Cellular Mobile Communications-
V Mobile Radio Propagation Dr. Nasir D. Gohar
Wireless Communications, Principles and Applications Ch.5 [Small Scale Path Loss] by RAPPAPORT Propagation Characteristics Observed in Macro / Micro Cells by H. L. Bertoni
Mobile Radio Propagation Modeling a Radio Channel: Most difficult part in Radio System Design Highly unpredictable as compared to fixed wireline media Transmission path and its parameters keeps changing instantaneously
Changes in path profile Obstructions Environmental changes Typically done statistically based on field measurements[System Specific]
Physical survey *1 Computer simulation using certain models and terrain data *2
*1
Gives some real picture of the environment and data can be incorporated to new system design and predict its performance
*2
Empirical Models and Terrain data may not be quite updated
T.S. Rappaport Ch 4-5
NDG Notes
2
Mobile Radio Propagation The Multi-path Environment The received signal is made up of a sum of attenuated, phase shifted and time delayed versions of the transmitted signal. Propagation modes include diffraction, transmission and reflection.
T.S. Rappaport Ch 4-5
NDG Notes
3
Mobile Radio Propagation The Mult-ipath Environment - contd. Assuming receiver is stationary and there is no direct path, the received signal can be expressed as a sum of delayed components or in terms of phasor notation: N
Pulse train
er (t ) = ∑ai p (t − ti ), i =1
N
A single pulse
er (t ) = ∑ ai cos (2πf c + φ i ) i =1
Where: ai is the amplitude of the scattered signal, p(t) is the transmitted signal (pulse) shape, ti is the time taken by the pulse to reach the receiver, N is the number of different paths taken by the signal to reach receiver, and fc is the carrier frequency T.S. Rappaport Ch 4-5
NDG Notes
4
Mobile Radio Propagation The Multi-path Environment - contd.
a Diffracted wave a b
Antenna
y=a+b
a & b are in phase
b Reflected wave a b
No direct path
Antenna
y=0
a & b are out of phase by π Complete fading
T.S. Rappaport Ch 4-5
NDG Notes
5
Mobile Radio Propagation Fading Is due to multi-path propagation.
With respect to a stationary base station, multipath propagation creates a stochastic standing wave pattern, through which the mobile station moves.
Caused by shadowing: when the propagation environment is changing significantly, but this fading is typically much slower than the multipath fading.
Modem design is affected mainly by the faster multipath fading, which can be normally assumed6 to T.S. Rappaport Ch 4-5 NDG Notes
Mobile Radio Propagation
Signal strength relative to 1uV (db)
Fading - Types
Slow (Long) Term Fast (Short) Term (Also known as Rayleigh Fading) 30 Fast fading 20
10 Slow fading 0 0
5
10
15
20
25 Distance (λ)
Exact representation of fading characteristics is not possible, because of infinite number of situation.
T.S. Rappaport Ch 4-5
NDG Notes
7
Mobile Radio Propagation
Fading - Slow (Long) Term Slower variation in mean signal strength as the receiver moves behind buildings and the propagation paths are obscured Variations of up to 20dB will cause handovers and change quality-of-service Caused by shadowing: Terrain configuration: Results in local mean (long term fading) attenuation and fluctuation. The built environment (rural and urban areas etc.), between base station and the mobile unit: Results in local mean attenuation T.S. Rappaport Ch 4-5
NDG Notes
8
Mobile Radio Propagation
Fading- Fast (Short) Term Describes the constant amplitude fluctuations in the received signal as the mobile moves. Caused by multipath reflection of transmitted signal by local scatters (houses, building etc.) Observed over distances = λ/2 Signal variation up to 30 dB. Is a frequency selective phenomenon. Can be described using Raleigh statistics, (no line of sight). Can be described using Rician statistics, (line of sight). Causes random fluctuations in the received power, and also Distorts the pulse carrying the information. T.S. Rappaport Ch 4-5
NDG Notes
9
Mobile Radio Propagation
Fading- Fast (Short) Term - contd. A received signal amplitude is given as the sum of delayed components. In terms of phasor notation it is given as: N
er (t ) = ∑ ai cos (2πf c + φ i ) i =1
Or N
N
i =1
i =1
er (t ) = cos( 2πf ct ) ∑ ai cos (φi ) − sin( 2πf ct ) ∑ ai sin( φi )
In-phase T.S. Rappaport Ch 4-5
Quadrature NDG Notes
10
Mobile Radio Propagation
Fading- Fast (Short) Term - contd. The phaseφi can be assumed to be uniformly distributed in the range (0, 2π), provided the locations of buildings etc. are completely random. This for large N, the amplitude of the received signal is:
er (t ) = X cos( 2πf ct ) − Y sin( 2πf ct ) where
N
N
i =1
i =1
X = ∑ ai cos (φi ), Y = ∑ ai sin(φi )
X and Y are independent, identically distributed Gaussian random variables. T.S. Rappaport Ch 4-5
NDG Notes
11
Mobile Radio Propagation Fading- Fast (Short) Term - contd. The envelope of the received signal is: 2
2 0.5
A = (X + Y )
Which will be Raleigh distributed.
