research article resource allocation for vertical...

Hindawi Publishing CorporationInternational Journal of Antennas and PropagationVolume 2013 Article ID 456760 7 pageshttpdxdoiorg1011552013456760

Research ArticleResource Allocation for Vertical Sectorization inLTE-Advanced Systems

Lei Song Mugen Peng and Yan Li

Wireless Signal Processing and Network Lab Key Laboratory of Universal Wireless Communications (Ministry of Education)Beijing University of Posts and Telecommunications Beijing 100876 China

Correspondence should be addressed to Mugen Peng pmgbupteducn

Received 28 October 2013 Accepted 17 November 2013

Academic Editor Zhaobiao Lv

Copyright copy 2013 Lei Song et alThis is an open access article distributed under the Creative Commons Attribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Massive multiple input multiple output (MIMO) technology has been discussed widely in the past few years Three-dimensionalMIMO (3DMIMO) can be seen as a promising technique to realize massive MIMO to enhance the performance of LTE-Advancedsystems Vertical sectorization can be introduced by means of adjusting the downtilt of transmitting antennasThus the radiowavefromabase station (BS) to a group of user equipments (UE) can be divided into two beamswhich point at twodifferent areaswithin acell Intrasector interference is inevitable since the resources are overlapped In this paper the influence of intrasector interference isanalyzed and an enhanced resource allocation scheme for vertical sectorization is proposed as amethod of interference cancellationCompared with the conventional 2D MIMO scenarios cell average throughput of the whole system can be improved by verticalsectorization System level simulation is performed to evaluate the performance of the proposed scheme In addition the impactsof downtilt parameters and intersite distance (ISD) on spectral efficiency and cell coverage are presented

1 Introduction

Long Term Evolution (LTE) proposed by the Third Gen-eration Partnership Project (3GPP) is aimed at achievingenhanced performance in terms of cell throughput andsystem overhead As is generally known LTE-Advanced isaddressed to achieve higher data rates and betterUEsrsquoQualityof Service (QoS) than LTE [1 2] Based on OrthogonalFrequency Division Multiple Access (OFDMA) capacity ofwireless systems can be significantly increased by usingmultiantenna methods As one of the key techniques in LTE-Advanced systemsmassiveMIMOhas beenwidely discussedcurrently and 3D MIMO can be seen as an effective methodto approach large-scale MIMO [3ndash6] Compared with theconventional 2D antenna 3D antenna is comprised of manyelement units Each element is deployed at different heightsin an individual 3D antenna Thus difference of elements invertical domain should not be neglected

Benefiting from Active Antenna Systems (AAS) manyactive antenna technologies such as 3D beamforming andvertical sectorization could be easily implemented Vertical

sectorization can be realized by multiple antenna elementsby which two separate beams with their distinct antennaparameters are arranged to serve vertically split inner andouter cells In our work performance with various electricaltilt angles have been simulated to observe the optimizationspace of the overlap between two vertical sectors and the gainin terms of cell capacity

One of the major advantages of introducing verticalsectorization into LTE-Advanced systems is to improve cellcapacity In [7] capacity gain brought by electrical downtiltin standard 3-sector BS (Base Station) configuration could beup to 484 for 1500m intersite distance for 3G WCDMAsystems In [8] 457 capacity improvement was observedfor 3-sector BS with 1732m intersite distance by optimizingantenna parameters for LTE systems In [9] performanceof vertical sectorization with Near Electrical Tilt (NET)and Remote Electrical Tilt (RET) is evaluated for differentnetwork deployment scenarios by means of 3D system levelsimulation

By applying the separate antenna elements into transmit-ter each BS can dominate twice as many frequency resources

