Beruflich Dokumente
Kultur Dokumente
net/publication/313387969
CITATIONS READS
64 1,878
2 authors:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Naser Al-Falahy on 21 February 2019.
Technologies for
5G Networks:
Challenges and
Opportunities
Extensive mobile data traffic has led researchers and designers to begin
developing fifth-generation (5G) networks. This article reviews potential
technologies for 5G and concludes that radical changes to the network
paradigm will be necessary.
T
he significant growth in wireless broadband • higher numbers of connected devices (massive
traffic has had a major impact on future connectivity) stemming from the Internet of
mobile network architectures. Such archi- Things (IoT), which will support massive ma-
tectures will need to deal with increasing demands, chine-to-machine (M2M) communications and
including applications; and
• improved energy consumption—5G must be a
• high traffic volume (massive capacity), involv- green network to reduce its carbon footprint.1
ing increases on the order of several magni-
tudes—the future requirement is a 1,000× Mobile communication has evolved from
increase in data traffic for 2020 and beyond; voice-only services into a complex, interconnect-
• increased indoor and small cell/hotspot traf- ed environment with multiple services built on
fic, which will make up the majority of mobile a system that supports myriad applications and
traffic volume—today, roughly 70 percent of provides high-speed access to a massive number
mobile traffic happens indoors; in the future, of subscribers and machines.2
indoor data traffic and hotspot areas could ex- Here, we examine potential technologies that
ceed this figure; could define the 5G standard in the next few
12 IT Pro January/February 2017 Published by the IEEE Computer Society 1520-9202/17/$33.00 © 2017 IEEE
Internet
MW macrocell
lan on
C-p aling
e
Mobile core ta
Da plane
n
Sig
network Data U-
on
Femtocell & Wi-Fi
(indoor solution)
Massive-MIMO base
mm-wave small cell station with
beamforming
Dense small cell deployment
In-field
devices
Re
so
Sensors
ur
ce
all
Weather
oc
at
devices
ion
Sensors (Signaling + data)
Smart
meters Massive MIMO/
Massive M2M traffic due to D2D communications beamforming
billions of connected
machines
Figure 1. 5G multitier network architecture. The next generation of network technologies includes
macrocells (bands <3 GHz); small cells (millimeter-wave); femtocells and Wi-Fi (millimeter-wave); massive
multiple-input, multiple-output (M-MIMO) with beamforming; and device-to-device (D2D) and machine-to-
machine (M2M) communications. Solid arrows indicate wireless (fronthaul) links, whereas the dashed arrows
indicate backhaul links.
years. We look at possible challenges and try to The answer to the question of evolution or revo-
answer important questions that should be inter- lution will thus depend highly on the changes that
esting to researchers in this field. shape 5G,4 which could be any of the following:
computer.org/ITPro 13
Telecommunications Net working
computer.org/ITPro 15
Telecommunications Net working
Massive MIMO and beamforming. M-MIMO One of the ways to reduce latency is through
is considered when a network is equipped with a dense small cells (as discussed) and D2D com-
large number of antennas at its base stations that munications, as follows: If two devices are in
can accommodate many co-channel users at a close proximity, their communications can be
time. Beamforming is the concentration of power handled via D2D communication without con-
in a certain direction with a limited beam width suming network resources. D2D can handle lo-
but a large gain. cal traffic efficiently. It is an important option for
Beamforming and M-MIMO are key enabling applications that require low latency. D2D has
technologies for 5G systems. M-MIMO can signif- already been studied as a 4G technology (release
icantly improve signal strength, which could result 12) in the Third-Generation Partnership Project
in much higher cell throughput and better cell- (3GPP), and its adoption is driven by its impor-
edge performance than traditional 4G systems.11 tance for safety and disaster applications and
One challenge with M-MIMO is pilot con- low-latency applications.4
tamination4 from nearby cells as the number of In this area, the challenges are efficient prox-
antennas increases. Researchers must optimize imity detection, network integration, and native
pilot orthogonality without consuming network support in future 5G networks.
resources.
Second, due to the “massive” number of anten- Spectral Efficiency Improvement
nas used, accurate channel estimation is a chal- Spectral efficiency improvement is vital in 5G
lenging issue even with time division duplexing to deliver ultra-fast data speeds to more smart-
(TDD) due to huge costs and complexity. A more phones and tablets than ever before. Spectral
sophisticated algorithm is necessary to enable ac- efficiency (bit/sec/Hz) can be increased by in-
curate channel estimation in frequency division creasing the modulation order, through D2D
duplexing (FDD) and to reduce signaling overhead. communications (as discussed), M-MIMO, and
Finally, the physical size of M-MIMO is a point the adoption of new efficient transmission wave-
of concern because it requires a very large-scale forms, as discussed next.
architecture.12 It will thus face pushback from the Orthogonal frequency division multiplexing
public and property owners regarding potential (OFDM) is a powerful and inherent way to ad-
environmental issues. Moreover, larger-sized tow- dress the problem of intersymbol interference
ers will create extra technical challenges, which (ISI). Instead of sending information on a single
will cause further pushback. However, a success- carrier, OFDM uses multiple carriers to transmit
ful marriage that could address the physical size simultaneous subframes after dividing the main
issue is to pair M-MIMO with the millimeter- stream and modulating each subframe on a dif-
wave band.13 A realistic array size will become ferent subcarrier frequency, which helps combat
possible that facilitates M-MIMO installation. multipath and ISI.
