cellular architecture and key technologies for 5g wireless communication networks

IEEE Communications Magazine • February 2014122 0163-6804/14/$25.00 © 2014 IEEE

INTRODUCTION

The innovative and effective use of informationand communication technologies (ICT) isbecoming increasingly important to improve theeconomy of the world [1]. Wireless communica-tion networks are perhaps the most critical ele-ment in the global ICT strategy, underpinningmany other industries. It is one of the fastestgrowing and most dynamic sectors in the world.

The European Mobile Observatory (EMO)reported that the mobile communication sectorhad total revenue of €174 billion in 2010, there-by bypassing the aerospace and pharmaceuticalsectors [2]. The development of wireless tech-nologies has greatly improved people’s ability tocommunicate and live in both business opera-tions and social functions.

The phenomenal success of wireless mobilecommunications is mirrored by a rapid pace oftechnology innovation. From the second genera-tion (2G) mobile communication system debutedin 1991 to the 3G system first launched in 2001,the wireless mobile network has transformedfrom a pure telephony system to a network thatcan transport rich multimedia contents. The 4Gwireless systems were designed to fulfill therequirements of International Mobile Telecom-munications-Advanced (IMT-A) using IP for allservices [3]. In 4G systems, an advanced radiointerface is used with orthogonal frequency-divi-sion multiplexing (OFDM), multiple-input multi-ple-output (MIMO), and link adaptationtechnologies. 4G wireless networks can supportdata rates of up to 1 Gb/s for low mobility, suchas nomadic/local wireless access, and up to 100Mb/s for high mobility, such as mobile access.Long-Term Evolution (LTE) and its extension,LTE-Advanced systems, as practical 4G systems,have recently been deployed or soon will bedeployed around the globe.

However, there is still a dramatic increase inthe number of users who subscribe to mobilebroadband systems every year. More and more

ABSTRACT

The fourth generation wireless communica-tion systems have been deployed or are soon tobe deployed in many countries. However, withan explosion of wireless mobile devices and ser-vices, there are still some challenges that cannotbe accommodated even by 4G, such as the spec-trum crisis and high energy consumption. Wire-less system designers have been facing thecontinuously increasing demand for high datarates and mobility required by new wirelessapplications and therefore have started researchon fifth generation wireless systems that areexpected to be deployed beyond 2020. In thisarticle, we propose a potential cellular architec-ture that separates indoor and outdoor scenar-ios, and discuss various promising technologiesfor 5G wireless communication systems, such asmassive MIMO, energy-efficient communica-tions, cognitive radio networks, and visible lightcommunications. Future challenges facing thesepotential technologies are also discussed.

5G WIRELESS COMMUNICATION SYSTEMS:PROSPECTS AND CHALLENGES

Cheng-Xiang Wang, Heriot-Watt University and University of Tabuk

Fourat Haider, Heriot-Watt University

Xiqi Gao and Xiao-Hu You, Southeast University

Yang Yang, ShanghaiTech University

Dongfeng Yuan, Shandong University

Hadi M. Aggoune, University of Tabuk

Harald Haas, University of Edinburgh

Simon Fletcher, NEC Telecom MODUS Ltd.

Erol Hepsaydir, Hutchison 3G UK

Cellular Architecture and KeyTechnologies for 5G WirelessCommunication Networks

WANG_LAYOUT.qxp_Layout 1/30/14 1:29 PM 22

IEEE Communications Magazine • February 2014 123

people crave faster Internet access on the move,trendier mobiles, and, in general, instant com-munication with others or access to information.More powerful smartphones and laptops arebecoming more popular nowadays, demandingadvanced multimedia capabilities. This hasresulted in an explosion of wireless mobiledevices and services. The EMO pointed out thatthere has been a 92 percent growth in mobilebroadband per year since 2006 [2]. It has beenpredicted by the Wireless World ResearchForum (WWRF) that 7 trillion wireless deviceswill serve 7 billion people by 2017; that is, thenumber of network-connected wireless deviceswill reach 1000 times the world’s population [4].As more and more devices go wireless, manyresearch challenges need to be addressed.

One of the most crucial challenges is thephysical scarcity of radio frequency (RF) spectraallocated for cellular communications. Cellularfrequencies use ultra-high-frequency bands forcellular phones, normally ranging from severalhundred megahertz to several gigahertz. Thesefrequency spectra have been used heavily, mak-ing it difficult for operators to acquire more.Another challenge is that the deployment ofadvanced wireless technologies comes at the costof high energy consumption. The increase ofenergy consumption in wireless communicationsystems causes an increase of CO2 emission indi-rectly, which currently is considered as a majorthreat for the environment. Moreover, it hasbeen reported by cellular operators that theenergy consumption of base stations (BSs) con-tributes to over 70 percent of their electricity bill[5]. In fact, energy-efficient communication wasnot one of the initial requirements in 4G wire-less systems, but it came up as an issue at a laterstage. Other challenges are, for example, aver-age spectral efficiency, high data rate and highmobility, seamless coverage, diverse quality ofservice (QoS) requirements, and fragmenteduser experience (incompatibility of differentwireless devices/interfaces and heterogeneousnetworks), to mention only a few.

