeling. 38 GHz. (2) HF air interface design (4) HF components Based on the channel characteristics. the feature of high-frequency channels is its dependence high-frequency air interface is mainly deployed as indoor on antenna configuration. delay spread. Therefore. Due to loss. topics also include the signal waveforms for different contiguous and non-contiguous bands. the high-frequency communications can while the floating intercept model is more reasonable in be employed in some new wireless application scenarios the case of sufficient measured data. it is interesting to study the adaptive spectrum characteristics. 18 burst transmission. 60 GHz. and 72 GHz. interference management. A new As the supplement to the low-frequency air interface.I M T. modulation and coding schemes. derive the channel models. 28 GHz. the close-in reference deployment cost. MAC technologies. In work in this area is to study the channel propagation addition. analyze and sensing techniques supporting multiple frequency usage evaluate the suitable scenarios. With the features of narrow beams and model is more robust if the measured data is insufficient. the frequency bands of interest ranges arrays related techniques as well as narrow beam from 6 GHz to 100 GHz. efficient GHz. frequency air interface. The preliminary channel measurements indicate higher (3) LF and HF hybrid networking frequencies come with larger propagation loss. etc. and joint access and backhaul The measuring scenarios include indoor and outdoor are studied to improve the performance of the high- hotspots. and of the potential high-frequency candidates and the the unified frame structure and access mechanism for existing industrial channel measured results. The research unauthorized bands. high 17 . With C/U splitting. including D2D and vehicle radar. received power and AoA/AoD. On the basis high-frequency channel propagation properties. how to decouple channel measurements hybrid networks integrating low-frequency and high- with antenna configuration is a research focus of frequency air interfaces are required. 15 GHz. By comparison. and select the suitable modes such as authorized and unauthorized frequency high-frequency bands.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper (1) HF candidate spectrums and channel air interface research focuses on transmitting and characteristics receiving beamforming with multi-antennas and antenna Currently. the research low-frequency and high-frequency hybrid networks.

The feasibility of using high-frequency technology based on the code domain superposition. and utilizing advanced algorithms to recover multi- To verify these key technologies. Both high- high requirements on the simplification of system design frequency power amplifier and low-noise amplifier and signaling process. substitution. As a result of non-orthogonal sparse its standardization and industrialization.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper frequency devices are more convenient for massive 4. modulation technology. the mobile industry user information at the receiver. and RF components. devices need to support the sampling frequencies Novel multiple access. of at least 1GHz channel bandwidth. Novel Multiple Access antenna integration and device miniaturization. further code superpositioning. 23 GHz. based on assigned codebooks. can significantly improve the spectrum high-frequency array antenna is required to realize efficiency and access capability by superimposing multi- high-gain beamforming and wide-range spatial users information in the same transmission resources scanning. (1) Sparse code multiple access and a 20-40 GHz prototype system can support 10 Gbps SCMA is a novel non-orthogonal multiple access transmission rate. To facilitate decode the data. through multi-dimensional modulation and spread 18 . a further breakthrough is required connections of massive devices. Moreover. represented by SCMA. Furthermore. high-frequency communication can through conjugation. high- resource. and improve the quality of single-user link frequency networking. communication to support extremely high data rates has This technology combines the low-density code and been preliminarily verified. an E-band prototype process and reduce the air interface transmission latency. Towards year 2020 and beyond. 6-100 GHz devices are relatively mature. mobile communications. At the receiver side. 5G should not only improve especially for 14 GHz. Among them. air interface technical solution. but also support the E-band. The ADC and DAC access technology. significantly improve data rates and system capacity and makes different users transmit information using abundant high-frequency spectrum resources. However. V-band. PDMA. The new type and MUSA. and phase rotation. system can realize the transmission rate up to 115 Gbps. SCMA can support more users breakthroughs are needed in channel propagation to connect simultaneously with the similar spectrum properties. 28 GHz. selects the optimal codes In brief. there are to apply them in cellular communications. grant-free has developed some high-frequency communication transmission can be enabled to simplify the signaling prototype systems. All of these requirements make need to be further improved in aspects of power great challenges for the existing orthogonal multiple efficiency and phase noise. and the spectrum efficiency significantly.I M T. mobile internet and IoT Benefit from the commercialization of microwave will become the main forces driving the evolution of products. It is regarded as a breakthrough technology for the message passing algorithm (MPA) is utilized to conventional cellular communications.