There are several possible challenges. 5G wave-
Latency Reduction forms should establish a set of requirements,
Latency is the time it takes a signal to complete such as high spectral efficiency, low latency, and
a single, full transaction. Besides achieving high limited cost and complexity. 5G systems will have
data rates, latency reduction becomes vital to en- several strategies to fulfill these requirements,
able energy savings and long battery lifetimes. such as dense small cell deployments and use of
Current 4G latency is about 15 ms based on the the millimeter-wave band, which will be directly
1-ms subframe. This latency is considered perfect influenced by the modulation format used at the
for current applications; however, 5G will intro- physical layer.14
duce technologies such as tactile Internet, two- Will OFDM be the dominant theme in 5G?
way real-time gaming, cloud-based applications, OFDM is not exempt from drawbacks, and its
and augmented reality that cannot be supported adoption by 5G should not be taken for granted.
at current latencies (see Figure 3). Therefore, 5G The main disadvantage of OFDM is its high peak-
should support latencies lower than 1 ms, which to-average-power ratio (PAPR), which decreases
will have a major impact on design choices at all power amplifier efficiency. Also, cyclic-prefix
layers.5 (CP) insertion decreases spectral efficiency.
Figure 3. Data speed vs. latency. According to the required services and applications, this figure
shows a tradeoff between latency and data speed, and the potential of next-generation networks to
support their demands. UHD: ultra-high definition.
New schemes can be further utilized to im- human devices. With the increased availability
prove spectral efficiency; those schemes include of mobile broadband, connectivity has become a
realistic issue for M2M communication.
• nonorthogonal multiple access (NOMA), which However, the massive traffic growth expected
can ensure that multiple users share a wireless from machine-type communication as a result of
medium and experience the same diversity as billions of connected devices will cause the net-
orthogonal multiple access techniques; work to become congested. So, a several orders of
• filter bank multicarrier (FBMC), in which sig- magnitude increase in network connectivity and
nals on each subcarrier are shaped by a well- capacity is required, which can be met with net-
designed filter to suppress signal side lobes and work densification, dense small cell deployment,
limit its band; and and M-MIMO.
• sparse coded multiple access (SCMA), in which Moving data access to the cloud will also play
different data are directly mapped to code a significant part in 5G, so that the network can
words of different codebooks, where each code be accessed from anywhere. Network function
represents a spread transmission layer. virtualization (NFV) can make functions with
hardware compatibility issues run on cloud
As an example, with successive interference can- computing infrastructure. Thus, there will be a
celling (SIC) receivers, NOMA has improved higher reuse of network infrastructure than in
throughput in macrocells by up to 30 percent the current network. In addition, CoMP can turn
compared to orthogonal multiple access schemes.6 interference into useful signals.5,11 Table 2 illus-
trates some challenges that can be tackled with
Massive Connectivity and the IoT specific technologies in the 5G system.
In the long term, it is expected that all devices
that benefit from network connectivity will even- Migrating Technologies to 5G
tually become connected,11 and the number In 5G, coverage holes are expected due to high
of connected devices will exceed the number of path loss at the millimeter-wave band. Thus, the
computer.org/ITPro 17
Telecommunications Net working
4G system will be required to cover the overall of users and devices to seamlessly connect in
area during the early stages of 5G deployment. 5G smart cities (IoT) by 2020 and beyond (w w w.
must use the primary MW band in addition to the itu.int/en/ITU-R/study-groups/rsg5/rwp5d/imt-
complementary millimeter-wave band. This spec- 2020/Pages/default.aspx).
trum must be migrated to 5G, or poor coverage At WRC 2015, the main objective focused on
can be expected. In addition, among the new fea- adding extra spectrum for mobile communica-
tures heralded by 5G, D2D transfers could have a tions below 6 GHz. However, the massive growth
prominent role. The adoption of D2D transfers is in global mobile traffic cannot be fulfilled by this
driven by safety and disaster systems, applications addition alone. 5G will need to access and extend
requiring low latency, and network traffic offload- its operation to the millimeter-wave band to en-
ing. M2M is the engine for the future IoT; CoMP able multi-Gbps data rates. Therefore, it was de-
technology and carrier aggregation will also be cided that at WRC in 2019, the identification of
transferred for better spectral efficiency and QoS. bands above 6 GHz will be included.