2 International Journal of Antennas and Propagation

Figure 1 vertical sectorization and 3 times 2 sector layout

as before The number of physical resource blocks (PRBs) isdoubled for vertical sectorization networks Severe intracellinterference will occur to reduce the signal to interferenceand noise ratio (SINR) of UEs and degrade performanceof the whole networks when the resources are scheduled todifferent groups of UEs at the same time In this paper anenhanced resource allocation scheme based on coordinatedcommunication is proposed Compared with the traditionalfrequency reuse scheme the proposed scheme is able toeliminate severe intracell interference and hence providebetter SINR of UE

The rest of this paper is organized as follows Section 2describes the system model of the vertical sectorization inLTE-Advanced systems and the deployment details Section 3proposes an enhanced scheduling scheme for vertical sec-torization and performance of the proposed scheme with

different scenario parameters is simulated and analyzed inSection 4 The conclusion of this paper is given in Section 5

2 System Model

Since a cell has 3 sectors and each sector only has oneBS in theregular systems the LTE-Advanced systems without verticalsectorization are referred to as 3 times 1 networks The LTE-Advanced systems with vertical sectorization are referred toas 3 times 2 networks since each sector is partitioned into 2subsectors as is shown in Figure 1

It is assumed that the total transmission power of a sectorof 3 times 2 networks remains the same as 3 times 1 networks andthe power is equally split between two subsectors The signallobe radiated by the antenna is split into two lobes withdifferent downtilts The first lobe serving the UEs near BS isgenerated by the bottom row of antenna elements by usinga higher downtilt (ie NET) while the second one servingthe UEs close to cell border is generated by the top row ofantenna elements using a lower downtilt (ie RET)The samesystem bandwidth is used for both subsectors The amountof resources is doubled but the power per physical resourceblock (PRB) is halved

Supposing that there are total119872 hexagonal cells with 119870active UEs served within the coverage of each sector Thetotal number of BSs in the system is 3119872 Each BS whichcorresponds to one sector is equipped with 119873

119905transmitting

antennas while each UE has119873119903receiving antennas For each

sector divided into inner sector and outer sector the bottomand the top rows of antenna elements can be seen as innersector and outer sector antenna elements respectively Thereare two groups of UEs in one sector inner sector UEs andouter sector UEs which means that UEs are located near theserving BS and far away from the serving BS respectivelyThus taking inner sector UE for example the received signalat the 119896th UE served by BS119898 is as follows

y(119896)119898-inner = H

(119896)

119898-innerW(119896)

119898-inners(119896)

119898-inner⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

desired signal

+H(119896)119898-outerW

(119896)

119898-outers(1198961015840)

119898-outer⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

intrasector interference

+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119896)119899-innerW

(119896)

119899-inners(119908)

119899-inner +3119872

sum

119899 =119898

119870

sum

119908=1

H(119896)119899-outerW

(119896)

119899-outers(119908)

119899-outer⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

inter-sector interference

+ n(119896)119898

(1)

where s(119896)119898-inner is the 119897 times 1 transmitted vector for inner sector

UE 119896 served by BS 119898 and 119897 is the number of data streamss(119896)119898-inner = [radic119901119896119904

(119896)

1198981 radic119901119896119904

(119896)

1198982 radic119901119896119904

(119896)

119898119897]119879

and 119901119896is the

transmit power of each data stream at UE 119896 It is assumedthat each transmitting antenna has 119873V elements in verticaldimension and the receiving antennas are still conventional2D antennas Thus H(119896)

119898-inner is the (119873119903 times 119873119905) channel matrixfrom the inner sector antenna elements in BS 119898 to UE 119896and W(119896)

119898-inner is the (119873119905 times 119897) precoding matrix n(119896)119898

is theadditive white Gaussian noise with zero mean and varianceE(n(119896)119898n(119896)119867119898) = 1205902

3 Proposed Scheme for VerticalSectorization

31 Intracell Interference Brought by Vertical SectorizationFor 3 times 1 networks detailed analysis of detected SINR for thereceived signal is in [10] Due to the introduction of verticalsectorization as shown in Figure 2 the received power ofUE is halved and the source of interference is much morecomplex than before The expression of the SINR can bedepicted as follows