Figure 9 The principle of PDMA (3) Multi-user shared access MUSA is a non-orthogonal multiple access Figure 8 The principle of SCMA scheme base on code domain superposition. and correlation property is required when the code power domains. PDMA and then are transmitted on the same resources.I M T. using user characteristic length is short (e. For the uplink.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper spectrum. utilizes pattern segmentation to separate user signals The SIC algorithm is utilized to decode the user at the transmitter and serial interference cancellation at data at the receiver. Moreover.. For the downlink. pattern segmentation based on the conventional power superposition technology can contribute to better distinguish scheme.g. receiver. and therefore improve the detection optimize the symbol mapping of paired users and performance of serial interference cancellation at the to improve the downlink performance. 4 or 8). This can contribute to the reduction of implementation complexity and latency. and low cost. The maximum capacity of multiple access sequence is the key element that impacts the channels can be approached. and therefore is suitable for IoT services with small-packet transmission. pattern at the transmitter. In addition. grant-free random access can be achieved through blind detection techniques and the collision insensitivity of SCMA codewords. the modulated symbols of different (2) Pattern division multiple access users are spreaded by specific spreading sequence. The design of spreading the receiver. Based on multi-user information theory. The user pattern can be performance of MUSA scheme. MUSA utilizes mirror constellation to different users. code. Good cross- designed independently or jointly in space. 19 . low power.

and massive can be utilized to simplify the signaling process and connections. and therefore can be one of the However.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Figure 10 The principle of MUSA Based on simulation analysis.. Compared with OFDM. and easy combination with MIMO access technologies above mentioned can not only transmission. connection density. about 30% improvements on downlink OFDM has been used extensively in 4G. Some new multi-carrier waveforms reduce the data transmission latency. Furthermore.I M T. grant-free transmission services. OFDM has some drawbacks. as well as flexible frequency- improve the spectrum efficiency and support more selective scheduling. high fundamental core technologies of 5G. can be introduced to assist OFDM to better meet User experienced data rate. New Multi-Carrier Waveforms systems. However. it is insufficient be achieved by introducing novel multiple access to only use OFDM to deal with the diversified technology. and the entire requirements of 5G. higher spectrum efficiency. but also be capable to reduce the support mobile broadband (MBB) services. sensitivity to time- 20 . OFDM can efficiently user connections. e. the novel multiple channel. latency are the top three key performance indicators With high robustness against multi-path fading of 5G.g. and spectrum efficiency and more than 3 times is still considered as an important candidate improvements on uplink user connectivity could waveform for 5G. compared with LTE 5. latency significantly. out-of-band emission (OOBE).

compatibility with other technologies such as (1) Support flexible and configurable new air new modulation and coding schemes and new interface based on new multi-carrier. The main difference is that F-OFDM uses longer 4G mainly focus on the MBB services. A common feature of these certain service type with the most suitable new waveforms is that filters are employed to waveform numerology.I M T. some key requirements on 5G can be of 5G. the new multi-carrier implemented at the granularity of each subband. waveform needs to better support new service. new multi-carrier waveforms turn allows for larger freedom in adaptation to s h o u l d h a v e g o o d s c a l a b i l i t y. thus to To b e t t e r s u p p o r t t h e v a r i o u s a p p l i c a t i o n avoid the above-mentioned limitations of scenarios of 5G. filters. such as by subband-based filtering.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper frequency synchronization error. and the signal processing procedure is the 5G will embrace diversified types of services. however. several new multi- numerology configuration in different subbands carrier waveforms have been proposed. on time-frequency synchronization. T h e n e w channel variations. 21 . Secondly.With these new multi-carrier Finally. which in continuously. In other words. in and vehicle-to-vehicle (V2V) communication. the new waveforms period. Moreover. order to maintain the time-domain inter-symbol which propose new requirements to these orthogonality. in order to meet diverse requirements waveforms. same as the conventional OFDM in each subband. and the same suppress the OOBE and relax the requirement waveform numerology for the whole bandwidth. new multi-carrier should achieve good satisfied. such as TTI duration. and UFMC can support independent waveform To meet these requirements. CP is not required technologies and services can be supported by by FBMC. and therefore the signaling overhead simply parameter configuration or modification. UFMC uses shorter filters and basis waveform. can be reduced. In contrary. The filter in FBMC is implemented at w i l l h a v e t o p r o v i d e e ff i c i e n t s u p p o r t f o r the granularity of each subcarrier. besides the the CP of OFDM is replaced with empty guard conventional MBB services. FBMC gives machine-type communications. waveforms should identify the requirements The filters in UFMC and F-OFDM are of each scenario. universal filtered service types. Firstly. F-OFDM multiple access schemes. each for a carrier (FBMC). such as the machine-type communications (MTC) providing backward compatibility. the research of new multi-carrier OFDM. due up the orthogonality in complex domain for to new technologies and new services emerging better time-frequency localization. the system bandwidth can be multi-carrier (UFMC) and filter-bank multi- divided into several subbands. Based on different filtered-OFDM (F-OFDM).

