Decentralized Femtocell Transmission Regulation in Spectrum-Sharing Macro and Femto Networks Xiaoli Chu King’s College London, UK OPTNet 2011, Sheffield,

Decentralized Femtocell Transmission Regulation in

Spectrum-Sharing Macro and Femto Networks

Xiaoli ChuKing’s College London, UK

OPTNet 2011, Sheffield, 14 September 2011

- 2 -

Outline• Introduction

• Collocated spectrum-sharing macro and femto cells▫ Motivation

▫ Contribution

▫ System model

▫ Outage probability analysis

▫ Femtocell location and transmit power

• Simulation results▫ Analytical results verified by simulations

• Conclusion

Introduction

- 4 -

Business opportunities

New user terminals New applicationsNew markets

- 5 -

Technical challenges

Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2010–2015

• Current 2G and 3G networks will not be able to meet future mobile data traffic demands

• Most of the data traffic is performed indoors, where coverage is the worst

• As a result, vendors and operators are desperately looking for new solutions

- 6 -

Solutions: Femtocells• Femtocells are low-power wireless access points (FAPs) that

operate in licensed spectrum to connect standard mobile devices to a mobile operator’s network using residential DSL or cable broadband connections [Source: Femto Forum].▫ Improve indoor coverage▫ Unload traffic from overburdened macrocells▫ Likely to be user-deployed

Collocated Spectrum-Sharing Macro and Femto Cells

- 8 -

Motivation• Spectrum-sharing macro and femto cells

▫ Benefits Spectrum-sharing allows for increased spectral efficiency and better spatial reuse

▫ Challenges Spectrum-sharing suffers from inter-cell interference and creates dead spots where UE QoS cannot be guaranteed.

- 9 -

Contribution• Analysis of downlink (DL) outage probabilities (OPs)

▫ Closed-form macro and femto DL OP lower bounds embracing the randomness of transmit power employed by different interfering FAPs.

• Analysis includes both Rayleigh flat fading and shadowing ▫ Our work accounts for path loss, Rayleigh fading, lognormal (LN)

shadowing, and LN interfering FAP power, and allows different DL (SIR) targets and OP constraints for macro and femto cells.

• Decentralized resource allocation▫ Decentralized strategy to regulate FAP’s transmit power and

usage of radio resources to guarantee a satisfactory macro and femto DL coverage.

- 10 -

• OFDMA downlink of collocated spectrum-sharing macrocell and closed-access femtocells

▫ A central MBS covers a disc area with radius rM

▫ Femtocells of radius rF are randomly distributed on R2 as a spatial Poisson point process (SPPP) with a density of F.

▫ NF femtocells per cell site on average

▫ UF indoor UEs per femtocell, each located on femtocell edge

▫ MBS transmit power PM,Tx is evenly distributed among RBs

▫ FAP transmit power PF,Tx is evenly distributed among RBs

▫ Each FAP transmits with a probability within an RB.

▫ Spatial intensity of co-channel FAPs is uF = F.

▫ Macro-to-macro interference and thermal noise are ignored.

System Model

MBS

rM

MUE

Femto coverage circle

Macro coveragecircle

FAP

rF

FUE

R2

PM,Tx

PF,Tx

- 11 -

Channel Model• Path loss follows the IMT-2000 channel model

• fc is the carrier frequency in MHz, d is the distance of the link, and denotes the wall-penetration loss.

• Each frequency subchannel sees Rayleigh flat fading and lognormal shadowing

- 12 -

Femtocell DL SIR• The received SIR of an indoor FUE at the femtocell edge

▫ PF = PF,TxGFAPGUE, PM = PM,TxGMBSGUE;

▫ DFM is the distance from the MBS to the FUE, DFFi is the distance from interfering FAP i to the FUE;

▫ HF, HFM and HFFi are unit-mean exponential channel power gains;

▫ QF ~ LN(F, 2F2), QFM ~ LN(FM, 2FM

2) and QFFi ~ LN(FF, 2FF2)

denote lognormal shadowing, = 0.1ln10;

▫ is the set of FAPs having access to the given RB, with intensity uF.

iiii DQHPDQHP

rQHPFFFM

F

FFFFFF1

FFFFMFMFM1

FMM

FFF1

FFFSIR

macro intef femto intef

- 13 -

Femto Outage Probability• Outage probability of an indoor FUE w.r.t. the target SIR F

• For an indoor FUE at a distance dFM from the MBS

• Based on the stochastic geometry theory

FFM

FFFF

FM

FF

FFFFFF1

FFFFM

FFF ,SIRPPPSIRP

FF

I

S

I

S

DQHPI

S

iiii

N

n

M

mb

mn

bbamn

m

mmn

eba

eeuvw

dP

rPFdD

1 1~

2~2

F

~~22FF

2FM

2FFMF

FMFMF

FFFMFMFMFF

~2

exp1

~~,~~;SIRP

FFFMFMFF

FM

F

Prob of macro-to-femto interf. being strong enough to

create outage

Prob of femto-to-femto and macro-to-femto interf. causing outage

- 14 -

Macrocell DL SIR• The received SIR of an outdoor MUE is

▫ DM is the distance from the MBS to the MUE, DMFi is the distance from FAP i to the MUE;

▫ HM and HMFi denote unit-mean exponential channel power gains;

▫ QM ~ LN(M, 2M2) and QMFi ~ LN(MF, 2MF

2) denote lognormal shadowing.

iiii DQHP

DQHPMF

M

MFMFMF1

MFF

MMM1

MMMSIR

femto intef

- 15 -

Macro Outage Probability• Outage probability of an MUE w.r.t. the target SIR M

• For an MUE at a distance dM from the MBS

• Based on the stochastic geometry theory

MMFMFMF

1MFF

MMM MF

PSIRP

iiii DQHP

S

M

m

mm bu

vdD

1 MF

M

MF

MFMMMMM

2~22expexp1SIRP

2MF

2MF

MF

MF

2

MF

MFM

~2~2exp

MF

P

- 16 -

Minimum MBS-to-FAP Distance

• P(SIRF < F) ≤ F and P(SIRM < M) ≤ M, where 0 ≤ F, M < 1

• P(SF/IFM < F|DFM = dFM) is a monotonically decreasing function of dFM.