All the above issues are putting more pres-sure on cellular service providers, who are facingcontinuously increasing demand for higher datarates, larger network capacity, higher spectralefficiency, higher energy efficiency, and highermobility required by new wireless applications.On the other hand, 4G networks have just aboutreached the theoretical limit on the data ratewith current technologies and therefore are notsufficient to accommodate the above challenges.In this sense, we need groundbreaking wirelesstechnologies to solve the above problems causedby trillions of wireless devices, and researchershave already started to investigate beyond 4G(B4G) or 5G wireless techniques. The projectUK-China Science Bridges: (B)4G Wireless MobileCommunications (http://www.ukchinab4g. ac.uk/) isperhaps one of the first projects in the world tostart B4G research, where some potential B4Gtechnologies were identified. Europe and Chinahave also initiated some 5G projects, such asMETIS 2020 (https://www.metis2020. com/) sup-ported by EU and National 863 Key Project in5G supported by the Ministry of Science andTechnology (MOST) in China. Nokia Siemens

Networks described how the underlying radioaccess technologies can be developed further tosupport up to 1000 times higher traffic volumescompared to 2010 travel levels over the next 10years [6]. Samsung demonstrated a wireless sys-tem using millimeter (mm) wave technologieswith data rates faster than 1 Gb/s over 2 km [7].

What will the 5G network, which is expectedto be standardized around 2020, look like? It isnow too early to define this with any certainty.However, it is widely agreed that compared tothe 4G network, the 5G network should achieve1000 times the system capacity, 10 times thespectral efficiency, energy efficiency and datarate (i.e., peak data rate of 10 Gb/s for lowmobility and peak data rate of 1 Gb/s for highmobility), and 25 times the average cell through-put. The aim is to connect the entire world, andachieve seamless and ubiquitous communica-tions between anybody (people to people), any-thing (people to machine, machine to machine),wherever they are (anywhere), whenever theyneed (anytime), by whatever electronicdevices/services/networks they wish (anyhow).This means that 5G networks should be able tosupport communications for some special sce-narios not supported by 4G networks (e.g., forhigh-speed train users). High-speed trains caneasily reach 350 up to 500 km/h, while 4G net-works can only support communication scenariosup to 250 km/h. In this article, we propose apotential 5G cellular architecture and discusssome promising technologies that can bedeployed to deliver the 5G requirements.

The remainder of this article is organized asfollows. We propose a potential 5G cellulararchitecture. We describe some promising keytechnologies that can be adopted in the 5G sys-tem. Future challenges are highlighted. Finally,conclusions are drawn.

A POTENTIAL 5G WIRELESSCELLULAR ARCHITECTURE

To address the above challenges and meet the5G system requirements, we need a dramaticchange in the design of cellular architecture. Weknow that wireless users stay indoors for about80 percent of time, while only stay ourdoorsabout 20 percent of the time [8]. The currentconventional cellular architecture normally usesan outdoor BS in the middle of a cell communi-cating with mobile users, no matter whether theystay indoors or outdoors. For indoor users com-municating with the outdoor BS, the signals haveto go through building walls, and this causes veryhigh penetration loss, which significantly dam-ages the data rate, spectral efficiency, and ener-gy efficiency of wireless transmissions.

One of the key ideas of designing the 5G cel-lular architecture is to separate outdoor andindoor scenarios so that penetration loss throughbuilding walls can somehow be avoided. This willbe assisted by distributed antenna system (DAS)and massive MIMO technology [9], where geo-graphically distributed antenna arrays with tensor hundreds of antenna elements are deployed.While most current MIMO systems utilize twoto four antennas, the goal of massive MIMO sys-

One of the key ideas

of designing the 5G

cellular architecture is

to separate outdoor

and indoor scenarios

so that penetration

loss through building

walls can be some-

how avoided. This

will be assisted by

distributed antenna

system (DAS) and

massive MIMO

technology.

WANG_LAYOUT.qxp_Layout 1/30/14 1:29 PM Page 123

tems is to exploit the potentially large capacitygains that would arise in larger arrays of anten-nas. Outdoor BSs will be equipped with largeantenna arrays with some antenna elements (alsolarge antenna arrays) distributed around the celland connected to the BS via optical fibers, bene-fiting from both DAS and massive MIMO tech-nologies. Outdoor mobile users are normallyequipped with limited numbers of antenna ele-ments, but they can collaborate with each otherto form a virtual large antenna array, whichtogether with BS antenna arrays will constructvirtual massive MIMO links. Large antennaarrays will also be installed outside of everybuilding to communicate with outdoor BSs ordistributed antenna elements of BSs, possiblywith line of sight (LoS) components. Large anten-na arrays have cables connected to the wirelessaccess points inside the building communicatingwith indoor users. This will certainly increase theinfrastructure cost in the short term while signifi-cantly improving the cell average throughput,spectral efficiency, energy efficiency, and datarate of the cellular system in the long run.