Rayleigh Probability density function Exponential A or power P T.S. Rappaport Ch 4-5
NDG Notes
12
Mobile Radio Propagation
Raleigh Distribution If the impulse response h(τ , t) of the mobile radio station is time invariant (without a significant deterministic ) and has zero mean, then the envelope of the impulse response has a Rayleigh Distribution given as: r2 r p ( r ) = 2 exp − 2 σ 2σ
where σ2 is the total power in the multi-path signal T.S. Rappaport Ch 4-5
NDG Notes
13
Mobile Radio Propagation Rice Fading If however the impulse response has a non zero mean then there is a significant component of the direct path (line of sight, specular component) signal and the magnitude of the impulse response has a Ricean distribution
Ricean distribution is the combination of Rayleigh signal with the direct line of sight signal. The distribution is:
r 2 + s 2 rs r I p( r ) = 2 exp − 2 0 2 σ 2σ σ
σ2 is the power of the line of sight signal and I0 is a Bessel function of the T.S. Rappaport Ch 4-5 NDGfirst Notes kind 14
Mobile Radio Propagation Fading
Multi-path create small-scale fading effects
Rapid changes in the signal strength due to movement and/or time Random frequency modulation due to Doppler shift on different multipath propagation paths Time dispersion due to multipath propagation delay T.S. Rappaport Ch 4-5
NDG Notes
15
Mobile Radio Propagation
V V
V Stationary
V
Field strength
Fast Fading – Cases 1: Stationary Mobile
t
V
Mobile is stationery surrounded by moving cars. The number of fading depends on the : Traffic flow Distance between the mobile and moving cars T.S. Rappaport Ch 4-5
NDG Notes
16
Mobile Radio Propagation
No scattered signals θ
V
Field strength
Fast Fading – Cases 2: Non-stationary Mobile Signal level
t
The received signal at the mobile is: j ( 2 πf t −βx cos θ) r x = Vt
s = Ae
Amplitude
T.S. Rappaport Ch 4-5
Wave number =2π/ λ Transmitting frequency NDG Notes
17
Mobile Radio Propagation SMALL SCALE MULTI-PATH PROPAGATION Multi-path reception of several versions of the same transmitted radio signal arriving at certain location [Rx] causes fading effects.
Rapid Fluctuations in Signal Strength [over very short distance or time period] Random Frequency Modulation due to Varying Doppler Shifts on Different Multi-Path Signals Time Dispersion / Echoes due to Multi-Path Propagation Delay
Multi-path propagation occurs even when there is LOS path In multi-path channel environment, the received signal at any point would be a vector sum of all the versions/components of the signal, reaching at the point, which may widely differ in magnitude and phase.