International Journal of Antennas and Propagation 3

SignalWeak interferenceStrong interference

Figure 2 Limitation of vertical sectorization

User 1 User 2 User 3 User 4

Severe inner-outer cell interference

User 5 User 6 User 7

Inner cell resource block

Outer cell resource block

Figure 3 Intrasector interference brought by vertical sectorization

SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W

sum3119872

119899=0 119899 =119898119875119896

119871W119867H119896

119899(H119896119899)119867W⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 1198751198961015840

119871W119867H119896

119898-outer(H119896

119898-outer)119867W⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(2)

where 119875119896119871= 119875119896

119905119909119875119896

loss2 119875noise = 1205902W119867W

32 Previous Work on Resource Allocation For a sector invertical sectorization networks the inner sector and outersector independently schedule radio resources for the UEsin their respective dominance areas Figure 3 illustrates anexample of a possible scheduling condition The signalsbecomemutual interference tomostUEs and such schedulingconfliction could happen all the time Furthermore the pathlosses experienced by UEs belonging to inner sector orouter sector are basically identical because the two beamsare radiated from the same active antenna Therefore thereceived signal and interference are mainly determined bythe vertical antenna gains Thus UEs in the area where twovertical antenna patterns overlap (defined as sector edge UEsin this paper) will suffer severe Intrasector interference

Fortunately the same active antenna serving inner sectorand outer sector gives us the opportunity to avoid Intrasectorinterference through cooperative scheduling The previousscheduling scheme [11] is shown in Figure 4 Instead ofscheduling resources independently by the inner sector andouter sector the radio resources are scheduled orthogonallyto UEs in the same manner of that in the 3 times 1 networks Inthis case the SINR of UEs would be improved largely

33 Proposed Resource Allocation Scheme As is analyzed inthe previous subsection without proper interference can-cellation mechanism the utilization of vertical sectorizationwould worsen the performance of sector-edge UEs In orderto provide better service for these UEs an enhanced resourceallocation scheme is presented in Figure 5 UE A and UEB are both sector-edge UEs as mentioned above Based oncoordinated communication the same part of resources fromthe two subsectors can be allocated to UE A Another similar

part of resources from the two subsectors orthogonal tothose allocated to UE A can be allocated to UE B In thiscase for each UE not only the Intrasector interference iseliminated but the power of signal is also doubled Themutual interference between sector-edge UEs can be avoidedby frequency orthogonality The specific methods to achievethe scheme are as follows

Step 1 Divide the UEs accessed in vertical sectorizationnetworks into three sets sector-edge UE set inner sector UEset and outer sector UE set Reference Signal Received Power(RSRP) principle is utilized to decide whether the UE belongsto sector-edgeUE set If the difference of RSRP between innersector and outer sector is less than a settled threshold the UEcan be recognized as a sector-edge UE The judging criteriacan be expressed as follows

1003816100381610038161003816119875BS times 119866loss times 119866shadow times 119866inner-antenna minus 119875BS

times 119866loss times 119866shadow times 119866outer-antenna1003816100381610038161003816 ltThr

(3)

Step 2Divide thewhole resources in each sector into two sets119866 and 119865 where 119866 cap 119865 = 0 Resources in set 119866 are used forserving sector-edge UEs while resources in set 119865 are used forinner sector UEs and outer sector UEsStep 3 For a sector judge whether sector-edge UEs exist inthe current scheduling period If the number of sector-edgeUEs 119873sector-edgeUE gt 0 both inner sector and outer sectorassign the PRBs in set 119866 to sector-edge UEs After that theresources in 119865 are allocated to other UEs Otherwise for nosector-edge UE case PRBs in119866 and 119865 are allocated to all UEsin the sector without any restriction