Using such a cellular architecture, as indoorusers only need to communicate with indoorwireless access points (not outdoor BSs) withlarge antenna arrays installed outside build-ings, many technologies can be utilized that aresuitable for short-range communications withhigh data rates. Some examples include WiFi,femtocell, ultra wideband (UWB), mm-wavecommunications (3–300 GHz) [7], and visiblelight communications (VLC) (400–490 THz)[10]. It is worth mentioning that mm-wave andVLC technologies use higher frequencies nottraditionally used for cellular communications.These high-frequency waves do not penetratesolid materials very well and can readily beabsorbed or scattered by gases, rain, andfoliage. Therefore, it is hard to use these wavesfor outdoor and long distance applications.However, with large bandwidths available, mm-wave and VLC technologies can greatlyincrease the transmission data rate for indoorscenarios. To solve the spectrum scarcity prob-lem, besides finding new spectrum not tradi-tionally used for wireless services (e.g. ,mm-wave communications and VLC), we canalso try to improve the spectrum utilization ofexisting radio spectra, for example, via cogni-tive radio (CR) networks [11].

The 5G cellular architecture should also be aheterogeneous one, with macrocells, microcells,small cells, and relays. To accommodate high-mobility users such as users in vehicles and high-speed trains, we have proposed the mobilefemtocell (MFemtocell) concept [12], whichcombines the concepts of mobile relay and fem-tocell. MFemtocells are located inside vehiclesto communicate with users within the vehicle,while large antenna arrays are located outsidethe vehicle to communicate with outdoor BSs.An MFemtocell and its associated users are allviewed as a single unit to the BS. From the userpoint of view, an MFemtocell is seen as a regu-lar BS. This is very similar to the above idea ofseparating indoor (inside the vehicle) and out-door scenarios. It has been shown in [12] thatusers using MFemtocells can enjoy high-data-

rate services with reduced signaling overhead.The above proposed 5G heterogeneous cellulararchitecture is illustrated in Fig. 1.

PROMISING KEY5G WIRELESS TECHNOLOGIES

In this section, based on the above proposedheterogeneous cellular architecture, we discusssome promising key wireless technologies thatcan enable 5G wireless networks to fulfill perfor-mance requirements. The purpose of developingthese technologies is to enable a dramatic capac-ity increase in the 5G network with efficient uti-lization of all possible resources. Based on thewell-known Shannon theory, the total systemcapacity Csum can be approximately expressed by

(1)

where Bi is the bandwidth of the ith channel, Piis the signal power of the ith channel, and Npdenotes the noise power. From Eq. 1, it is clearthat the total system capacity Csum is equivalentto the sum capacity of all subchannels and het-erogeneous networks. To increase Csum, we canincrease the network coverage (via heteroge-neous networks with macrocells, microcells,small cells, relays, MFemtocell [12], etc.), num-ber of subchannels (via massive MIMO [9], spa-tial modulation [SM] [13], cooperative MIMO,DAS, interference management, etc.), bandwidth(via CR networks [11], mm-wave communica-tions, VLC [10], multi-standard systems, etc.),and power (energy-efficient or green communi-cations). In the following, we focus on some ofthe key technologies.

MASSIVE MIMOMIMO systems consist of multiple antennas atboth the transmitter and receiver. By addingmultiple antennas, a greater degree of freedom(in addition to time and frequency dimensions)in wireless channels can be offered to accom-modate more information data. Hence, a signif-icant performance improvement can beobtained in terms of reliability, spectral effi-ciency, and energy efficiency. In massive MIMOsystems, the transmitter and/or receiver areequipped with a large number of antenna ele-ments (typically tens or even hundreds). Notethat the transmit antennas can be co-located ordistributed (i.e., a DAS system) in differentapplications. Also, the enormous number ofreceive antennas can be possessed by one deviceor distributed to many devices. Besides inherit-ing the benefits of conventional MIMO sys-tems, a massive MIMO system can alsosignificantly enhance both spectral efficiencyand energy efficiency [9]. Furthermore, in mas-sive MIMO systems, the effects of noise andfast fading vanish, and intracell interferencecan be mitigated using simple linear precodingand detection methods. By properly using multi-user MIMO (MU-MIMO) in massive MIMOsystems, the medium access control (MAC)layer design can be simplified by avoiding com-

∑∑≈ +⎛

⎝⎜

⎞

⎠⎟C B

P

Nlog 1i

i

psum 2

ChannelsHetNets

IEEE Communications Magazine • February 2014124

The 5G cellular archi-

tecture should also

be a heterogeneous

one, with macrocells,

microcells, small cells,

and relays. To

accommodate high-

mobility users such

as users in vehicles

and high-speed

trains, we have pro-

posed the mobile

femtocell concept,

which combines the

concepts of mobile

relay and femtocell.



plicated scheduling algorithms [14]. With MU-MIMO, the BS can send separate signals toindividual users using the same time-frequencyresource, as first pro. Consequently, these mainadvantages enable the massive MIMO systemto be a promising candidate for 5G wirelesscommunication networks.