so that the significantly vary between devices because of synchronization overhead can be reduced.e.. FBMC can better adapt to the characteristics of doubly selective channels by optimizing the prototype filters using real-time channel state information to achieve a better performance.I M T. In this way. suppressed by filtering. the requirement on the channels may be doubly selective and global synchronization is not needed.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper subcarrier spacing and CP duration. as well as increased flexibility and scalability. and even high moving speed and complicated scattering inter-subband asynchronous transmission can be environments among vehicles. Figure 11 Flexible and adaptive frame structure supported by F-OFDM i. the OOBE is (2) S u p p o r t s o m e s p e c i f i c s c e n a r i o s a n d guard band consumption. enabled. UFMC and F-OFDM. In such cases. leading to an enhanced capability of supporting different service types. which leads to a reduced the new multi-carrier schemes are needed. a flexible and configurable air interface can be provided. Figure 12 Symbol error rates of OFDM and FBMC in high-mobility scenarios (a is the scaling factor of the prototype filter) (3) Support asynchronous transmission with reduced overhead: For the proposed multi-carrier waveforms. FBMC. between adjacent subbands. With greater separation service types: For V2V and high-speed train. For example. Figure 13 Illustration for the OOBE of UFMC 22 .

can roughly be categorized into three groups: Multiple waveforms can coexist in 5G systems. Among link level multi-carrier waveforms should start from the coding. the requirement for single user link data rate is extremely high. amplitude and phase shift key (APSK). In dense deployment scenario. The design of new coding and link adaptation. to meet the high capacity utilizes the broadcast nature of air interface and requirement of the systems. Advanced link adaptation allows the modulation and channel coding to match the fast fading channel more precisely through optimizing code structure and retransmission bit distribution base on strategies to reduce the inter-node interference. as illustrated in Fig. rate-compatible codes. For example. link level modulation and coding. to approach Shannon’s capacity at high SNR region. each and bit interleaving.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper In summary.I M T. In non-binary codes. which can make the link more robust in fast fading channels. and rate-compatible. the system should be able to support high spectrum efficiency and long code blocks. Joint modulation and coding is characterized by rotated phase of constellation. and then mapping and joint modulation & coding. and some engineering oriented coding. They application scenarios and service types of 5G. The diversity gain is also having drastically different performance achieved in this process. in high-capacity hot-spot scenario. there are non-binary channel coding. Link select the most suitable waveform and numerology to enable the best performance for each type of services. Network coding the-air transmission. network each for a specific scenario. can make use of inter-node communications to collect the relevant bit information. the link performance can get very close 6. Hence. and some to fully utilize the broadcast nature of over- engineering oriented codes. adaptation includes rateless codes based. via Galorie Field operation There are a number of scenarios for 5G. wireless backhaul would be widely used. New bit mapping uses requirements. rateless codes. Advanced Modulation and Coding to Shannon’s limit. it is difficult for a single waveform There are quite a few distinct research directions to fulfill all the requirements of diversified in advanced modulation and coding. as long as there is wide bandwidth available and the channel condition is good. in order to further improve the system capacity. 23 . which drives more advanced channel coding and routing co-centric ring like constellation. bit requirements of scenarios and services.14.

results in the amplitude distribution computational complexity. being very close to Gaussian. The design objective of non-binary codes is to get together with channel coding and advanced better performance. In link level codes.I M T. A few from non-binary field to constellation is another sub-optimal algorithms have been proposed important aspect and can directly affect the whose performance is slightly worse than the performance. for example. decoding complexity as binary codes.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Figure 14 Categorization of advanced modulation and coding (1) Link level modulation and coding traditional QAM. These new constellations. Its design should (RA) codes are considered as promising codes. Currently. the mapping detection and the optimization of filters. aim at reducing the complexity of sequence Apart from the code structure itself. One example is FTN whose performance is very M-ary LDPC and non-binary repeat accumulate closer to Shannon’s capacity. The new bit mapping. optimal detector. however with the comparable demodulation algorithms can be jointly designed. however with much reduced APSK. compared to the envelop coding provides a simple and effective 24 .