Beamforming and M-MIMO are key enabling The ITU-R Working Party 5D will define the
technologies for the millimeter-wave band, so technical performance requirements for next-
their transfer can be taken for granted. generation systems and develop an evaluation pro-
cess to occur between 2016 and 2017. According
On the Way Toward 5G to the ITU timeframe, standardization and pro-
The trend in future mobile networks (5G) has posals will be studied in 2018. From 2018 through
shown a different pattern from that of exist- 2020, an evaluation will be held by external
ing networks, because the main objective has groups, and the definition of new radio interfaces
changed from enabling users to connect wire- will be included in the most recent International
lessly to the Internet to enabling massive numbers Mobile Telecommunication system (IMT-2020),
Feature 1G 2G 3G 4G 5G
Deployment 1980 1990 2001 2010 2020 or beyond (www.itu.
int/en/ITU-R/study-groups/
rsg5/rwp5d/imt-2020/
Pages/default.aspx)
Frequency 800 MHz 900 MHz 2,100 MHz 2,600 MHz 3–90 GHz15
band
Speed 2 Kbps 64 Kbps 2 Mbps 1 Gbps Higher than 1 Gbps
Technology Analogue Digital cellular Code division Long-Term Multi-radio access
cellular multiple access, Evolution technology, Wi-Fi,
Universal Mobile Advanced, Wi-Gig16
Telecommunications Wi-Fi
System
Services Voice Digital voice, Higher quality High data Very high data rate13
SMS, packet audio and video rate, wearable to fulfill extreme user
(General Packet calls, mobile devices demands, device-to-device,
Radio Service), broadband machine-to-machine,
low-rate data Internet of Things
Multiplexing Frequency Time division Code division Orthogonal Orthogonal frequency-division
division multiple access multiple access frequency- multiplexing, filter bank
multiple division multicarrier, nonorthogonal
access multiple access multiple access14
Handover No Horizontal Horizontal Horizontal/ Horizontal/vertical
vertical
Switching Circuit Circuit/packet Packet All packet All packet17
Core Public Public switched Packet network Internet Internet
network switched telephone
telephone network
network
similar to what happened for IMT-2000, and has achieved a stable connection at 1.2 Gbps in
IMT-Advanced. Table 3 shows the difference be- a mobile environment from a vehicle at a speed
tween 5G and old mobile network generations, of 100 km/h at 28 GHz. In addition, Nokia has
listing changes to several features.13–17 used the 73-GHz carrier with 2-GHz band-
width to achieve a speed of 10 Gbps with latency
Current Development to 5G Realization around 1 ms.20 The ITU set a timeframe for 5G
Given that current mobile phones operate in fre- systems; its IMT-2020 group reviewed many re-
quencies between 0.8 to 2.5 GHz, they are capable search proposals and will soon set the first 5G
of download speeds of only 230 Mbps. Therefore, network design.13
mobile devices operating in the millimeter-wave
band are essential to cope with the higher-speed
A
data transmissions required from 5G. Fujitsu has s the demand for high-speed and low-la-
developed a millimeter-wave prototype receiver tency applications increases dramatically,
small enough to be incorporated into a mobile the 5G system should have the technol-
phone.3 This receiver has achieved 20-Gbps ogy and flexibility to meet these those require-
download speeds. Fujitsu began field-testing in ments and support multifold increases in network
2016 and will launch the receiver in 2020. Fur- capacity and connectivity.
thermore, IEEE has developed the IEEE 802.11ad The extremely high data throughput and very
standard, which operates at 60 GHz and supports low latency required from 5G cannot be satisfied
a speed of 7 Gbps within a short distance.18 only through evolution or modification of the ex-
Samsung has announced that it has achieved isting 4G network. Therefore, researchers must
7.5 Gbps, the fastest ever 5G data transmission focus on technologies that would have a major
rate in a stationary environment.19 The company impact on system performance. This will come
computer.org/ITPro 19
Telecommunications Net working
by introducing radical changes at the base station 12. C. Lin et al., “Toward Green and Soft : A 5G Perspective,”
(component) and network (core and backhaul) IEEE Comm. Magazine, vol. 52, no. 2, 2014, pp. 66–73.
levels. The most prominent technologies and as- 13. L.J. Young, “Telecom Experts Plot a Path to 5G,” IEEE
pects that currently have this ability are millime- Spectrum, 6 Oct. 2015; spectrum.ieee.org/telecom/wire-
ter-wave band, dense deployment of small cells, less/telecom-experts-plot-a-path-to-5g.