SPATIAL MODULATIONSpatial modulation, as first proposed by Haas etal., is a novel MIMO technique that has beenproposed for low-complexity implementation ofMIMO systems without degrading system per-formance [13]. Instead of simultaneously trans-mitting multiple data streams from the availableantennas, SM encodes part of the data to betransmitted onto the spatial position of eachtransmit antenna in the antenna array. Thus,the antenna array plays the role of a second (inaddition to the usual signal constellation dia-gram) constellation diagram (the so-called spa-tial constellation diagram), which can be usedto increase the data rate (spatial multiplexing)with respect to single-antenna wireless systems.Only one transmit antenna is active at any time,while other antennas are idle. A block of infor-mation bits is split into two sub-blocks oflog2(NB) and log2(M) bits, where NB and M arethe number of transmit antennas and the sizeof the complex signal constellation diagram,respectively. The first sub-block identifies theactive antenna from a set of transmit antennas,while the second sub-block selects the symbolfrom the signal constellation diagram that willbe sent from that active antenna. Therefore,SM is a combination of space shift keying (SSK)and amplitude/phase modulation. Figure 2shows the SM constellation diagram with 4transmit antennas (NB = 4) and quadraturephase shift keying (QPSK) modulation (M = 4)

as an example. The receiver can then employoptimal maximum likelihood (ML) detection todecode the received signal.

Spatial modulation can mitigate three majorproblems in conventional MIMO systems: inter-channel interference, inter-antenna synchroniza-tion, and multiple RF chains [13]. Moreover,low-complexity receivers in SM systems can bedesigned and configured for any number oftransmit and receive antennas, even for unbal-anced MIMO systems. We have to point out thatthe multiplexing gain in SM increases logarith-mically with the increase in the number of trans-mit antennas, while it increases linearly inconventional MIMO systems. Therefore, the lowimplementation complexity comes at the expenseof sacrificing some degrees of freedom. Mostresearch on SM focuses on the case of a singlereceiver (i.e., single-user SM). Multi-user SMcan be considered as a new research direction tobe considered in 5G wireless communication sys-tems.

COGNITIVE RADIO NETWORKSThe CR network is an innovative softwaredefined radio technique considered to be one ofthe promising technologies to improve the uti-lization of the congested RF spectrum [9].Adopting CR is motivated by the fact that alarge portion of the radio spectrum is underuti-lized most of the time. In CR networks, a sec-ondary system can share spectrum bands withthe licensed primary system, either on an inter-ference-free basis or on an interference-tolerantbasis [9]. The CR network should be aware ofthe surrounding radio environment and regulateits transmission accordingly. In interference-freeCR networks, CR users are allowed to borrowspectrum resources only when licensed users donot use them. A key to enabling interference-

Figure 1. A proposed 5G heterogeneous wireless cellular architecture.

CR network

Large MIMOnetwork

Mobile femtocellnetwork

LOS

channel

........

Internet

Internet

Internet

.

.........

........

..........

...........

VLC

60 GHz

Gigabit Ethernet

WiFi

Femtocell

Core network

The 5G CR network

is an innovative soft-

ware defined radio

technique which has

been considered as

one of the promising

technologies to

improve the utiliza-

tion of the congest-

ed RF spectrum.

Adopting CR is moti-

vated by the fact

that a large portion

of the radio spec-

trum is underutilized

most of the time.


free CR networks is figuring out how to detectthe spectrum holes (white space) that spread outin wideband frequency spectrums. CR receiversshould first monitor and allocate the unusedspectrums via spectrum sensing (or combiningwith geolocation databases) and feed this infor-mation back to the CR transmitter. A coordinat-ing mechanism is required in multiple CRnetworks that try to access the same spectrum toprevent users colliding when accessing thematching spectrum holes. In interference-toler-ant CR networks, CR users can share the spec-trum resource with a licensed system whilekeeping the interference below a threshold. Incomparison with interference-free CR networks,interference-tolerant CR networks can achieveenhanced spectrum utilization by opportunisti-cally sharing the radio spectrum resources withlicensed users, as well as better spectral andenergy efficiency. However, it has been shownthat the performance of CR systems can be verysensitive to any slight change in user densities,interference threshold, and transmission behav-iors of the licensed system. This fact is illustratedin Fig. 3, where we notice that the spectral effi-ciency decreases quickly with the increase in thenumber of primary receivers. However, the spec-tral efficiency can be improved by either relaxingthe interference threshold of the primary systemor considering only the CR users who have shortdistances to the secondary BS. In [15], hybridCR networks have been proposed for adoptionin cellular networks to explore additional bandsand expand the capacity.

