Transmitted Power Formulation for the Implementation of ...

23
© 2005, it - instituto de telecomunicações. Todos os direitos reservados. Jessica Acevedo Flores Daniel Robalo Fernando J. Velez Transmitted Power Formulation for the Implementation of Spectrum Aggregation in LTE-Advanced over 800 MHz and 2 GHz Frequency Bands 7º Congresso Comité Portugués da URSI 22 de novembro de 2013 / Lisboa, Portugal

Transcript of Transmitted Power Formulation for the Implementation of ...

© 2005, it - instituto de telecomunicações. Todos os direitos reservados.

Jessica Acevedo Flores

Daniel Robalo

Fernando J. Velez

Transmitted Power Formulation

for the Implementation of

Spectrum Aggregation in

LTE-Advanced over 800 MHz and

2 GHz Frequency Bands

7º Congresso Comité Portugués da URSI

22 de novembro de 2013 / Lisboa, Portugal

2 22 de novembro de 2013 / Lisboa, Portugal

Outline

Introduction;

Objectives;

System model;

Topology;

Average SINR;

Normalized Transmitter Power;

Simulation results;

Conclusions and Future Work.

7º Congresso Comité Portugués da URSI

3

Introduction

Dynamic Spectrum Access (DSA) comprises Spectrum Aggregation

(SA) to alleviate the spectrum scarcity problem;

Radio Resource Management (RRM) guarantees Quality of Service

(QoS), since scheduling determines among packets that are ready for

transmission;

Common RRM (CRRM) ensures an efficient and coordinated use of

the available radio resources in heterogeneous networks scenarios.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

DL

DL

Operator 1

DL band

Operator 2

DL band

Operator 3

DL band

Shared by all 3

operators

DL

DL

4

Objectives

To propose a formulation to calculate the transmitter power required

to cover cells of different sizes whilst maintaining the average

Signal-to-Interference-plus-Noise Ratio (SINR) constant;

To consider a MultiBand-CRRM entity, for the 2 GHz and 800 MHz

frequency bands for a single operator scenario under a constant

average SINR using the LTE Simulator in [1];

To implement the integrated-CRRM [2], where CRRM and SA

functionalities are handled simultaneously, in order to facilitate the

best user allocation and maximize the total network throughput in

LTE systems.

[1] G. Piro, L. A. Grieco, G. Boggia, F. Capozzi and P. Camarda, “Simulating LTE Cellular Systems:

an Open Source Framework,” IEEE Transaction on Vehicular Technologies, Vol. 60, No. 2, Feb 2011,

pp. 498-513.

[2] O. Cabral, F. Meucci, A. Mihovska, F.J. Velez and N.R. Prasad. Integrated Common Radio

Resource Management with Spectrum Aggregation over Non-Contiguous Frequency Bands. Wireless

Personal Communications, Vol. 59, No. 3, Aug. 2011, pp. 499-523.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

5

System Model

Carrier frequency 800 MHz 2 GHz

Bandwidth, BW 5 MHz 5 MHz

Path loss model

Path Loss Model

The radio channel follows the ITU radio propagation COST-231 Hata

model for urban and suburban scenarios ;

R is the base station (BS)/user equipment (UE) maximum separation,

f is the carrier frequency, and Dhb is the BS antenna height (from the

average rooftop level).

Considering two carrier frequencies, 800 MHz and 2 GHz, Dhb = 15 m

and a UE antenna of 1.5 m, we obtain the following path loss model:

)(log)1041(40 ][10][3

][ KmKmhbdB RDL 80)(log21)(log18 ][10][10 MHzKmhb fD

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

6

Parameters for LTE DL budget for a data rate of 1 Mbps and a commercial

omnidirectional antenna.

System Model (II)

Transmitter – NodeB

a) Max. TX power (dBm) 50

b) TX antenna gain (dBi) 3 - 3.5 For 800 MHz and 2 GHz respectively

c) Body loss (dB) 2

d) EIRP (dBm) 51- 51.5 = a + b – c

Receiver UE

e) Node B noise figure (dB) 8 LTE specifications

f) Receiver noise floor (dBm) -99 =-174+10log (BW) + e

g) SINR (dB) -10 From simulations

h) Receiver sensitivity (dBm) -109 = f + g

i) Interference margin (dB) 3

j) Cable loss (dB) 1

k) RX antenna gain (dBi) 0

l) Fast fade margin (dB) 0

m) Maximum path loss (dBm) 156-156.5 = d – h – i – j + k – l

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

7

Topology

Hexagonal cell

deployment, with

frequency reuse

factor K = 3;

In co-channel

interference.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

8

Topology (II)

where:

PTx is BS transmitter power;

Pow is the power received from the own

cell;

α is the orthogonality factor;

Pnh is the total amount of interfering power

coming from the neighbour cells;

is the thermal noise power, given by:

being NF = 8 dB for LTE and BW = 5 MHz.

