Wireless Channel Models2 - poly.edufaculty.poly.edu/~tsr/Presentations/Wireless_Channel_Models.pdf · Title: F:\PowerPoint Presentations\CAMAD June 2000\Wireless_Channel_Models2.PDF

$Page 1: Wireless Channel Models2 - poly.edufaculty.poly.edu/~tsr/Presentations/Wireless_Channel_Models.pdf · Title: F:\PowerPoint Presentations\CAMAD June 2000\Wireless_Channel_Models2.PDF$
Wireless Channel Models and Wireless Channel Models and Simulators from RealSimulators from Real--World DataWorld Data

Prof. Theodore S. RappaportJames S. Tucker Professor of Engineering

Mobile and Portable Radio Research GroupBradley Dept. of Electrical and Computer Engineering

Virginia Polytechnic Institute and State University http://www.mprg.ee.vt.edu/

8th IEEE CAMAD International Workshop

June 22, 2000

© Copyright 2000 All rights reserved.© Copyright 2000 All rights reserved.

© Copyright 2000 Wireless Valley Communications, Inc. and T. S. Rappaport

MPRG

Introduction

Simulation Software: Simulation of Mobile Radio Channel Impulse Response Models (SMRCIM) and Simulation of Indoor Radio Channel Impulse Response Models (SIRCIM)

Implementation of geometric models for simulating Angle-of-Arrival (AOA) of multipath components

RF channel models utilized by SIRCIM and SMRCIM

Variations of the Geometrically Based Single-Bounce Elliptical Model (GBSBEM)


MPRG

The Basic Mobile Multipath Environment

mobile 1

base station

mobile 2

A t e t j t

1 2 1 2 1, 2

, ( )

, ( ) ( ) ϕ

δ τ −

θ 0 2 ,

θ 1 1 ,

θ 0 1 , θ 1 2 ,

A t e t j t 0 1 0 1

0 ,

( ) , ( ) ( ) ,1 ϕ δ τ −

beam of the base station antenna steered toward mobile

1

θ 2 1 ,

A t e t j t

0 2 0 2 0 2

, ( )

, ( ) ( ) , ϕ δ τ −

A t e t j t 1 1 1 1

1,1 ,

( ) , ( ) ( )

ϕ δ τ −

A t e t j t 2 1 2 1 2 ,

( ) , ( ) ( ) ,1 ϕ δ τ −

Al,k = Amplitude t l,k = time delay

f l,k = Carrier phase shift ?l,k = angle-of-arrival (AOA)

of lth signal component of kth mobile

Source: J.C. Liberti and T.S. Rappaport, Smart Antennas for CDMA Wireless Systems: Applications to IS-95 and Wideband CDMA, Prentice Hall, Prentice Hall, NJ, 1998.


MPRG

Channel Impulse Response

Complex baseband discrete representation of the channel impulse response:

∑ −=k

kj

kkkb tpeth k )()(),,( τφαφτ θ

Where:

= power of kth multipath component)(2kφα

= relative time delay of kth multipath componentkτ= phase of kth multipath componentkθ= narrow pulse approximating delta function)(tp

kφ = angle-of-arrival of kth multipath component


MPRG

Why do we want the Impulse Response?

∑x(t)

n(t)

),,( kkb th φτ r(t)

Modem developmentSpatial algorithm developmentTemporal Equalization

Position location and trackingSmart antennas3G, 4G


MPRG

RF Channel Modeling Background

Mid 1970s -- G.L. Turin at University of California, Berkeley models San Francisco urban/suburban mobile radio channel

Suzuki and Hashemi -- fit measured data to statistical models with intention of developing simulation software Simulation of Urban Propagation (SURP)

Devasirvatham and Cox measures in-building environments at 850 MHz with wideband channel sounding equipment

Transmitted 100ns pulses and received them on three-trace oscilloscopes

Saleh and Valenzuela propose a statistical model for the indoor channel based on measured wideband multipath profiles


MPRG

Continuation of RF Channel Modeling Background

Mid 1980s -- T. S. Rappaport at Purdue University collects data from factories for future of wireless applications

1989 -- S. Seidel, using Rappaport’s research and other published data, creates the initial concept of SIRCIM

Early 1990s -- W. Huang & M. Feurestein acquire data from outdoor environments and begin early development of SMRCIM

1998-99 -- J.E. Nuckols implements Liberti’s GBSBEM models in SIRCIM and SMRCIM for simulation of AOA

Mid 1990s -- MPRG students measure noise in various indoor environments and add noise models to SIRCIM

1990s -- International researchers publish data used to improve SIRCIM and SMRCIM (e.g. Hashemi, Devasirvatham, Pahlavan)


MPRG

Model Environments

SIRCIM -- Indoor environmentsOPEN PLAN -- Factories, department stores, warehouses

HARD PARTITIONED -- Apartments, school buildings, homes, hospitals

SMRCIM -- Outdoor environmentsMICROCELLULAR -- Campuses, microcells URBAN -- Heavily built up city areasSUBURBAN -- Outskirts of a city, open farmland next to small neighborhoods

SOFT PARTITIONED -- Cubicles, voting booths, restaurants, grocery stores

Source: S. Y. Seidel, “UHF Indoor Radio Channel Models for Manufacturing Environments,” Masters Thesis in Electrical Engineering, Virginia Polytechnic Institute and State University, Blacksburg VA, Aug. 1989.