No inner-outer cell interference




Figure 4 Previous scheduling scheme for vertical sectorization

Sector edge UE

UE AUE B

Figure 5 Proposed resource allocation scheme

Step 4 Apply beamforming algorithm as a manner of jointprecoding for sector-edge UEs Take sector-edge UE 119894for example h119894inner and h119894outer are the channels generatedby inner sector and outer sector antenna elements respec-tively The equivalent channel from sector 119898 to UE 119894 is

defined as Heff where Heff = [h119894innerh119894

outer] V119894119898sim

SVD(Heff) V119894119898

is the V matrix from SVD decom-position of Heff The beamforming vector v

119894119898is the

first column of matrix V119894119898 The received signal can be

derived

y(119894)119898= H(119894)119890119891119891

k119894119898s(119894)119898⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

desired signal

+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-innerW

(119894)

119898-inners(119908)

119898+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-outerW

(119894)

119898-outers(119908)

119898

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ n(119894)119898 (4)

Thus the SINR of sector-edge UE 119894 can be expressed asfollows

SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W + 119875119896119871W119867H119896

119898-outer(H119896

119898-outer)119867

W

sum3119872

119899=0 119899 =119898119875119871W119867H119896

119899-inner(H119896

119899-inner)119867

W + sum3119872119899=0 119899 =119898 119875119871W119867H119896119899-outer(H

119896

119899-outer)119867W

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(5)

4 Simulation Results

In this section system level performance of the proposedresource allocation scheme is evaluated in terms of cellaverage spectral efficiency and SINR of the UEs Detailedsimulation parameters are listed in Table 1

Figure 6(a) is the 2 times 2 cell average spectral efficiency ofdifferent downtilt configurations with sector radius of 288mFor these downtilt configurations the performance of 3D ver-tical sectorization has 16 27 15 and 26 enhancementcompared to 2D MIMO networks respectively Althoughsystem capacity gain is decreased due to the orthogonality


20

18

16

14

12

10

08

06

04

02

00

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(a)

150

135

120

105

090

075

060

045

030

015

000

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(b)

2624222018161412100806040200

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

RET and NET

2D MIMO3D vertical sectorization

Previous schemeProposed scheme

(c)

222018161412100806040200

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET



(d)

Figure 6 System level results of 2D MIMO 3D vertical sectorization previous scheme and proposed scheme (a) 2 times 2 cell average spectralefficiency with sector radius of 288m (b) 2 times 2 cell average spectral efficiency with sector radius of 288m (c) 4 times 2 cell average spectralefficiency with sector radius of 500m (d) 4 times 2 cell average spectral efficiency with sector radius of 500m

of resources there is still a substantial improvement in theperformance of throughput Since coordinated transmissionis applied into our proposed scheme UEs located at thesector edge area are allocated twice as many resources asbefore As can be seen in the figure the proposed scheme isable to achieve an obvious gain in throughput compared tothe previous resource allocation scheme About 29 2824 and 25 of the gain compared to 2D MIMO can beachieved by the enhanced scheme with different RET andNET configurations

The cell average spectral efficiency of 2times2 500m scenariois shown in Figure 6(b) Since the radius of sector increasesthe distribution of UEs within a sector changes a lot and

the elevation angles of UEs in the area decrease The sameRET and NET configuration cannot cover as many UEs as in288m scenario Although the performance of coverage is notso good as Figure 6(a) the proposed scheme still achieves aperformance gain compared to the previous scheme and 2DMIMO networks Cell average spectral efficiency has utmost8 enhancement compared to 2D MIMO networks

The cell average spectral efficiency of 4 times 2 288mscenario is shown in Figure 6(c) For each downtilt config-uration 3D vertical sectorization can achieve more than 30enhancement compared with 2DMIMO networks Since theresources of inner sector and outer sector are allocated toUEs in an orthogonal manner the throughput of cell average




1

09

08

07

06

05

04

03

02

01

0minus30 minus20 minus10 0 10 20 30 40

SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)