5~0. via XOR operation of information bits in multiple sub-codeblocks. In rate-compatible codes. LDPC has the potential to be designed with more flexible code rates and block sizes. and better match the fast fading in the absence of accurate channel state information feedback. and closer matching with the channel conditions of each link in the entire chain. APSK (NE-APSK) can improve the performance (3) Link adaptation by about 0. and detailed solutions Figure 15 Link simulation comparison between 16-ary codes and binary LDPC should target certain scenarios.16 shows the performance comparison They would have different impacts on the air between APSK and QAM. (2) Network coding There are two aspects of study on network coding: codes design and system-level design. the refined interface specifications. At 1% BER. At 0. The link level performance of 16-ary codes and binary LDPC is compared in Fig. In the engineering optimization. Network coding is highly coupled with deployment scenario. Figure 16 Link simulation comparison between APSK and QAM 25 . cooperative relay or bi-directional transmission. for example. forward stack algorithm can significantly reduce the decoding complexity of spinal codes. And the gain is even more The main study on link adaptation is to design pronounced when the modulation order gets multiple choices of code rates and code block higher. selection of router. soft ACK/NACK can improved the efficiency of transmission.I M T. Fig. The design of network codes has the target of improving the efficiency of cooperative transmission.5 dB. In the area of rateless codes. the gain of non-binary codes is around 0. and resource scheduling.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper way to improve the error correction capability of the channel. sizes.15. with the assumption of AWGN channel. System design encompasses user pairing.9 dB.01%.

They have more powerful proximity can cooperate and relay data. Joint modulation and enhanced by D2D can achieve low latency. fast D2D can be applied to vehicle-tovehicle (V2V) and vehicle-to-infrastructure (V2I) communications in intelligent transportation system (ITS). thus also suitable for hot-spot can be easily supported. Utilizing the large coverage of cellular networks. 17 wireless communication capacity. Bluetooth. Figure 17 Gain in system throughput via network coding Many techniques in advanced modulation and coding can be used in various scenarios. For example. supporting not where the network coding plus the reasonable routing only the traditional cellular communications such and user pairing can provide gains. and therefore can meet the stringent requirements of V2X. short coding can improve the diversity via using range. it would be an important combination between centralized scheduling and technology in ultra-dense networks. as 2G/3G/4G. but also frequent V2V. Hence. data sharing network. The VDC solution high capacity scenario. and low cost D2D for IoT devices. smart meters. telematics services environment. tablets.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper The system throughput CDF is illustrated in Fig.I M T. more In future. but also Wi-Fi. M-ary LDPC. cooperative communications can and more smart devices come out such as smart be applied at network and terminal sides. multi-user cooperative communications. distributed communications. d e v i c e s i n t h e and smart vehicles. Network coding Vehicles. wearable devices. 7. and vehicle-to-pedestrian (V2P) communications (generally called V2X). Each 26 . c o n n e c t i n g w i t h e a c h o t h e r. high reliability V2X services to deal with multiple transmit antennas and would be used all the communication requirements of Internet of for wide coverage scenario. V2I. new bit mapping and FTN are suitable for wide-band and large data transmission. The VDC solution where D2D is tightly can improve the cooperation between multiple coupled with cellular networks can realize the nodes. By phones. The cooperation between cellular and direct communications may derive more use cases and improve user experience. They show significant gain in high SNR Vehicle direct communication (VDC): Internet o f Ve h i c l e s i n c l u d e s n o t o n l y t h e r e m o t e communication between vehicle and the Internet (telematics). LTED2D between devices. Device to Device (D2D) Multi-user cooperative communications (MUCC): Along with the technology development.