T.S. Rappaport Ch 4-5
NDG Notes
18
Mobile Radio Propagation Multipath Channel: Impulse Response Model A mobile radio channel can be modeled as a linear filter with time varying Impulse Response [IR]. IR contains all the necessary information required to simulate or analyze any type of radio transmission through the channel. IR is a useful channel characterization which can be used to predict and compare the performance of many mobile communication systems/transmission bandwidths for any given particular channel condition. x(t)
h(t)
y(t)
Fig-1: Impulse Response Model of a LTI system. The output signal y(t) can be obtained from the convolution of the input signal x(t) with the IR = h(t) of the system. If x(t) is band limited signal, y(t) is also band limited regardless h(t) is band limited or not. So, we can assume LTI system to be band limited if the input is band limited. T.S. Rappaport Ch 4-5
NDG Notes
19
Mobile Radio Propagation Multipath Channel: Impulse Response Model Suppose an MS is moving along the ground at a constant velocity v. At a certain fixed location distant d, the system can be assumed as LTI system. Due to multi-path reception, various multi-path components will have different delays associated with them because of difference in distance covered. So, the IR of this LTI system is a function of time as well as of distance, i.e. IR = h(d,t)
v spatial position
d
Fig-2: The Mobile Radio Channel as a Function of Time and Space
T.S. Rappaport Ch 4-5
NDG Notes
20
Mobile Radio Propagation Multipath Channel: Impulse Response Model ∝
y(d, t) = x(t)⊗h(d,t) = ∫ - ∝ x(τ) h(d, t- τ) dτ y(d, t) = t∫ - ∝ x(τ) h(d, t- τ) dτ when h(d,t) = 0 for t>0. When MS is moving at a constant speed v, then d = vt, so, y(vt, t) = t∫ - ∝ x(τ) h(vt, t- τ) dτ Since v is constant, so, y(vt, t) is a function of time, y(t) = t∫ - ∝ x(τ) h(vt, t- τ) dτ = x(t)⊗h(vt,t) = x(t)⊗h(d,t) As the velocity can be assumed constant over a small time interval (short distance), so, Mobile Radio Channel can be assumed as a time varying multi-path channel where t represents time variation due to MS in motion and τ represents the channel multi-path delay for a fixed value of t. T.S. Rappaport Ch 4-5
NDG Notes
21
Mobile Radio Propagation Multipath Channel: Impulse Response Model ∝
y(t) = ∫ - ∝ x(τ) h(t, τ) dτ = x(t)⊗h(t, τ) and r(t) = c(t) ⊗ ½(hb (t, τ) ) x(t)
c(t)
h(t, τ ) = Re { hb (t, τ) exp (jwct)}
h(t, τ ) = ½ hb (t, τ)
y(t)
r(t)
Fig-3: Base band channel response model and its equivalent RI
Here c(t) and r(t) are the complex envelop of x(t) and y(t) such that x(t) = Re { c(t) exp (j2π fc t)} and y(t) = Re { r(t) exp (j2π fc t)} Average Power of BP signal x2 (t) = ½ |c(t) |2, over bar denotes ensemble average of stochastic signal. T.S. Rappaport Ch 4-5
NDG Notes
22
Mobile Radio Propagation Multipath Channel: Impulse Response Model If τ0 represents the excess time delay of the first arriving multipath signal and then ∆τ = τ1 - τ0 but as τ0 = 0, so, ∆τ = τ1 . τi = i x . ∆τ , and if there are N multi-path components, then maximum excess delay of the channel is N x ∆τ This model can be used to analyze the transmitted signals having BW which are less than 1/2 ∆τ hb (t, τ) = Σ ai (t, τ) exp [j(ωc τi(t)+ φi(t, τ) )] δ (t, τi(t)). Sum over I = 0 to N-1 where ai (t, τ) and τi(t)) are real amplitude and excess delay of ith multipath component at time t. The power delay profile of the channel is found by taking the spatial average of | hb (t, τ) |2 over a local area. The received power delay profile is given P(t, τ) ≈ k| hb (t, τ) |2 , k is the gain relating the transmitted power in the probing pulse to the total power received. T.S. Rappaport Ch 4-5
NDG Notes
23
Mobile Radio Propagation Multipath Channel: Impulse Response Model If a CW signal is transmitted into the channel, let the complex envelop be given by c(t) = 2, then, instantaneous complex envelop of received signal is given as r(t) = Σ ai exp [jθi(t, τ) )] , summation over i = 0 to N-1 and instantaneous received power is | r(t)|2= |Σ ai exp [jθi(t, τ) )]|2 , summation over i = 0 to N-1
EXERCISE: Do Example-4.2 and Example-4.3 carefully at home.