D2D, M2M, and M-MIMO with beamforming. 14. P. Banelli et al., “Modulation Formats and Waveforms for
With help from these technologies, future wire- 5G Networks: Who Will Be the Heir of OFDM?” IEEE
less systems will include myriad smart features Signal Processing Magazine, vol. 31, no. 11, 2014, pp. 80–93.
and applications to make 5G the most intelligent 15. F. Khan and Z. Pi, “mmWave Mobile Broadband
and dominant wireless technology thus far. (MMB): Unleashing the 3-300GHz Spectrum,” Proc.
34th IEEE Sarnoff Symp., 2011, pp. 1–6.
Acknowledgments 16. K. Zheng et al., “10 Gb/s HetSNets with Millimeter-
This work is sponsored by the ministry of higher education and Wave Communications: Access and Networking—
scientific research, University of Anbar, Iraq. Challenges and Protocols,” IEEE Comm. Magazine, vol.
53, no. 1, 2015, pp. 222–231.
References 17. R.S. Sapakal and S.S. Kadam, “5G Mobile Technol-
1. S. Chen and J. Zhao, “The Requirements, Chal- ogy,” Int’l J. Advanced Research in Computer Eng. and
lenges, and Technologies for 5G of Terrestrial Mobile Technology, vol. 2, no. 2, 2013, pp. 568–571.
Telecommunication,” IEEE Comm. Magazine, vol. 52, 18. T. Nitsche et al., “IEEE 802.11ad : Directional 60
no. 5, 2014, pp. 36–43. GHz Communication for Multi-Gigabit-per-Second
2. 5G: What Is It?, white paper, Ericsson, 2014, pp. 1–10. Wi-Fi,” IEEE Comm. Magazine, vol. 52, no. 12, 2014,
3. J. Boyd, “Fujitsu Makes a Terahertz Receiver Small Enough pp. 132–141.
for a Smartphone,” IEEE Spectrum, 6 Oct. 2015; http:// 19. J. Gozalvez, “Samsung Electronics Sets 5G Speed Re-
spectrum.ieee.org/tech-talk/telecom/wireless/fujitsu cord at 7.5 Gb/s,” IEEE Vehicular Technology Magazine,
-makes-a-terahertz-receiver-small-enough-for-a-smartphone. Mar. 2015, pp. 12–16.
4. F. Boccardi et al., “Five Disruptive Technology Direc- 20. Looking Ahead to 5G, white paper, Nokia Solutions and
tions for 5G,” IEEE Comm. Magazine, vol. 52, no. 2, 2014, Networks, 2013, pp. 1–16.
pp. 74–80.
5. J.G. Andrews et al., “What Will 5G Be?” IEEE J. Select- Naser Al-Falahy is a PhD candidate in mobile telecom-
ed Areas in Comm., vol. 32, no. 6, 2014, pp. 1065–1082. munications at the University of Salford, UK, and a member
6. P.K. Agyapong et al., “Design Considerations for a 5G of the electrical engineering department at the University of
Network Architecture,” IEEE Comm. Magazine, vol. Anbar, Iraq. His research interests include mobile commu-
52, no. 11, 2014, pp. 65–75. nications, radio network planning and optimization, and
7. F. Khan and Z. Pi, “An Introduction to Millimeter- millimeter wave communications. Al-Falahy has worked at
Wave Mobile Broadband Systems,” IEEE Comm. Mag- Motorola as an optimization engineer for mobile networks.
azine, vol. 59, no. 6, 2011, pp. 101–107. He received his MSc and BSc in electronics and commu-
8. S.G. Larew et al., “Air Interface Design and Ray Trac- nications from Al-Nahrain University–Baghdad. Contact
ing Study for 5G Millimeter Wave Communications,” him at naser_falahi@yahoo.com.
Proc. IEEE Globecom Workshops, 2013, pp. 117–122.
9. F. Khan, Z. Pi, and S. Rajagopal, “Millimeter-Wave Omar Y. Alani is the program leader of computer net-
Mobile Broadband with Large Scale Spatial Process- works in the School of Computing, Science, and Engineer-
ing for 5G Mobile Communication,” Proc. 50th Ann. ing at the University of Salford, UK. His research interests
Allerton Conf. Comm., Control, and Computing, 2012, include wireless multimedia communications, radio resource
pp. 1517–1523. management, location and mobility management in next-
10. T. Bai and R.W. Heath, “Coverage Analysis for Mil- generation mobile communication systems, diversity and
limeter Wave Cellular Networks with Blockage Ef- smart antenna techniques, and ad hoc and sensor networks.
fects,” Proc. IEEE Global Conf. Signal and Information Alani is a member of the Institution of Engineering and
Processing, 2013, pp. 727–730. Technology and an editor of the International Journal
11. Future Technology Trends of Terrestrial IMT Systems, of Mobile Communication. He received a PhD in tele-
report ITU-R M.2320-0, Int’l Telecomm. Union, communication engineering from De Montfort University,
2014. UK. Contact him at o.y.k.alani@salford.ac.uk.