MPRG

Physical Channel Model

Channel simulated as receiver moves over local area “track”

User selectable track length of 4.5, 10, 20, 40, or 80? Mobile direction, a, ranges from 0° and 360°

t i

Tx Rx

a delay Multipath

locationReceiver

locationr Transmitte direction bilevehicle/mo

i =

=

==

τ

α

x

x

R

T

Motion track

Large-scale path loss model for each multipath component

64 time delay bins used to represent hb(t) at any instant of position in local areaComplete doppler, AOA, and second-order effects are modeled

Small-scale model:

Verified by numerous research programs, papers, observed resultsdn model with wide range of values for n, s


MPRG

Simulated Multipath Delay Profile

Multipath delay profile generated in SIRCIM along a 20? trackOPEN PLAN indoor environment with TR separation of 11.2 m

Excess delay

Relative squared

amplitude of

multipath

Distance along track


MPRG

Geometrically Based Single-Bounce Elliptical Model

1995 - (GBSBEM) developed by MPRG student, Joe Liberti

Dbm

am

Scatterer Region

x

y

Base stationMobile

Assumption of low antenna heights -- scattering near base station as likely as near mobile

222

21

2

Dcbc

a mmm

m −== ττ

separationR -T delay multipath

light of speed

axisminor

axismajor

m

==

=

=

=

D

c

b

a

m

m

τ

2121



MPRG

Variations of GBSBEM

Aisle Elliptical Model -- constricted scattering environments

Random Elliptical Model -- unconstrained scattering environments

Outdoor

Indoorcity blocks with rows of buildings

hallways

Outdoor

Indoorsuburban and urban areas

factories, warehouses


MPRG

Aisle Elliptical Model

Ellipse represents set of all scatterers causing multipath delay of t i

? W

s

Tx RxAOA

t 1

t 2delay Multipath

locationReceiver

locationr Transmitte locationscatterer

width aisle

i ==

==

=∆

τx

x

R

TS

w

Aisle represents hallways or rows of buildingsModel results in end fire clustering at 0° and 180°

NOTE: Not drawn to scale

Source: J. Eric Nuckols, “Implementation of Geometrically Based Single-Bounce Models for Simulation of Angle-of-Arrival of Multipath Delay Components in the Wireless Channel Simulation Tools, SMRCIM and SIRCIM,” Masters Thesis in Electrical Engineering, Virginia Polytechnic Institute and State University, Blacksburg VA, Dec. 1999.


MPRG

Random Elliptical Model

delay Multipath

locationReceiver locationr Transmitte

locationscatterer

i =

===

τx

x

RTS

s

TxRx

AOA

t 1

t 2

? x

y

?

f(?)

2p

π21

Ellipse represents set of all scatterers causing multipath delay of t i

Multipath likely to arrive from any direction -- useful in open areas? uniform and random over 2p -- determines location of scatterer

0



MPRG

Analysis of Aisle Elliptical Model - Geometry

Associate 64 randomly chosen cases with each delay t i

Four “CASES” corresponding to geometry for each of four scatterer locations

Simulate AOA based on user entered parameters in conjunction with geometry

CASE I of IV

ωθ ±= 4AOA



MPRG

Analysis of Random Elliptical Model

Generate 64 randomly chosen angles ? associated with each delay t iBased on the angles, ?, determine the scatterer location for each delay Using the geometry and scatterer locations simulate the AOA

±= −

TRSR

TRSRAOA

''

','cos 1



MPRG

Typical Aisle Elliptical Results

End fire clustering

TR-separation distance vs. multipath path distance (size of ellipse)

As delay increases, aisle quickly constrains AOA to small angles

Hallways will tend to guide the waves

Source: Output generated by commercial license of SIRCIM Plus 4.0. This software is copyrighted and licensed by Wireless Valley Communications, Inc.


MPRG

Typical Random Elliptical Results

Not constricted to end fire clustering

Multipath likely to arrive from any direction

Open areas -- well suited to this model due to random scatterers

Source: Output generated by commercial license of SIRCIM Plus 4.0. This software is copyrighted and licensed by Wireless Valley Communications, Inc.


MPRG

Phase Generation for Narrowband Simulations

Uniform random phases generated for initial receiver locationsAOA used to recover scatterer locations along ellipseSmall changes in radio path length from scatterer to receiver location on track account for change in phase

With motion along track, incremental

changes in path length reflect small

changes in narrowband phase



MPRG

Who uses these tools?

SIRCIM and SMRCIM are being used world-wide by many companies and research institutions such as:

Oklahoma Christian University

Tantivy Communications

Australian Department of Defense

Chinese University of Hong Kong

Institute for Advanced Engineering, Korea

ITT Aerospace Communications

Korea Advanced Institute of Science and Technology

Georgia Tech

Purdue University

University of Kansas

Motorola

SiemensU.S. Department of Defense GAT Systems

Texas Instruments

RF Micro Devices

Broadcom

Nokia


MPRG

Conclusion

Development of the Random Elliptical and Aisle Elliptical Models from GBSBEM ModelImplementation of these models in SIRCIM and SMRCIM

As a Bradley Industrial Fellow have commercialized SIRCIM and SMRCIM into world class products

SIRCIM and SMRCIM are licensed by Wireless Valley Communications in Blacksburg

Future Work Current work in MPRG to build measurement apparatuses Gather site specific AOA measurements and generate statistics to compare with geometrical models

Many companies and universities now have easy to use software tools that provide them with life-like channel information in addition to AOA of multipath delay signals

Wireless Channel Models2 - poly.edufaculty.poly.edu/~tsr/Presentations/Wireless_Channel_Models.pdf · Title: F:\PowerPoint Presentations\CAMAD June 2000\Wireless_Channel_Models2.PDF

Documents