URSI-GA2005 2005/10/24, New Delhi, India Double Directional Cluster - - PowerPoint PPT Presentation

ursi ga2005
SMART_READER_LITE
LIVE PREVIEW

URSI-GA2005 2005/10/24, New Delhi, India Double Directional Cluster - - PowerPoint PPT Presentation

URSI-GA2005 2005/10/24, New Delhi, India Double Directional Cluster Properties Investigated from a Series of Ultra Wideband Propagation Measurements in Home Environments Katsuyuki Haneda, Jun-ichi Takada UWB Technology Institute, National


slide-1
SLIDE 1

URSI-GA2005

2005/10/24, New Delhi, India

Double Directional Cluster Properties Investigated from a Series of Ultra Wideband Propagation Measurements in Home Environments

Katsuyuki Haneda, Jun-ichi Takada

UWB Technology Institute, National Institute of Information and Communication Technology / Tokyo Institute of Technology

Takehiko Kobayashi

UWB Technology Institute, National Institute of Information and Communication Technology / Tokyo Denki University

slide-2
SLIDE 2

Table of contents

  • Aim of this study
  • Measurement environments
  • Analyses of measured data
  • Channel parameters
  • Physical interpretation of the parameters
  • Summary
  • Acknowledgement
slide-3
SLIDE 3

Aim of this study

  • The double-directional UWB

measurement enables us to develop propagation models which are:

– Applicable both for MIMO and UWB systems – Antenna-independent – Quite high resolution using UWB signal – Available over wide frequency range of UWB

slide-4
SLIDE 4

UWB sounding system

Synthetic array

  • Vector Network Analyzer & spatial scanner

GPIB

  • Data acquisition

VNA PC

  • 3-D (x-y-z)

scanner and UWB antenna

  • frequency

sweeping from

3.1 to 10.6 GHz

UWB antenna

  • Omni-directional

monopole antenna

  • flat group delay

characteristics

  • Measurement control

via GPIB

Tx side Rx side

Preamp (30dB)

slide-5
SLIDE 5

Measurement environments

  • In a Japanese wooden house, total of 7

channels were measured:

1. LOS room 2. OLOS Room-corridor 3. NLOS Room-corridor 4. NLOS Inter-room 5. NLOS Inter-floor 6. OLOS Indoor-outdoor 7. NLOS Indoor-outdoor

slide-6
SLIDE 6

Measurement environments

  • First floor
slide-7
SLIDE 7

Measurement environments

  • Second floor
slide-8
SLIDE 8

Specifications of the experiment

  • Frequency range: 3.1 to 10.6 GHz
  • Array shape

– Virtual array, 4 x 4 squared shape URA in horizontal plane (Rx), – 4 elements ULA (Tx) – 48mm element spacing

  • Antennas at Tx and Rx

– UWB monopole antennas – Fluctuation of group delay characteristics is less than 0.1 ns in the considered bandwidth

  • Receiving SNR: at least 20 dB
  • Calibration: function of VNA, back-to-back
slide-9
SLIDE 9

Analyses on measured data (1/3)

1. Path extraction (parameters estimation)

– SAGE algorithm for UWB signal – Antenna deconvolution – Residual components: diffuse components

2. Clusterization of the paths

– In double-directional manner, i.e., in DOD-DOA- TOA domain – With a heuristic approach – Cluster: group of paths with the similar parameters

3. Analyses of clusters

– We introduced channel parameters to grasp channel behavior intuitively

slide-10
SLIDE 10

Example of power spectrum (1/2)

  • Original spectrum and estimated components by

the SAGE

Red: Spectrum of measured data Green: Detected paths by the SAGE

slide-11
SLIDE 11

Example of power spectrum (1/2)

  • Residual spectrum after the extraction of 100

waves

Red: Residual spectrum

slide-12
SLIDE 12

Diffuse components

  • 95 dB
  • 100 dB
  • 105 dB
  • 110 dB
slide-13
SLIDE 13

Channel parameters (1/5)

1. Number of clusters 2. Number of paths in each cluster 3. Standard deviation of the path distribution within clusters w.r.t. azimuth and delay 4. Cluster power ratio

→ Power ratio between the strongest cluster and

  • ther clusters (in case of LOS, strongest cluster is

replaced by LOS path)