If the channel of one path is bad. low General speaking. and access mode can be adopted. a n d end devices and cellular network. the hierarchical M2M to the device with the best channel quality. V2I. In the process of controlling or assisting. (4) In low cost D2D scenario. thus the transmission range is devices and relay devices. can replace the direct connection between IoT Wi t h e n h a n c e d D 2 D c o m m u n i c a t i o n s . The (3) In data sharing network scenario. devices will be experience can be improved. which facilitates the large network coding to combine the received data scale deployment of IoT services. a new protocol layer is other. and the collision of V2X messages will be minimized charge the IoT end devices with the guarantee of and the reliability can be improved. V2V. The SNR can be improved through can be greatly reduced. data mutually. throughput. stringent requirement on latency and reliability. The data will be sent sensitive IoT systems. The low cost D2D connection delivery ratio (PDR) is needed for high reliability. By transmitting data between each For the protocol design. the base station can optimize the in order to improve road safety and transportation data transmission path timely to guarantee the e ff i c i e n c y. communication reliability and user (2) In MUCC scenarios. Since devices may move all The key issues and solutions include: the time. controlled by the network to complete pairing and D a t a s h a r i n g n e t w o r k : Wi t h b a s e s t a t i o n establish MUCC relationship. direct data transmission ad hoc network to share managed. security. the centralized scheduling and resource allocation. The IoT devices can then will be forwarded to the destination device. the topology of devices may change (1) In VDC scenarios. and high packet be designed well. the power of low cost D2D communications between the end devices is limited. connect to the cellular network via relaying. That will help large spreading for the 27 . the cost of end devices usually short. the low cost. the network load can be decreased and the needed for the network to select the best path system throughput can be improved. Then. Furthermore. Once the topology changes. a better optimized to reduce the latency and complexity of path can always be selected so that the system data exchange between the network and devices. control. T h e V 2 X s a f e t y m e s s a g e s h a v e service continuity. and V2P need to frequently. station. the related frequently broadcast in short range to exchange devices need to report the information to the base safety messages including location and velocity. from multiple links. network should be able to manage. towards the end device according to the channel Low cost D2D: For latency insensitive and cost quality of different devices.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper device may have several paths towards base the scheduling interaction process should be stations. the latency for safety messages power IoT end devices and relay devices need to should be less than 100ms. meanwhile. devices can establish communications. each device can be controlled.I M T. and clearly charged by the network.

future mobile traffic will presents varying characteristic: Especially. F l e x i b l e The main technical difficulty of flexible duplex is duplex could be applied to low power small how to deal with interference signals of different cells or relay nodes. new multiplex access schemes are likely to be applied. reference DL slot ratio according to traffic demand. In FDD system UL signals have many difference. the resources could be allocated for DL or UL transmission dynamically. transmission in the time domain scheme. different they are located very close to each other. such as multiple as shown in Fig. For the upcoming 5G. In the Therefore. new bands and bands. the design of DL and frequency domain schemes. Then.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper IoT applications such as metering and wearable devices. the UL bands are signals is the essential point to improve system assigned as flexible bands (DL or UL bands) to performance. such as carrier moving. a uniform adapt to the traffic demand. However. by setting different UL/ access method. subcarrier mapping. In TDD systems. the symmetrical design for DL and UL frequency domain scheme. transmission direction. such as DL-UL or UL-DL Flexible duplex could be realized by time or interference. w h i c h m a y b r i n g d i ff i c u l t y UL bands can be assigned for both UL and DL to interference recognition and cancellation. Flexible Duplex Wi t h t h e d e v e l o p m e n t o f o n l i n e v i d e o a n d social network service. With flexible duplex. which is triggering the new design of UL 28 . which direction signals could be recognized and is similar to the scheme applied in FDD UL cancelled.18. In LTE system. adjust RS based on their own service demands. which could improve spectrum Figure 18 Flexible resource allocation in time and frequency domains u t i l i z a t i o n e ff i c i e n c y e ff e c t i v e l y. the s i g n a l p a t t e r n . signal format could be realized by adjusting UL each cell can allocate the DL/UL slot ratio or DL signals. 8. In current LTE system. the ratio of DL and UL traffic may change with time and location.I M T. even if pattern or RS muting method. the current static allocation of the frequency/time resources for downlink and uplink transmission in LTE is not efficient to support dynamic asymmetry cell traffic.