Figure 7 System level results of UEsrsquo receive SINR CDF for 2D MIMO 3D vertical sectorization previous scheme and proposed scheme(a) 2 times 2 288m scenario (b) 2 times 2 500m scenario



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)


UEs decreases By applying the enhanced resource allocationthe proposed scheme outperforms the previous scheme in allthe scenarios It can be seen that as the increasing of sectorradius there are not too many sector edge UEs existing sothe throughput gain brought by the proposed scheme is notso obvious in the 500m scenario as shown in Figure 6(d)

Figure 7 shows the UEsrsquo receiving SINR CDF of 2DMIMO 3D vertical sectorization the previous scheme andthe proposed scheme For 2times2 288m scenario and 4times2 288mscenario performance of 3D vertical sectorization descendsapproximately 7 dB compared with that of 2D MIMO dueto the introduction of sector interference The average SINR


Table 1 Simulation configuration parameters

Parameters AssumptionsCell type MacrocellCellular layout 19 cells each with 3 sectorsSector radius 288m 500mBandwidth 5MChannel model WINNER II

Antenna number Transmitter 2Tx 4TxReceiver 2Rx

Antenna configuration Copolarized linear array

Antenna spacing Transmitter 40120582Receiver 05120582

Number of elevation elements 2Regular antenna gain 14 dBiBS max TX power 46 dBmBS height 32mUE height 15mUE numbers 10UEs per sectorScheduler Round-robinReceiver MMSETraffic model Full buffer

of the previous scheme outperforms that of 3D verticalsectorization by 5 dB Since sector edge UEs in the proposedscheme can be served much better than before the averageSINR of the proposed scheme has 1 dB gain over that of theprevious scheme

When the sector radius is increased UEs are locatedfarther away to each other and sector edge UEsrsquo impacts onsystem performance are reduced largely As is depicted inFigure 8 theUEsrsquo average SINRof the two resource allocationschemes is less than that of 2D MIMO In spite of thisthe SINR of UEs in the proposed scheme and the previousscheme still outperform that of 3D vertical sectorization

5 Conclusion

In this paper modeling of active antenna and structure ofvertical sectorization have been introduced Compared with2D MIMO scheme vertical sectorization can enhance thecell throughput significantly butwill suffer serious Intrasectorinterference To solve this problem the interference sufferedby UEs is analyzed and an enhanced resource allocationscheme for vertical sectorization in LTE-Advanced systemsis proposed System level simulation has been performedto evaluate the performance of the proposed scheme Sim-ulation results can prove that it is an effective method toeliminate intracell interference and increase system capacityBeyond that simulation results for vertical sectorization withdifferent downtilt parameters and sector radius parametersare presented which indicate that the proposed scheme canimprove the SINR of sector-edge UEs and maintain the cellaverage throughput gain

Acknowledgments

This work was supported in part by the State Major ScienceandTechnology Special Projects (Grant no 2012ZX03001028-004 2013ZX03001001 and 2012ZX03001031) the NationalNatural Science Foundation ofChina (Grant no 61222103 no61072058) the BeijingNatural Science Foundation (Grant no4131003) and the Program forNewCentury Excellent Talentsin University (NCET-10-0262)

References

[1] M Peng and W Wang ldquoTechnologies and standards for TD-SCDMA evolutions to IMT-advancedrdquo IEEE CommunicationsMagazine vol 47 no 12 pp 50ndash58 2009

[2] M Peng D Liang Y Wei J Li and H Chen ldquoSelf-config-uration and self-optimization in LTE-advanced heterogeneousnetworksrdquo IEEE Communications Magazine vol 51 no 5 pp36ndash45 2013

[3] F Rusek D Persson B K Lau et al ldquoScaling upMIMO oppor-tunities and challenges with very large arraysrdquo IEEE Signal Proc-essing Magazine vol 30 no 1 pp 40ndash60 2013