T.S. Rappaport Ch 4-5
NDG Notes
24
Mobile Radio Propagation Small-Scale Multipath Measurements Channel Sounding Techniques Used to determine Power Delay Profile/Channel Impulse Response of any channel
Direct RF Pulse System
Fig-01: Direct RF Channel Impulse Response Measurement System
T.S. Rappaport Ch 4-5
NDG Notes
25
Mobile Radio Propagation Small-Scale Multipath Measurements Channel Sounding Techniques Spread Spectrum Channel Impulse Response Measurement System
Fig-02: Spread Spectrum Channel Impulse Response Measurement System
T.S. Rappaport Ch 4-5
NDG Notes
26
Mobile Radio Propagation Small-Scale Multipath Measurements Channel Sounding Techniques Spread Spectrum Channel Impulse Response Measurement System Advantages Rejects pass-band noise Improves the coverage range of a given Tx power No need of Tx and Rx PN synchronization-thanks to sliding correlator Due to inherent processing gain, Tx power required is considerably less Disadvantages No real time measurements Measurement time for power delay profile may be relatively excessive Due to non-coherent detection, information about the phases of individual multi-path components is lost
T.S. Rappaport Ch 4-5
NDG Notes
27
Mobile Radio Propagation
Important Multi-path Channel Parameters
The Mean Excess Delay τ : It is also called first moment of the power delay profile and is defined as ………………..E-01
The RMS Delay Spread σ τ : It is the square root of the second central moment of the power delay profile and is given as ………………..E-02
where ………………..E-03
* These delays are measured wrt the first detectable multi-path
signal component arriving at Rx at τ0 = 0. ** Typical values for στ are in usec for outdoor channels and nsec for indoor mobile channels.
T.S. Rappaport Ch 4-5
NDG Notes
28
Mobile Radio Propagation
Important Multi-path Channel Parameters
The Maximum Excess Delay τX : The max. excess delay at which a multi-path component is within X dB from the strongest multi-path component (may not necessarily arriving at τ0 ). Also known as excess delay spread of the power delay profile.
Fig-3: Example of an indoor power delay profile, mean excess delay, rms delay spread and excess delay spread
T.S. Rappaport Ch 4-5
NDG Notes
29
Mobile Radio Propagation
Important Multi-path Channel Parameters
Coherence Bandwidth Bc It is a defined relation derived from the rms delay spread στ. Range of frequencies over which the mobile channel can be considered flat (a channel that passes all frequencies with almost equal gain and linear phase). Bc ≈ 1/ 50 στ [for frequency correlation function of above 0.9] ≈ 1/5 στ [for frequency correlation function of above 0.5]
EXAMPLE-01: A multi-path power delay profile is given in Fig-04. Calculate mean excess delay, rms delay spread, and max. excess delay(10 dB). Estimate the 50% coherence bandwidth of the channel. Would it be suitable for AMPS or GSM without equalizers? T.S. Rappaport Ch 4-5
NDG Notes
30
Mobile Radio Propagation
Important Multi-path Channel Parameters
Doppler Spread BD and Coherence Time TC Describe the time varying nature of the channel in a small-scale region.