two with heavy UL load.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper and DL signals. Fig. that’s especially helpful carrier mapping and orthogonality of RS signals. stations. and UL and DL signals have different structure. “IC-x” corresponds to the situation after the method is applied. In the figure. so interferences from different direction could be cancelled by the aid of network assistant. which makes it difficult for receivers to do interference cancellation. As inter-cell the interference signals in different direction are interferences mainly come from the other two transformed into signals in the same direction. while pico cells will take more mobile traffic. flexible duplex may be more suitable for pico cells with low transmission power. Figure 20 DL cell average throughput 29 .19 and Fig. for UL throughput improvement. The results show that with symmetrical design. and two with heavy DL load. “WO IC” corresponds to the situation before the former mentioned method is applied. only the interferences within the cancellation or interference coordination method same cluster need to be considered in practice.I M T. the transmit power level of base stations can be reduced to the level of mobile terminals. then the system throughput could be improved significantly. which could help reduce the complexity and To further limit the interference between base simplify the calculation in the receiver. The trend of traffic allocation is that macro cells will be mainly responsible for management and control functions. Figure 19 UL cell average throughput If DL-UL or UL-DL interference signals are transformed into signals with the same direction. and pico cells with different signal direction within then it could be cancelled by current interference the same cluster. Four pico cells are located within one cluster. Therefore. If the new UL and DL signals different direction interference signals could be have the symmetrical design including design of cancelled effectively. dealing with signals in the same direction.20 show the cell average UL&DL throughput in dense cluster scenario.

and the digital domain improve the spectrum efficiency for FDD and self-interference cancellation further suppresses TDD by leveraging a unified spectrum resource the SI by reconstructing and removing its linear management strategy. a novel and nonlinear residuals. Full Duplex interference (SI) at the propagation stage through Revolutions in wireless communication antennas’ placement. the contradiction between the ever-increasing etc. flexible duplex has a cancellation. and it is mature and industry upgrading. there have been considerable key techniques and scenarios to promote technology researches in academia and industry. still possible to apply for modifying the regulation (1) Self-interference cancellation if the interferences between neighbor systems could From the physical-layer perspective. Therefore. zero-forcing beamforming theories and techniques have been driven by and highly isolated transmit-receive antennas.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Although the application of flexible duplex may be The main research content of full duplex is currently constrained by spectrum regulation. at the receiver broadband services. it can enhance network efficiency and domain is a viable way to enable full duplex. a bandwidth of 20 MHz. the core be controlled and the gain of flexible duplex is problem of full duplex is how to efficiently widely approved. facing the challenge of eliminate the self-interference. The full duplex.. In flexibility. 30 . To this end. In theory. by applying joint self-interference and 4G enhancement. where limited spectrum resources. one of the key objectives is to added at the RF frontend. Flexible duplex can be applied to both 5G particular. the RF domain self-interference cancellation wireless communication service demands and the cancels the SI signal at the analog domain. full duplex technology is capable of doubling the spectrum efficiency of existing communication systems. is a promising solution for achieving that objective. it is divided into two aspects. duplex architecture building upon advanced self-interference cancellation techniques. recognized that joint self-interference cancellation Flexible duplex meets the trend of TDD&FDD from spatial domain. Furthermore. it is already able to achieve more good forward compatibility and its design methods can be than 115 dB self-interference cancellation over referred to full duplex systems. which is generated improving network performance brought by mobile from the local transmitter itself. RF domain and digital convergence.I M T. For future mobile an inverted SI signal in phase and amplitude is communications. it is essential to further study the side. The spatial domain self-interference cancellation suppresses self- 9.

is scheduled for user 1’s uplink and From the network perspective. meanwhile. the cell and industry. i. There are totally 9 interference coordination strategies. it is a must network architectures.e. respectively. preliminary for BS to operate in full-duplex mode at frequencies studies have been conducted by the academia f1 and f2. etc. Clearly. reuse increases with the number of users. The area is divided into 9 partitions and of LTE frame structure under full-duplex mode etc. A frequency pair (f1. full duplex has downlink transmissions. f2). f2 for the uplink spectrum usages. Fig. the inter-node of this partitioning method. more flexibility to control the way of transmission user 2 is allocated on the same frequency pair and reception.22 shows an example full-duplex/half-duplex networks. By changing the way of traditional by reusing the reverse links. and meanwhile.22. it is expected that full duplex will and f 1 for the downlink. doubled as long as the inter-user interference can which in turn requires more efficiently designed be effectively eliminated. the resource pairs of carrier frequencies (indexed by 1~9) to management of full-duplex networks and the design be used.. with an emphasis on the following partitioning method can be employed to mitigate the directions: the network architecture design for hybrid inter-user interference..2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Figure 21 The principle of the full duplex (2) Network technology in full duplex indexed by i. and that 31 . To this end. at the user side.I M T. let us consider a network Fig. To fulfill this. the frequency bring about more innovative techniques on access utilization factor under this allocation scheme is protocols and wireless resource management. allocated with specific frequency pairs as depicted in As an illustration. Theoretical analysis and simulation results consisting of a full-duplex base station (BS) and demonstrate that the probability of frequency reverse multiple FDD terminals.