• Minimum dFM required for P(SIRF < F|DFM = dFM) ≤ F

▫ = HFQF/(HFMQFM) approximately follows a LN

distribution

• Any UE located less than dFM,min from the MBS

should be associated with the macrocell.

FM

F

1

FFFM

2FM

2FFMFF

1FMF

minFM,

~~,~~;

rP

FPd

MBS

FAP

rF

FUE

dFM,min

rM

- 17 -

FAP Transmit Power• Femtocells’ transmit power should be within the range

[PF,Tx,min, PF,Tx,max]

• PF,Tx,max is delimited by network standard.

• PF,Tx,min is chosen as the minimum PF,Tx that makes an FUE at the macrocell edge meet Pr(SF/IFM<F|DFM= rM) ≤F.

where is the inverse CDF of the LN RV evaluated at F.

2FM

2FFMFF

1MFMFAP

FFFMBSTxM,minTx,F, ~~,~~;FM

F

FrG

rGPP

2FM

2FFMFF

1 ~~,~~; F

- 18 -

FAP Self-Regulation• FAP at a distance d (dFM,min ≤ d ≤ rM) from the MBS,

▫ For an RB, if P(LB)F,Tx(d) min{P(UB)

F,Tx(d), PF,Tx,max}, then the FAP can transmit in the RB with PF,Tx set in the range [P(LB)

F,Tx(d), min{P(UB)

F,Tx(d), PF,Tx,max}] for simultaneously meeting both the macro and femto DL OP constraints;

▫ otherwise, the FAP can only transmit in the RB with P(LB)F,Tx(d)

and at a reduced probability.

2

FM2FFMFF

1FMFAP

FFFMBSTxM,LBTxF, ~~,~~;FM

F

FdG

rGPdP

2MF

dBmmaxTx,F,dBmMUB

maxTx,F,dBmminTx,F,2

2MF

2

dBmmaxTx,F,2

dBmMUB

maxTx,F,2

MF

dBmmaxTx,F,dBmMUB

maxTx,F,

9

18exp

PrPP

PrPPrP

Simulations and Results

- 20 -

Simulation Setup• FAPs and MUEs are randomly dropped within the macrocell

coverage, following two independent SPPPs.

Parameters Values Parameters Values

10 dB, 15 dB PM,Tx 43 dBm

M, FM 4 PF,Tx 23 dBm

F 3 GMBS 15 dBi

FF, MF 3.5 GFAP 2 dBi

M 8 dB GUE 0 dBi

F 4 dB rM 1000 m

FF 12 dB rF 30 m

MF, FM 10 dB UF 2

fc 2000 MHz M 5 dB

M, F 0.1 F 10 dB

- 21 -

Outage Probability• DL OP vs. the distance from the MBS, for NF = 30 and 100, = 10 dB.

- 22 -

• Simulated DL OP vs. the distance from the MBS, when the femtocell regulation strategy is employed at each FAP.

Performance of Femto Self Reg

- 23 -

Femto Self Reg• FAP transmit power and vs. the distance from the MBS, when using

the proposed femtocell regulation strategy.

- 24 -

Conclusions• OFDMA downlink of collocated spectrum-sharing

macrocell and closed-access femtocells ▫ Closed-form analytical expressions for outage probabilities

▫ Analytical expression of minimum MBS-to-FAP distance

▫ Simulation results have verified the accuracy of analytical results.

• Interference caused by femtocells has to be limited by▫ regulating femtocell transmit power, which depends on the

distance from the MBS; or

▫ restricting the probability of each femtocell transmitting in each RB, which can be controlled in both frequency and time domains.

- 25 -

Further Information• This research has been supported by the UK EPSRC Grants

EP/H020268/1, CASE/CNA/07/106, and the RCUK UK-China Science Bridges Project (EP/G042713/1): R&D on (B)4G Wireless Mobile Communications.

• Related publications and submissions: ▫ X. Chu, Y. Wu, D. López-Pérez and H. Wang, “Decentralized femtocell transmission

regulation in spectrum-sharing macro and femto networks,” IEEE VTC 2011-Fall, San Francisco, USA, Sep 2011.

▫ X. Chu, Y. Wu and H. Wang, “Outage probability analysis for collocated spectrum-sharing macrocell and femtocells,” IEEE ICC 2011, Kyoto, Japan, Jun 2011.

▫ X. Chu, Y. Wu, L. Benmesbah and W. K. Ling, “Resource allocation in hybrid macro/femto networks,” IEEE WCNC 2010 WS, Sydney, Australia, Apr 2010.

▫ X. Chu, Y. Wu, D. López-Pérez and X. Tao, “On providing downlink services in collocated spectrum-sharing macro and femto networks,” IEEE Trans. Wireless Commun., under review.

- 26 -

Thank You !Thank You !

Xiaoli Chu

[email protected]