MOBILE FEMTOCELL

The MFemtocell is a new concept that has beenproposed recently to be a potential candidatetechnology in next generation intelligent trans-portation systems [12]. It combines the mobilerelay concept (moving network) with femtocelltechnology. An MFemtocell is a small cell thatcan move around and dynamically change itsconnection to an operator’s core network. Itcan be deployed on public transport buses,trains, and even private cars to enhance servicequality to users within vehicles. Deployment ofMFemtocells can potentially benefit cellularnetworks. First, MFemtocells can improve thespectral efficiency of the entire network. Todemonstrate this fact, Fig. 4 compares the aver-age spectral efficiency of the direct transmis-sion scheme and an MFemtocell-enhancedscheme with two resource partitioning schemes(i.e., orthogonal and non-orthogonal resourcepartitioning schemes) as a function of the per-centage of users associated with MFemtocells.Also, the comparison is done between maxi-mum signal-to-noise ratio (MAX-SNR) andproportional fairness (PF) scheduling algo-rithms. We can see that increasing the percent-age of users that communicate with the BSthrough MFemtocells leads to an increase inspectral efficiency, which is much better com-pared to the case where users communicatedirectly with the BS (i.e., the direct transmis-sion scheme). Second, MFemtocells can con-tribute to signaling overhead reduction of thenetwork. For instance, an MFemtocell can per-form a handover on behalf of all its associatedusers, which can reduce the handover activitiesfor users within the MFemtocell. This makesthe deployment of MFemtocells suitable forhigh-mobility environments. In addition, theenergy consumption of users inside an MFem-tocell can be reduced due to relatively shortercommunication range and low signaling over-head.

VISIBLE LIGHT COMMUNICATIONVisible light communication uses off-the-shelfwhite light emitting diodes (LEDs) used forsolid-state lighting (SSL) as signal transmittersand off-the-shelf p-intrinsic-n (PIN) photodi-odes (PDs) or avalanche photo-diodes (APDs)as signal receivers [10]. This means that VLCenables systems that illuminate and at the sametime provide broadband wireless data connectiv-ity. If illumination is not desired in the uplink,infrared (IR) LEDs or indeed RF would beviable solutions. In VLC, the information is car-ried by the intensity (power) of the light. As aresult, the information-carrying signal has to bereal valued and strictly positive. Traditional dig-ital modulation schemes for RF communicationuse complex valued and bipolar signals. Modifi-cations are therefore necessary, and there is arich body of knowledge on modified multi-carri-er modulation techniques such as OFDM forintensity modulation (IM) and direct detection(DD). Data rates of 3.5 Gb/s have been report-ed from a single LED. It has to be noted thatVLC is not subject to fast fading effects as thewavelength is significantly smaller than the


Figure 2. SM constellation diagram using four transmit antennas (NB = 4)and QPSK modulation.

01(11)

00(11)

11(11)

10(11)

Signal constellation for thefirst transmit antenna

.....

Signal constellation for thesecond transmit antenna

Spatial constellation

Re

Im

Re11

Im

10

01

00

01(00)

00(00)

11(00)

10(00)



detector area. While the link-level demonstra-tions are important steps to prove that VLC is aviable technique to help mitigate spectrum bot-tlenecks in RF communications, it is essential toshow that full-fledged optical wireless networkscan be developed by using existing lightinginfrastructures. This includes MU access tech-niques, interference coordination, and others.To this end, let us assume multiple light fixturesin a room. Each of these light fixtures is envis-aged to function as a very small optical BSresulting in a network of very small cells calledoptical attocells. This is in analogy to femtocellsin RF communications and in recognition of thefact that a single room can be served by many ofthese very small cells. An optical attocell coversan area of 1–10 m2 and distances of about 3 m.It is well known in cellular RF communicationsthat small cells have significantly contributed torecent improvements in network spectral effi-ciency. However, the main limiting factor isinterference. Optical attocells are less subject tointerference since lightwaves do not propagatethrough walls. The ratio of the area spectralefficiency (ASE) in bits per second per Hertzper square meter attained for the attocell net-work and the ASE for the femtocell network isillustrated in Fig. 5 against a varying number offemtocells per floor. The number of opticalaccess points per room varies from one to four.The gains diminish as the number of femtocellsper floor is increased as expected, but the gainis still above 100 for 20 femtocells per floor and4 optical attocells per room. The maximum gainin ASE is close to 1000. As an example, let usassume a typical ASE of 1.2 b/s/Hz/m2 for theoptical attocell network and a bandwidth of 10MHz for LED and RF. This would mean thatusers can on average share a total of 300 Mb/sin a 5 m × 5 m × 3 m room in the case of theoptical attocell network. In the case of the RFfemtocell network and the best case of 20 fem-tocells per floor, the attainable capacity for thesame room is only about 3 Mb/s.

GREEN COMMUNICATIONSThe design of 5G wireless systems should takeinto account minimizing the energy consump-tion in order to achieve greener wireless com-munication systems [5]. Wireless systemoperators around the world should aim toachieve such energy consumption reductions,which consequently contribute to the reductionof CO2 emissions. The indoor communicationtechnologies are promising deployment strate-gies to get better energy efficiency. This isbecause of the favorable channel conditionsthey can offer between the transmitters andreceivers. Moreover, by separating indoor trafficfrom outdoor traffic, the marcocell BS will haveless pressure in allocating radio resources andcan transmit with low power, resulting in a sig-nificant reduction in energy consumption. VLCand mm-wave technologies can also be consid-ered as energy efficient wireless communicationsolutions to be deployed in 5G wireless systems.For example, in VLC systems the consumedenergy in one bulb is much less than that in itsRF-based equivalents for transmitting the samehigh-density data.