SINR

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

9

with,

=

,,

=

,

=

where

are the antenna gains

PTx = 1 dB;

= 3 dBi;

= 3.5 dBi;

= 0 dBi.

Topology (III)

Received power and Interference

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

10

Topology (IV)

PL, at 800 MHz:

PL, at 2 GHz:

Path loss

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

11

Topology (V)

SINR as a function of the MS-BS distance for the two cell radii

(R ϵ {300,1500}m) and PTx = 1 dBW

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

12

Average SINR

where

= 6·

Acellnh is the total affected

cell area;

Acellnh=

is the total

integration area to calculate

(neighbour) cell

interference.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

13

Average SINR: Power and interference (II)

where

=

is the total integration area to compute the

signal/power and own cell interference;

is the Fraunhofer distance and is different for each frequency band,

depending on antenna specifications.

cellr

ryx

A

xy

RxTxTxTxMHzow GGPyxPP10

log2.376.119

800

2210

3

1

,

10),,(

nhcellr

ryx

A

xy

RxTxTxTxMHznh GGPyxPP10

log2.376.119

800

2210

3

1

,

10),,(

cellr

ryx

A

xy

RxTxTxTxGHzow GGPyxPP10

log37.6+128.1

2

2210

3

1

,

10),,(

nhcellr

ryx

A

xy

RxTxTxTxGHznh GGPyxPP10

log37.6+128.1

2

2210

3

1

,

10),,(

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

14

Average SINR (III)

Average power and interference within a cell as a function of the

inter-cell distance with PTx = 1 dBW and α = 1.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

15

Average SINR (IV)

Average SINR as a function of the cell radius with different

orthogonality values (α ϵ {0.9, 1}) and constant PTx = 1 dBW for

the 800 MHz and 2 GHz frequency bands.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

16

Average SINR (V)

Average SINR as a function of the cell radius (m) with three

values of PTx(-10, 1 and 10 dBW).

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

17

Normalized Transmitter Power

Normalized PTx required to achieve a selected high average SINR (dB),

near the maximum, as a function of the cell radius at 800 MHz and 2 GHz.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

18

Video Traffic Throughput

Video traffic simulation setup:

Traced-based video sessions have been addressed for

simulations, these applications send packets based on realistic

video trace files;

We have considered a video bit rate of 128 kbps;

Modified Largest Weighted Delay First (MLWDF) scheduler.

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

19

Results for video Traffic Throughput

Average Throughput vs. number of users for different cell radii

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

20

Results for video Traffic Throughput (II)

Average Throughput as a function of cell radius, R, for 40 users (2 GHz

and 800 MHz bands) and 2X40 users with CRRM

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

21

We propose an innovative formulation to calculate the transmitter

power required to cover cells of different sizes whilst maintaining

the average SINR constant, in the context of SA and MB-scheduler

to switch the users between two frequency bands -800 MHz and 2

GHz- for LTE systems;

We considered a topology with K = 3 and the COST-231 Hata

model for the path loss, and obtained the average SINR in the cell;

To assess the possible impact of non-ideal orthogonalities,

(α = 0.9), we have also evaluated the impact of the DL

orthogonality factor in the average SINR. In LTE, in the ideal

situation (no multipath), only inter-cell interference is considered.

Conclusions and Future Work

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

22

Future work:

To complete the implementation of the integrated-CRRM;

To conceive a cost/revenue performance analysis of our LTE

scenario in the context of SA with MB-scheduler;

To develop a similar formulation for tri-sector cells and K = 1,

in order to evaluate our system in terms of output parameters,

such as throughput and frame error rate.

Conclusions and Future Work (II)

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI

23

Thank You!

Q & S

22 de novembro de 2013 / Lisboa, Portugal

7º Congresso Comité Portugués da URSI