[4] H Huh G Caire H C Papadopoulos and S A RamprashadldquoAchieving ldquomassive MIMOrdquo spectral efficiency with a not-so-large number of antennasrdquo IEEE Transactions onWireless Com-munications vol 11 no 9 pp 3226ndash3239 2012

[5] M Peng X ZhangWWang andH-H Chen ldquoPerformance ofdual-polarized MIMO for TD-HSPA evolution systemsrdquo IEEESystems Journal vol 5 no 3 pp 406ndash416 2011

[6] T Zhou M Peng W Wang and H Chen ldquoLow-complex-ity coordinated beamforming for downlink multi-cell SDMAOFDMSystemrdquo IEEETransactions onVehicular Technology vol62 no 1 pp 247ndash255 2013

[7] J Niemela T Isotalo and J Lempiainen ldquoOptimum antennadowntilt angles formacrocellularWCDMAnetworkrdquoEURASIPJournal onWireless Communications and Networking vol 2005Article ID 610942 2005

[8] O N C Yilmaz S Hamalainen and J Hamalainen ldquoAnalysisof antenna parameter optimization space for 3GPP LTErdquo inProceedings of the 70th Vehicular Technology Conference Fall(VTC rsquo09-Fall) pp 1ndash5 Anchorage Alaska USA September2009

[9] O N C Yilmaz S Hamalainen and J Hamalainen ldquoSystemlevel analysis of vertical sectorization for 3GPPLTErdquo inProceed-ings of the 6th International Symposium on Wireless Com-munication Systems (ISWCS rsquo09) pp 453ndash457 Tuscany ItalySeptember 2009

[10] F Shu L Lihua C Qimei and Z Ping ldquoNon-unitary codebookbased precoding scheme for multi-user MIMO with limitedfeedbackrdquo in Proceedings of the IEEE Wireless Communicationsand Networking Conference (WCNC rsquo08) pp 678ndash682 LasVegas Nev USA April 2008

[11] M Peng W Wang and H-H Chen ldquoMultiuser pairing-upschemes under power constraints for uplink virtual MIMOnetworksrdquo Mobile Networks and Applications vol 15 no 2 pp298ndash309 2010

International Journal of

AerospaceEngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

RoboticsJournal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Active and Passive Electronic Components

Control Scienceand Engineering

Journal of



RotatingMachinery


Hindawi Publishing Corporation httpwwwhindawicom

Journal ofEngineeringVolume 2014

Submit your manuscripts athttpwwwhindawicom

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics


Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

SensorsJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation



DistributedSensor Networks



Figure 1 vertical sectorization and 3 times 2 sector layout

as before The number of physical resource blocks (PRBs) isdoubled for vertical sectorization networks Severe intracellinterference will occur to reduce the signal to interferenceand noise ratio (SINR) of UEs and degrade performanceof the whole networks when the resources are scheduled todifferent groups of UEs at the same time In this paper anenhanced resource allocation scheme based on coordinatedcommunication is proposed Compared with the traditionalfrequency reuse scheme the proposed scheme is able toeliminate severe intracell interference and hence providebetter SINR of UE

The rest of this paper is organized as follows Section 2describes the system model of the vertical sectorization inLTE-Advanced systems and the deployment details Section 3proposes an enhanced scheduling scheme for vertical sec-torization and performance of the proposed scheme with

different scenario parameters is simulated and analyzed inSection 4 The conclusion of this paper is given in Section 5

2 System Model

Since a cell has 3 sectors and each sector only has oneBS in theregular systems the LTE-Advanced systems without verticalsectorization are referred to as 3 times 1 networks The LTE-Advanced systems with vertical sectorization are referred toas 3 times 2 networks since each sector is partitioned into 2subsectors as is shown in Figure 1