FUTURE CHALLENGES IN5G WIRELESS COMMUNICATION

NETWORKS

Although there have been some developments inthe above potential key 5G wireless technolo-gies, there are still many challenges ahead. Dueto the limited space, in this section we only dis-cuss some of these challenges.

OPTIMIZING PERFORMANCE METRICSThe evaluation of wireless communication net-works has been commonly characterized by con-sidering only one or two performance metricswhile neglecting other metrics due to high com-

Figure 3. The average system spectral efficiency of a CR network as a functionof the number of primary receivers with different values of interference thresh-olds Q (number of secondary receivers = 20).

Number of primary receivers

Spec

tral

eff

icie

ncy

(b/s

/Hz)

1000

2

4

6

8

10

12

14

Q = 20 dB

Q = 10 dB

Q = –10 dB

2000 300 400 500

Simulation resultsAnalytical results

Figure 4. Average spectral efficiency of system-level MFemtocells with multi-user scheduling and resource partitioning schemes.

Percentage of users within MFemtocells

Ave

rage

spe

ctra

l eff

icie

ncy

(b/s

/Hz)

20% 40%0%2

2.5

3

3.5

4

4.5

5

5.5

6

60% 80% 100%

Direct transmission, MAX SINRSingle transmission, PFNon-orthogonal, MAX SINRNon-orthogonal, PFOrthogonal, MAX SINROrthogonal, PF



plexity. For a complete and fair assessment of5G wireless systems, more performance metricsshould be considered. These include spectralefficiency, energy efficiency, delay, reliability,fairness of users, QoS, implementation complexi-ty, and so on. Thus, a general framework shouldbe developed to evaluate the performance of 5Gwireless systems, taking into account as manyperformance metrics as possible from differentperspectives. There should be a trade-off amongall performance metrics. This requires high-com-plexity joint optimization algorithms and longsimulation times.

REALISTIC CHANNEL MODELS FOR5G WIRELESS SYSTEMS

Realistic channel models with proper accura-cy-complexity trade-off are indispensable forsome typical 5G scenarios, such as massiveMIMO channels and high-mobility channels(e.g., high-speed train channels and vehicle-to-vehicle channels). Conventional MIMO chan-nel models cannot be directly applied tomassive MIMO channels in which differentantennas may observe different sets of clus-ters. Massive MIMO channel models shouldtake into account specific characteristics thatmake them different from those in convention-al MIMO channels, such as the spherical wave-front assumption and non-stationaryproperties. Also, 3D massive MIMO models,which jointly consider azimuth and elevationangles, are more practical but more complicat-ed. Some existing massive MIMO channelmodels are briefly summarized and classifiedin Table 1 [14].

Compared to conventional low-mobility wire-less channels, high-mobility channels havegreater dynamics and possibly more severe fad-ing, and are essentially non-stationary. How tocharacterize non-stationary high-mobility chan-nels is also very challenging.

REDUCING SIGNAL PROCESSING COMPLEXITYFOR MASSIVE MIMO

One technical challenge in developing massiveMIMO systems is the signal processing complex-ity. As transmit and receive signals are quitelengthy, the search algorithms must be per-formed over many possible permutations of sym-bols. In the current literature, massive MIMOresearch is often treated as a detection problembased on a search motivated by the well-knownML criterion. The existing detection algorithmsassume that the channel has been perfectly esti-mated, which appears to be an unreasonableassumption given the size of the channel matrixand thus amount of channels to be tracked. Apossible solution to this problem is to apply theSM concept to massive MIMO systems. In thiscase, the spatial signature of each antenna needsto be different from the point of view of thereceiver because data is encoded into the choiceof transmit antenna active in the transmit array.It is therefore possible that channel estimationdoes not need to be exact but rather be merelysufficient to distinguish each transmit antenna.This may be a reasonable prospect, especially ifthe receive array is large, in which case eachtransmit antenna would have a quite detailedand thus distinct spatial signature.

INTERFERENCE MANAGEMENT FORCR NETWORKS

A major issue in interference-tolerant CR net-works in 5G is how to reliably and practicallymanage the mutual interference of CR and pri-mary systems. Regulating the transmit power isessential for the CR system to coexist with otherlicensed systems. An interference temperaturemodel is introduced for this purpose to charac-terize the interference from the CR to thelicensed networks. Interference cancellationtechniques should also be applied to mitigate theinterference at CR receivers. Another issue ininterference-tolerant CR networks is that a feed-back mechanism is important to periodicallyinform the CR network about the current inter-ference status at the licensed system. A practicalsolution is that the interference state informa-tion can be sent from licensed systems and col-lected by a central unit (or a third party system).Any CR network should first register to the cen-tral unit in order to be updated regarding theallowed spectrum and interference. Alternative-ly, the CR transmitters can listen to beacon sig-nals transmitted from the primary receivers andrely on the channel reciprocity to estimate thechannel coefficient. In this case, the CR trans-mitters can cooperate among themselves to reg-ulate the transmit power and prevent theinterference at the primary receivers being abovethe threshold.