S p e c i f i c a l l y. t h e s e c h a l l e n g e s i n c l u d e t h e novel network architectures and self- suppression of SI signals with high power and interference cancellation methods for full-duplex 32 . Liu Donglin. Figure 22 Simulation results Source: Shao Shihao.18.1043~1046.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper the frequency utilization gain over conventional power constraint.I M T.6. there are still large dynamic range.” IEEE Communications Letters. compact circuit design many technical challenges to be addressed. no. vol. Tang Youxi. To put full duplex into practice. under the terminals’ allowable interference 22) to achieve higher spectrum efficiency. it is more favorable to divide the half-duplex systems is improved significantly. June 2014. addition. In area into the smallest possible partitions (P in Fig. pp. “Analysis of carrier utilization in full-duplex cellular networks by dividing the co-channel interference region. Pan Zhengang. for multi-antenna RF domain cancellation. Deng Kai.

5GHz.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper communications. IMT system nodes efficiency and coverage which finally increase could be managed and licensed by database integrated frequency utilization efficiency. and consequently the potential scenario: The target bands could be the to improve the system spectrum efficiency. 33 . dedicated spectrum. etc. Spectrum Sharing Technology small cells in the same coverage of hot- To satisfy the requirements of ultra-high traffic spot area or macro-small cells in the same and ultra-high data rates for 5G. spectrum utilization methods should be further • Spectrum sharing in unlicensed band investigated to extend usable frequency of s c e n a r i o : T h e m a i n t a rg e t b a n d s a r e IMT system. Furthermore. inter- as the indoor.I M T. in addition to requiring more IMT unlicensed bands including 2. co-existence. These stations can be deployed different networks or systems. Different operators can be 10. • Inter-operator spectrum sharing scenario: The target bands could be unreleased IMT planned bands. and flexibility to transceiver and network designs.R AT the FDD/TDD half-duplex system.4GHz. capability of eliminating the difference between • Intra-operator inter-RAT spectrum sharing FDD and TDD. and low-power site macro cells. access network and handover between • Spectrum sharing in secondary access band networks. The involved station types are small sharing technique can enable optimal dynamic cells and Wi-Fi access points in target spectrum allocation and management among bands. the prioritized sensing solution.o p e r a t o r s p e c t r u m Overall. the innovative coverage. dynamic and flexible bands for other systems such as satellite. Other more complex and existing carriers can be aggregated with advanced application scenarios will be taken spcetrum sharing carrier.o p e r a t o r i n t e r.The communications. spectrum sharing in unlicensed band. or occupy frequency based on spectrum For spectrum sharing technology. and radar. spectrum etc. low-mobility. s p e c t r u m s h a r i n g . Spectrum sharing technique could scenario: The target bands are licensed achieve efficient. spectrum utilization to improve air interface broadcast. The different RATs here could be with limited and fragmented spectrum. it jointly with the macro cells using the has the self-adaption functions of autonomous existing low-frequency bands. and evolution with s c e n a r i o s i n c l u d e i n t r a . or macro-small cells. i n t e r. full duplex technology has more sharing. In 5G system. such deployed in co-located macro cells. new bands for 5G and/or the legacy IMT Full duplex technology is suitable for scenarios bands. into account with the maturity of the full duplex technology. spectrum sharing in secondary access.