It is assumed that the total transmission power of a sectorof 3 times 2 networks remains the same as 3 times 1 networks andthe power is equally split between two subsectors The signallobe radiated by the antenna is split into two lobes withdifferent downtilts The first lobe serving the UEs near BS isgenerated by the bottom row of antenna elements by usinga higher downtilt (ie NET) while the second one servingthe UEs close to cell border is generated by the top row ofantenna elements using a lower downtilt (ie RET)The samesystem bandwidth is used for both subsectors The amountof resources is doubled but the power per physical resourceblock (PRB) is halved

Supposing that there are total119872 hexagonal cells with 119870active UEs served within the coverage of each sector Thetotal number of BSs in the system is 3119872 Each BS whichcorresponds to one sector is equipped with 119873

119905transmitting

antennas while each UE has119873119903receiving antennas For each

sector divided into inner sector and outer sector the bottomand the top rows of antenna elements can be seen as innersector and outer sector antenna elements respectively Thereare two groups of UEs in one sector inner sector UEs andouter sector UEs which means that UEs are located near theserving BS and far away from the serving BS respectivelyThus taking inner sector UE for example the received signalat the 119896th UE served by BS119898 is as follows

y(119896)119898-inner = H

(119896)

119898-innerW(119896)

119898-inners(119896)

119898-inner⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

desired signal

+H(119896)119898-outerW

(119896)

119898-outers(1198961015840)

119898-outer⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119896)119899-innerW

(119896)

119899-inners(119908)

119899-inner +3119872

sum

119899 =119898

119870

sum

119908=1

H(119896)119899-outerW

(119896)

119899-outers(119908)

119899-outer⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ n(119896)119898

(1)

where s(119896)119898-inner is the 119897 times 1 transmitted vector for inner sector

UE 119896 served by BS 119898 and 119897 is the number of data streamss(119896)119898-inner = [radic119901119896119904

(119896)

1198981 radic119901119896119904

(119896)

1198982 radic119901119896119904

(119896)

119898119897]119879

and 119901119896is the

transmit power of each data stream at UE 119896 It is assumedthat each transmitting antenna has 119873V elements in verticaldimension and the receiving antennas are still conventional2D antennas Thus H(119896)

119898-inner is the (119873119903 times 119873119905) channel matrixfrom the inner sector antenna elements in BS 119898 to UE 119896and W(119896)

119898-inner is the (119873119905 times 119897) precoding matrix n(119896)119898

is theadditive white Gaussian noise with zero mean and varianceE(n(119896)119898n(119896)119867119898) = 1205902

3 Proposed Scheme for VerticalSectorization

31 Intracell Interference Brought by Vertical SectorizationFor 3 times 1 networks detailed analysis of detected SINR for thereceived signal is in [10] Due to the introduction of verticalsectorization as shown in Figure 2 the received power ofUE is halved and the source of interference is much morecomplex than before The expression of the SINR can bedepicted as follows










SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W

sum3119872

119899=0 119899 =119898119875119896

119871W119867H119896

119899(H119896119899)119867W⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 1198751198961015840

119871W119867H119896

119898-outer(H119896

119898-outer)119867W⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(2)

where 119875119896119871= 119875119896

119905119909119875119896

loss2 119875noise = 1205902W119867W








(3)









Sector edge UE

UE AUE B




outer] V119894119898sim

SVD(Heff) V119894119898


119894119898is the


derived

y(119894)119898= H(119894)119890119891119891

k119894119898s(119894)119898⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

desired signal

+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-innerW

(119894)

119898-inners(119908)

119898+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-outerW

(119894)

119898-outers(119908)

119898

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ n(119894)119898 (4)


SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W + 119875119896119871W119867H119896

119898-outer(H119896

119898-outer)119867

W

sum3119872

119899=0 119899 =119898119875119871W119867H119896

119899-inner(H119896

119899-inner)119867

W + sum3119872119899=0 119899 =119898 119875119871W119867H119896119899-outer(H

119896

119899-outer)119867W

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(5)