CONCLUSIONSIn this article, the performance requirements of5G wireless communication systems have beendefined in terms of capacity, spectral efficiency,energy efficiency, data rate, and cell averagethroughput. A new heterogeneous 5G cellular

Figure 5. The ratio of ASE attained for the optical attocell network to ASE forthe femtocell network against varying numbers of femtocells per floor.

Number of femtocells per floor64

100

0

ASE

VLC

/ASE

RF

200

300

400

500

600

700

800

900

1000

8 10 12 14 16 18 20

1 AP1 AP, FL = 17 dB2 AP2 AP, FL = 17 dB4 AP4 AP, FL = 17 dB



architecture has been proposed with separatedindoor and outdoor applications using DAS andmassive MIMO technology. Some short-rangecommunication technologies, such as WiFi, fem-tocell, VLC, and mm-wave communication tech-nologies, can be seen as promising candidates toprovide high-quality and high-data-rate servicesto indoor users while at the same time reducingthe pressure on outdoor BSs. We have also dis-cussed some potential key technologies that canbe deployed in 5G wireless systems to satisfy theexpected performance requirements, such as CRnetworks, SM, MFemtocells, VLC, and greencommunications, along with some technical chal-lenges.

ACKNOWLEDGMENTThe authors gratefully acknowledge the supportof this work from SNCS Research Center, theUniversity of Tabuk, the Ministry of HigherEducation in Saudi Arabia, the Opening Projectof the Key Laboratory of Cognitive Radio andInformation Processing (Guilin University ofElectronic Technology), Ministry of Education(No. 2013KF01), Hutchison 3G UK, Natural Sci-ence Foundation of China (No. 61231009), theSTCSM (No. 11JC1412300), the InternationalScience & Technology Cooperation Program.

REFERENCES[1] Commission of the European Communities, Staff Work-

ing Document, “Exploiting the Employment Potential ofICTs,” Apr. 2012.

[2] Euro. Mobile Industry Observatory, GSMA, Nov. 2011.[3] A. Hashimoto, H. Yorshino, and H. Atarashi, “Roadmap

of IMT-Advanced Development,” IEEE Microwave Mag.,vol. 9, no. 4, Aug. 2008, pp. 80–88.

[4] WWRF, L. Sorensen and K. E. Skouby, User Scenarios2020, report, July 2009; http://www.wireless-world-research.org.

[5] C. Han et al., “Green Radio: Radio Techniques to EnableEnergy Efficient Wireless Networks,” IEEE Commun.Mag., vol. 49, no. 6, June 2011, pp. 46–54.

[6] Nokia Siemens Networks, “2020: Beyond 4G: Radio Evo-lution for the Gigabit Experience,” white paper, 2011.

[7] A. Bleicher, “Millimeter Waves May Be the Future of 5GPhones,” IEEE Spectrum, Aug. 2013.

[8] V. Chandrasekhar, J. G. Andrews, and A. Gatherer,“Femtocell Networks: A Survey,” IEEE Commun. Mag.,vol. 46, no. 9, Sept. 2008, pp. 59–67.

[9] F. Rusek et al., “Scaling Up MIMO: Opportunities andChallenges with Very Large Arrays,” IEEE Sig. Proc.Mag., vol. 30, no. 1, Jan. 2013, pp. 40–60.

[10] H. Haas, “Wireless Data from Every Light Bulb,” TEDwebsite, Aug. 2011; http://bit.ly/tedvlc

[11] X. Hong et al., “Secondary Spectrum Access Networks:Recent Developments on the Spatial Models,” IEEEVehic. Tech. Mag., vol. 4, no. 2, June 2009, pp. 36–43.

[12] F. Haider et al., “Spectral Efficiency Analysis of MobileFemtocell Based Cellular Systems,” Proc. IEEE ICCT ’11,Jinan, China, Sept. 2011, pp. 347–51.

[13] M. D. Renzo et al., “Spatial Modulation for GeneralizedMIMO: Challenges, Opportunities, and Implementation,”Proc. IEEE, vol. 102, no. 1, Jan. 2014, pp. 56–103.

[14] C.-X. Wang and S. Wu, “Massive MIMO Channel Mea-surements and Modeling: Advances and Challenges”IEEE Wireless Commun.., submitted for publication.

[15] X. Hong et al., “Capacity Analysis of Hybrid CognitiveRadio Networks with Distributed VAAs,” IEEE Trans.Vehic. Tech., vol. 59, no. 7, Sept. 2010, pp. 3510–23.

BIOGRAPHIESCHENG-XIANG WANG [S’01, M’05, SM’08] ([email protected]) received his Ph.D. degree from Aal-borg University, Denmark, in 2004. He joined Heriot-WattUniversity, United Kingdom, as a lecturer in 2005 andbecame a professor in August 2011. His research interestsinclude wireless channel modeling and 5G wireless commu-

nication networks. He has served or is serving as an Editoror Guest Editor for 11 international journals, including IEEETransactions on Vehicular Technology (2011–), IEEE Trans-actions on Wireless Communications (2007–2009), and IEEEJournal on Selected Areas in Communications. He has edit-ed one book, and published one book chapter and about190 papers in journals and conferences.