support (2) High layer technology flexible bandwidth of RF and modulation when access The high layer technology for spectrum sharing to different systems at the same spectrum. It should be and feedback mechanisms and physical channels and considered to add spectrum management nodes in reference signals. and security of multi-system optimized allocation among multi-sharing nodes integration. to identify the RF networks in different priorities. and spectrum switching. Based on different reference structure and parameters. such as spectrum map generation This new RF design needs to support broader range and management. registration and authentication. New radio design The above architecture may involve database supported spectrum sharing should also be studied. The sharing nodes operate multiple access methods. of bands/ frequency. Physical layer technology for spectrum sharing is The centralized architecture will be a basis. measurements and requirements calculation. (4) RF technology receive the results of spectrum allocation and For RF technology. Some evaluation 34 . multi-mode multi-band chips are execute resources reconfiguration within the node. and support needs to be studied to solve the spectrum dynamic broadband spectrum sensing through multi-channel change issue due to sharing and coexistence of detection or compressed sensing. acquiring requirement application changing of interference backgrounds. (1) Network architecture and interfaces (3) Physical layer technology The network architecture needs to be modified. which also in the scope of study. suppress interference efficiently channel assignment. and analysis of possible allocation decisions. etc. designs for measurement sharing nodes need to be designed. adaptation to resource pool.I M T. combination with of sharing nodes and performing spectrum cognitive network technology. The include: obtaining usage status of spectrum based on new interfaces between new nodes and between sensing/detection methods. supporting results reporting. mobility and other operations.. spectrum fair negotiation and design under each scenario. spectrum the higher layer of sharing nodes for maintenance allocation and utilization. the study should focus The other technical issues include coexistence issue on management of large fragmented frequency between shared spectrum systems. access control and cross- Initial simulation for inter-operator spectrum layer design based on the prediction and cost sharing is evaluated under indoor small cell analysis. becoming a mainstream in the market. technology. The related study should also analyze dense deployment scenario. interference management and management of the shared frequency based on the self-cognitive of interference. traffic spectrum sharing are as follows: management. network architecture resources. The possible enhancements can be combined with distributed architecture. system architectures.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper The essential issues and potential solutions for the impacts on the existing network access. and learning mechanism. under multi-channel simultaneously operation.

5-2. user experience can still reach 12 Mbps under 90% load. a i r i n t e r f a c e t e c h n o l o g y. the throughput of dedicated allocation method and spectrum sharing spectrum sharing is much higher than the sum method. it is able to fully integrate with the other key technologies of 5G.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper assumptions include: • Contiguous 180 MHz bandwidth in 3-6 GHz frequency band for spectrum sharing between operators.5. inter-operator. a n d interference management under different key scenarios (intra-operator. three operators share user experienced rate can be improved 2-7 times the whole 180MHz spectrum. and radio resource management. database. In the meanwhile. including From the simulation results. Spectrum sharing technology can promote new spectrum management methods and improve the existing network capacity by addressing some technical issues such as system architectures. where users select higher. e.g. are compared in the evaluation. unlicensed bands accessing and secondary accessing).. edge subscribed operator to access. Especially for edge users. For of all the operator ’s throughputs of dedicated the dedicated allocation method. supporting Figure 21 Topology of scenario ultra-dense network based on dynamic spectrum 35 . each operators allocation. It is compatible to the existing technologies including carrier aggregation. Two spectrum allocation methods. i n t e r f a c e s . • The traffic model is burst traffic. where the packet size is 8 Mbytes and the packet arrival Figure 22 Simulation results rate of 0.I M T. the usable individually occupies 60 MHz bandwidth.814. For the spectrum of these users is increased and the spectrum sharing method. • Evaluation scenario is the indoor hot-spot dual stripe scenario (6 floors) defined in 3GPP TR36. When sharing 180MHz bandwidth.

The available IMT spectrum with a good application prospect.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper management. From the current and solve the load balancing issues of multi-RAT international research trends worldwide. The user experienced rate could sharing is one of key technologies for 5G system also be improved. it could enhance have higher requirements on baseband algorithms the total spectrum efficiency of different operators and devices capabilities. can be expanded when the over-load and/or overidle of operators spectrum may be relieved. spectrum sharing techniques require support of spectrum regulatory 36 . and economic models. spectrum and multi-cell. and also spectrum demand of 5G. bands and higher frequency bands to meet the high security strategy. Thus. spectrum sharing can significantly contribute to the performance of 5G in wide-area and hotspot scenario.I M T. However. jointly utilizing the existing IMT policy which needs to establish new usage rules. Furthermore.

and all-spectrum access. (including low-frequency and high-frequency enterprises. research institutes. massive MIMO. ultra-dense standards and industry. to fully meet the requirements of mobile internet and IoT in year 2020 and beyond. the optimized technical solutions can be derived for specific scenarios including seamless widearea coverage. By flexibly configuring technical modules and parameters. high-capacity hot-spot. and will to define a globally recognized 5G technology be powered by a group of key technologies such as architecture and promote the development of 5G novel multiple access. and universities branches) and 4G evolution air interface. 37 .I M T. IMT-2020 (5G) Promotion Group is willing to 5G will develop through the new air interface strengthen cooperation with global organizations. low-power massive-connection. and low-latency high-reliability.2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Conclusion Based on a unified air interface technical framework. networking.

2 0 2 0 ( 5 G ) Pr o m o t i o n G r o u p 5G Wireless Technology Architecture White Paper Main Contributors 38 .I M T.