20

18

16

14

12

10

08

06

04

02

00

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(a)

150

135

120

105

090

075

060

045

030

015

000

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(b)

2624222018161412100806040200

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

RET and NET



(c)

222018161412100806040200

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET



(d)









1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)




1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)













5 Conclusion


Acknowledgments


References














RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation


















SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W

sum3119872

119899=0 119899 =119898119875119896

119871W119867H119896

119899(H119896119899)119867W⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 1198751198961015840

119871W119867H119896

119898-outer(H119896

119898-outer)119867W⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(2)

where 119875119896119871= 119875119896

119905119909119875119896

loss2 119875noise = 1205902W119867W








(3)









Sector edge UE

UE AUE B




outer] V119894119898sim

SVD(Heff) V119894119898


119894119898is the


derived

y(119894)119898= H(119894)119890119891119891

k119894119898s(119894)119898⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

desired signal

+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-innerW

(119894)

119898-inners(119908)

119898+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-outerW

(119894)

119898-outers(119908)

119898

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ n(119894)119898 (4)


SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W + 119875119896119871W119867H119896

119898-outer(H119896

119898-outer)119867

W

sum3119872

119899=0 119899 =119898119875119871W119867H119896

119899-inner(H119896

119899-inner)119867

W + sum3119872119899=0 119899 =119898 119875119871W119867H119896119899-outer(H

119896

119899-outer)119867W

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(5)





20

18

16

14

12

10

08

06

04

02

00

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(a)

150

135

120

105

090

075

060

045

030

015

000

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(b)

2624222018161412100806040200

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

RET and NET



(c)

222018161412100806040200

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET



(d)









1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)




1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)













5 Conclusion


Acknowledgments


References














RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation
















Sector edge UE

UE AUE B




outer] V119894119898sim

SVD(Heff) V119894119898


119894119898is the


derived

y(119894)119898= H(119894)119890119891119891

k119894119898s(119894)119898⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟

desired signal

+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-innerW

(119894)

119898-inners(119908)

119898+

3119872

sum

119899 =119898

119870

sum

119908=1

H(119894)119898-outerW

(119894)

119898-outers(119908)

119898

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ n(119894)119898 (4)


SINR119896 =119875119896

119871W119867H119896

119898-inner(H119896

119898-inner)119867

W + 119875119896119871W119867H119896

119898-outer(H119896

119898-outer)119867

W

sum3119872

119899=0 119899 =119898119875119871W119867H119896

119899-inner(H119896

119899-inner)119867

W + sum3119872119899=0 119899 =119898 119875119871W119867H119896119899-outer(H

119896

119899-outer)119867W

⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟⏟


+ 119875noise

(5)





20

18

16

14

12

10

08

06

04

02

00

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(a)

150

135

120

105

090

075

060

045

030

015

000

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(b)

2624222018161412100806040200

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

RET and NET



(c)

222018161412100806040200

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET



(d)









1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)




1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)













5 Conclusion


Acknowledgments


References














RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation










20

18

16

14

12

10

08

06

04

02

00

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(a)

150

135

120

105

090

075

060

045

030

015

000

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET

(b)

2624222018161412100806040200

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

RET and NET



(c)

222018161412100806040200

Cell

aver

age s

pect

rum

effici

ency

(bit

sH

z)

10∘18

∘10

∘22

∘22

∘8∘

8∘18

∘

RET and NET



(d)









1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)




1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)













5 Conclusion


Acknowledgments


References














RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation












1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)




1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(a)



1

09

08

07

06

05

04

03

02

01


SINR (dB)

CDF

(b)













5 Conclusion


Acknowledgments


References














RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation


















5 Conclusion


Acknowledgments


References














RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation











RoboticsJournal of





Journal of



RotatingMachinery





VLSI Design



Shock and Vibration







Journal of



Volume 2014


SensorsJournal of





Propagation









research article resource allocation for vertical...

Documents