FOURAT HAIDER ([email protected]) received his Bachelor’sdegree in electrical and electronic engineering/communica-tion engineering from the University of Technology, Iraq, in2004, and his M.Sc. degree with Distinction at Brunel Uni-versity, United Kingdom, in 2009. Since December 2010, hehas been a Ph.D. student at Heriot-Watt University and theUniversity of Edinburgh, United Kingdom. His main researchinterests include spectral-energy efficiency trade-off, wire-less channel capacity analysis, femtocell and mobile femto-cell, and MIMO systems. He received the Best Paper Awardfrom IEEE ICCT 2011.

XIQI GAO [SM’07] ([email protected]) received his Ph.D.degree in electrical engineering from Southeast University,Nanjing, China, in 1997. He joined the Department ofRadio Engineering, Southeast University, in April 1992.Since May 2001, he has been a professor of informationsystems and communications. From September 1999 toAugust 2000, he was a visiting scholar at MassachusettsInstitute of Technology and Boston University. From August2007 to July 2008, he visited the Darmstadt University ofTechnology as a Humboldt scholar. His current researchinterests include broadband multicarrier communications,MIMO wireless communications, and signal processing forwireless communications.

XIAOHU YOU [SM’11, F’12] ([email protected]) received hisMaster’s and Ph.D. degrees from Southeast University,Nanjing, China, in electrical engineering in 1985 and 1988,respectively. Since 1990, he has been working with Nation-al Mobile Communications Research Laboratory at South-east University, where he held the rank of professor. Hisresearch interests include mobile communication systems,and signal processing and its applications. Now he is amember of the China 863 Expert Committee responsiblefor telecommunications.

YANG YANG [S’99, M’02, SM’10] ([email protected])received his Ph.D. degree in information engineering fromthe Chinese University of Hong Kong in 2002. He is cur-rently the director of the Shanghai Research Center forWireless Communications (WiCO) and a professor with theSchool of Information Science and Technology, Shang-haiTech University. His general research interests includewireless ad hoc and sensor networks, wireless mesh net-works, next generation mobile cellular systems, intelligenttransport systems, and wireless testbed development andpractical experiments.

DONGFENG YUAN [SM’01] ([email protected]) received hisM.Sc. degree from the Department of Electrical Engineer-ing, Shandong University, China, in 1988, and his Ph.D.degree from the Department of Electrical Engineering,Tsinghua University, China, in January 2000. Currently he isa full professor in the School of Information Science and

Table 1. Recent advances in massive MIMO channel models.

Channel model Complexity DescriptionReady formassiveMIMO?

Narrowband i.i.d.Rayleigh Low Uncorrelated model No

Narrowband CBSM(Weichselberger) Medium Jointly correlated model No

Narrowband CBSM(Kronecker) Medium Classic correlated model No

Wideband ellipticalGBSM [14] High Massive MIMO

properties considered Yes



Engineering, Shandong University, China. His currentresearch interests include cognitive radio systems, coopera-tive (relay) communications, and 5G wireless communica-tions.

HADI M. AGGOUNE [M’00, SM’09] ([email protected]) received his Ph.D. degree in electrical engi-neering from the University of Washington (UW). He is aProfessional Engineer registered in the State of Wash-ington, winner of the IEEE Professor of the Year Award,UW Branch, and l i s ted as an inventor in a patentassigned to Boeing. Currently he is a professor anddirector of the Sensor Networks and Cellular SystemsResearch Center, University of Tabuk, Saudi Arabia. Hisresearch interests include modeling and simulation oflarge-scale networks, sensors, visualization, and controland energy.

HARALD HAAS [S’98, AM’00, M’03] ([email protected]) holdsthe Chair of Mobile Communications in the Institute forDigital Communications at the University of Edinburgh. Hismain research interests are in the areas of wireless systemdesign and analysis as well as digital signal processing,with a particular focus on interference coordination in

wireless networks, spatial modulation, and optical wirelesscommunication. He holds more than 15 patents. He haspublished more than 50 journal papers, including a Sciencearticle, and more than 140 peer-reviewed conferencepapers.

SIMON FLETCHER ([email protected]) has respon-sibility for Products Strategy and Innovation for the emerg-ing communications infrastructure platforms of NEC’sGlobal Market product portfolio. He has core interests insustainability, 3GPP radio access technologies, and themanagement of innovation processes. As a director ofmobile VCE, a UK based research consortium, he acts asCEO and Chair of the Advisory Board. He also representsthe community as a director in the ICT-KTN.

EROL HEPSAYDIR ([email protected]) has been work-ing in the cellular mobile industry for 22 years. He haslaunched several mobile networks in Australia, Singapore,Korea, and the United Kingdom. Since 2001, he has beenresponsible for network technology. He is now head ofNetwork Technology for Hutchison 3G UK. He is currentlyworking on LTE and small cell deployment planning in theUnited Kingdom.