5. Percentage of the sum of extracted power by paths 6. Dynamic range of paths over diffuse (residual) components

cluster

N

path

N

φ

σ

τ

σ

cluster

K

PCT

dyn

P

slide-14
SLIDE 14

Channel parameters (2/5)

1. The num. of clusters / the num. of paths in each cluster

LOS room OLOS room-corridor NLOS room-corridor NLOS inter-room NLOS inter-floor OLOS indoor-outdoor

cluster

N

path

N

7 3 ~ 13 8 3 ~ 19 8 2 ~ 20 8 2 ~ 23 9 3 ~ 14 6 2 ~ 14 3 ~ 18 6

NLOS indoor-outdoor

slide-15
SLIDE 15

Physical interpretation (1/3)

  • Reinforcing inside the ceiling causes diffraction
slide-16
SLIDE 16

Physical interpretation (3/3)

  • Result of cluster extraction
slide-17
SLIDE 17

Channel parameters (3/5)

2. Standard deviation of paths in azimuth / delay domains

φ

σ

LOS room OLOS room-corridor NLOS room-corridor NLOS inter-room NLOS inter-floor OLOS indoor-outdoor

0.7 ~ 6.3

τ

σ

[deg] [ns]

0.06 ~ 1.41 3.0 ~ 10.8 0.61 ~ 1.88 0.9 ~ 9.3 0.50 ~ 1.79 0.3 ~ 11.9 0.15 ~ 1.28 1.2 ~ 13.1 0.28 ~ 1.90 0.7 ~ 10.5 0.11 ~ 1.83 0.11 ~ 1.85 0.9 ~ 6.4

NLOS indoor-outdoor

slide-18
SLIDE 18

Channel parameters (4/5)

3. Cluster power ratio and extracted power by paths

cluster

K

PCT

[dB] [%]

77.0

2.1 −

LOS room

51.0

4.8 −

OLOS room-corridor

52.9

4.3 −

NLOS room-corridor

71.8

1.2 −

NLOS inter-room

58.8

5.4 −

NLOS inter-floor

72.3 9.6

OLOS indoor-outdoor

7.4 77.5

NLOS indoor-outdoor

slide-19
SLIDE 19

Physical interpretation (2/3)

Tx Rx Metal shutter

  • Indoor-outdoor meas.
slide-20
SLIDE 20

Channel parameters (5/5)

4. Dynamic range of paths over diffuse components

dyn

P PCT

[dB] [%]

77.0 29.2

LOS room

51.0 24.0

OLOS room-corridor

52.9 25.5

NLOS room-corridor

71.8 30.5

NLOS inter-room

58.8 21.6

NLOS inter-floor

72.3 31.6

OLOS indoor-outdoor

77.5 34.0

NLOS indoor-outdoor

slide-21
SLIDE 21

Physical interpretation (3/3)

  • Indoor-outdoor scenarios

– The strongest cluster carries most of the propagating power – Large and

  • LOS scenario

– Contribution of power from scattered paths is large due to effective illumination of scattering objects by Tx and Rx – Large and (but small )

  • NLOS (OLOS) scenarios

– Channels are diffuse in which huge number of weak paths aggregate – Small and

cluster

K PCT ,

dyn cluster

P K PCT

dyn

P ,

dyn cluster

P K PCT

slide-22
SLIDE 22

Summary

  • Deterministic measurement and characterization
  • f propagation channels in a wooden house for

UWB / MIMO indoor applications

– Cluster analyses of paths – Introduction of channel parameters and its derivation – Physical interpretation of the parameters

  • Future works

– Constructing channel models – Evaluation of antenna effects for MIMO and UWB systems – Measurements in other environments

slide-23
SLIDE 23

Acknowledgement

  • The authors would like to thank the

members of NICT UWB Consortium and UWB Technology Institute in NICT:

– Prof. Dr. Kiyomichi Araki – Mr. Iwao Nishiyama – Dr. Honggang Zhang – Mr. Fumio Ohkubo – Dr. Makoto Yoshikawa – Dr. Akira Akeyama – Dr. Osamu Sasaki – Dr. Yuko Rikuta – Mr. Takahiro Miyamoto

slide-24
SLIDE 24

Analyses on measured data (2/3)

  • Cluster extraction?
slide-25
SLIDE 25

Double-directional UWB meas.

  • Double-directional measurement enables us to

separate antenna effects from channel model (M. Steinbauer et., al, 2001)

DOD DOA

Tx Rx Propagation Antenna + Propagation = Channel