Vous êtes sur la page 1sur 40

Administrivia

Reading list. Mailing list. Topics to be covered.


Need volunteers. Presenters choose readings (subject to our feedback). Project proposal.
1

Goal

Radio propagation is a whole course! Our goal is to understand the basics.


Understand propagation models. To better interact with simulators. For completeness. Radio propagation for CS audience.

Outline

Radio Waves Propagation Mechanisms and Models


Free space Refraction, diffraction, and scattering

Path Loss Models Small-Scale Fading Modulation

Radio Waves

Electromagnetic energy. Propagate from transmitting antenna. Microwave frequencies: 109-1012 Hz.

Spectrum
Wave length (m)

-1

-4

-7

-10

AM, AB,FM,TV micro

IR UV

12

15

18

Frequency (Hz)
5

Propagation Mechanisms

Free space (no obstacles): straight line (line-of-sight). With obstacles (multipath):
Reflection (or refraction) Diffraction Scattering

Propagation Models

Predict signal strength at receiver.


Attenuation or gain of original signal Factors:
Absorption, reflection, diffraction, scattering. Interference. Mobility.
7

Propagation Models

Different models for:


different scales:
large scale: in meters. Small scale: in wavelength.

different environments:
outdoor indoor

Free Space Propagation Model

Transmitter and receiver have clear line-of-sight. Typically, large-scale, outdoor. Examples:
satellite communication microwave links

Free Space Propagation Model

Pt Pr (d ) K 2 d

Pt is the transmitters power. K (constant) depends on antennae

gain, system loss factor, and carrier wavelength.


10

Refraction

If incident plane is perfect dielectric: part reflected, part transmitted (no loss). If perfect conductor, all is reflected.

11

Diffraction

Waves hit edge of obstacle. Secondary waves propagate into shadowed region. Excessive path length causes phase shift: Fresnel zones.

Obstruction

12

Diffraction: Fresnel Zones

Represent regions where secondary wavelengths have path length > LOS path by n l/2. Alternate zones differ by 1800: interference (+ or -).

13

Diffraction Models

Estimating diffraction loss caused by hills, buildings, etc. is essential. Knife-edge diffraction model: good first approximation. Multiple knife-edge diffraction: series of obstacles replaced by single equivalent obstacle.
T R

OBSTRUCTION 1

OBSTRUCTION 2

14

Scattering

Rough surfaces diffuse radio wave in all directions. How flat is a surface?
If really flat, modeled as reflective surfaces. Rayleigh criterion: defines critical bump height as function of l and .

15

Scattering Models

Nearby metal objects (street signs, etc.)


usually modeled statistically.

Large, distant objects


analytical model: radar cross section (RCS).

16

Path Loss Models

Path loss: measures signal attenuation along transmitterreceiver path. Most models use combination of analytical and empirical.
Empirical approach: fitting curves/expressions to re-create measured data. Validity: different environments?

17

Path Loss Models

Free Space Models


Log-distance Path Loss Model Log-normal Shadowing

Outdoor Models Indoor Models

18

Free Space Path Loss Models

Free space model only valid in far-field. Typically define reference distance (d0) determined from close-by measurements.

19

Log-Distance Path Loss Model

d PL (dB) PL (d 0 ) 10 n log d0

Average received signal power decreases logarithmically with distance. Path loss exponent n indicates rate loss increases with distance.
20

Log-distance Path Loss Model

Value of n depends on propagation environment.


Environment Free space Urban area cellular Shadowed urban cell In building LOS Obstructed in building Obstructed in factories n 2 2.7 to 3.5 3 to 5 1.6 to 1.8 4 to 6 2 to 3
21

Log-Distance Path Loss Model

Reference distance d0 also depends on propagation environment.


Large-coverage cell systems: d=1Km. Microcellular systems: d=1-100m

22

Log-Normal Shadowing Loss Model

Log-distance model doesnt different environmental obstructions. Same T-R distance may lead to different signal strengths in different environments.

23

Log-Normal Shadowing

Simple probabilistic model can account for random shadowing.


Just add Gaussian random variable to log-distance PL.

24

Log-Normal Shadowing

Simple probabilistic model can account for random shadowing.


Just add Gaussian random variable to log-distance PL. Gaussian distribution has distancedependent mean.

PL(d )[dB] PL(d0 ) X


25

Outdoor Models

Account for terrain profile. Models to predict loss over irregular terrain. Most models based on interpretation of measured data in service area.

26

2-ray Ground Reflection Model

Considers both direct and ground reflected paths. Applicable when d>>ht and d>>hr Pr ~ 1/d4.

ht

hr

d
27

Outdoor Models

Longley-Rice Model
Use 2-ray ground reflection Diffraction loss prediction using knife-edge models.

Durkins Model
Implemented in a simulator. Accesses topographic database. Determines whether free space, diffraction (single-, two-, etc.)

28

Urban Models

Okumura Model (1968)


Empirical study of path loss in Japanese cities. No analytical results. Useful for planning urban cellular systems. Includes correction factors for rural and suburban areas.

29

Urban Models

Hata Model
Empirical formulation of the path loss data provided by Okumura. Valid from 150MHz to 1500MHz.

Walfisch and Bertoni


Considers effects of rooftops and building heights through diffraction.
30

Indoor Models

Relatively new area. Smaller distances. Greater environment variability. Partition losses: same floor.
Walls, furniture, etc. Highly material dependent. Hard versus soft partitions.

Tabulated losses: material, freq.


31

Indoor Models

Partition losses: between floors


Type of construction and materials. Attenuation factor diminishes with number of floors. Typical values:
15dB for first floor. 6-10dB per floor for floors 2-5. 1-2dB per floor beyond 5 floors.
32

Small-Scale Fading

So far, large-scale propagation models. Mobility: rapid fluctuation of radio signal amplitude over small period of time/distance. Fading: caused by multiple versions of the same signal (multipath) which are received at (slightly) different times.

33

Small-Scale Fading: Factors

Multipath propagation:
Reflecting objects and scatterers. Multiple versions of transmitted signal arrive shifted in time and orientation. Fluctuations in signal strength.

34

Fading: Factors

Movement speed:
Relative motion between transmitter and receiver results in random frequency modulation (Doppler effect) on each multipath component.

35

Fading: Factors

Speed of surrounding objects:


Also induce Doppler effect on multipath omponents. If speed surr-objects > speed T-R , this effect dominates fading.

36

Fading: Factors

Transmission bandwidth:
If greater than channel bandwidth, received signal will be distorted. Channel bandwidth: can be quantified by the coherence bandwidth, related to the multipath structure of channel.

37

Doppler Effect

Caused by difference in path lengths traveled by signal from S to mobile (X -> Y). Positive Doppler (increased freq.) if mobile going towards signal.

l v

Y d

38

Small-Scale Multipath Measurements

Ways of measuring the multipath structure of a channel. Measurement techniques:


Direct pulse. Spread spectrum sliding correlator. Swept frequency.
39

Multipath Fading Channel Modeling

Statistical models:
Clarkes model
Uses scattering

Two-ray Rayleigh fading model Saleh and Valenzuela indoor model

40

Vous aimerez peut-être aussi