BD = fm = v/c * fc, the maximum doppler shift TC ≈ 1/ fm TC ≈ 0.423/ fm [A popular rule of thumb] EXAMPLE-02: Assuming the consecutive samples having high time correlation, calculate the proper spatial interval required for proper small-scale propagation measurements. If the test van runs at v = 120 km/h and fc = 900 MHz, how many samples would be required in 10 m interval. Assuming that measurements can be made in real time, how much time it take to make these measurements. What is Doppler spread for the channel? SOLUTION : Worst case Tc = 9/16 *Pi* fm = 9C/16*Pi* v fc = ? Samples should be taken at a rate less than 0.5 *Tc = ? ∆x = vTc/2 =? So, number of samples reqd. over 10 m distance = 10/ ∆x . T.S. Rappaport Ch 4-5
NDG Notes
31
Mobile Radio Propagation Types of Small-Scale Fading Depending on the signal parameters [BW, Symbol Period] and channel parameters [ rms delay spread and Doppler spread], in a multi-path environment transmitted signals under go different kinds of fading: Multi-path delay spread causes:
Flat Fading : BW of signal < BW of channel and delay spread < symbol period, most common fading, amplitude and phase of the transmitted signal is changed depending upon the multi-path but no frequency shift is observed [spectral characteristics preserved]. Frequency Selective Fading: BW of signal > BW of the channel and delay spread > symbol period
Doppler Spread leads to
Fast Fading: Higher BD, TC < symbol period, rapid fluctuations of signal strength due to very high variations in channel Impulse Response faster than base-band signal variations Slow Fading: Just opposite to fast fading T.S. Rappaport Ch 4-5
NDG Notes
32
Mobile Radio Propagation Fading Effects due to Multi-Path Time Delay Spread Flat Fading: In a multi-path radio channel, signal will undergo a flat fading if Bs << Bc and Ts >> rms delay spread of the channel
Spectral characteristics of the signal remain unchanged. Amplitude and phase of the transmitted signal will change in Flat Fading Channel. [Also known as Amplitude Varying Channel or Narrowband Channel]
Fig-1: Flat Fading Channel Characteristics
T.S. Rappaport Ch 4-5
NDG Notes
33
Mobile Radio Propagation Fading Effects due to Multi-Path Time Delay Spread Frequency Selective Fading: In a multi-path radio channel, signal will undergo a frequency selective fading if Bs > Bc and Ts < rms delay spread of the channel Amplitude and phase of the transmitted signal will change as well as spectral characteristics in Frequency Selective Fading Channel. [Also known as wideband Channel] ROT: Ts < 10 στ will make a channel FSFC.
Fig-2: Frequency Selective Fading Channel Characteristics
T.S. Rappaport Ch 4-5
NDG Notes
34
Mobile Radio Propagation Fading Effects due to Doppler Spread Fast Fading: In a multi-path radio channel, signal will undergoes a fast fading if Ts > Tc and Bs < BD Channel impulse response changes rapidly within the symbol duration. Leads to frequency dispersion which in turn increases signal distortion. Slow Fading: In a multi-path radio channel, signal will undergoes a slow fading if Ts << Tc and Bs >> BD Channel impulse response varies at a rate much slower than the transmitted base-band signal. {Static Channel] ** Fast and Slow Fading should not be confused with Large-Scale and Small-Scale fading. T.S. Rappaport Ch 4-5
NDG Notes
35
Mobile Radio Propagation Fading Effects due to Doppler Spread
Fig-3: Fading of signal as function of Symbol Period and Base-band Signal BW
T.S. Rappaport Ch 4-5
NDG Notes
36
Mobile Radio Propagation
Classification of the Channels
T.S. Rappaport Ch 4-5
Channel Corelation bandwidth
Signal Bandwidth/
Time-Flat Channel: A time-invariant channel which remains flat (constant) during at least the transmission time of one symbol. Frequency-Flat Channel: A frequency non-selective channel in which bandwidth of the transmitted signal Bs is less than coherence bandwidth Bc of the channel. Flat-Flat Channel: A time-flat and frequency flat channel. Non-Flat Channel: A time-variant and frequency selective channel Time-Flat Non-Flat
Flat-Flat
Frequenc y-Flat
Fig-4: Classification of the Channels
Symbol Time/Channel Correlation Time
NDG Notes
37
Mobile Radio Propagation Link Budget Calculation Large-Scale Fading Shadowing Effect Small-Scale [Fast] Fading
Fig-6: Link Budget Calculation for Fading Channels [AB-2000]
T.S. Rappaport Ch 4-5
NDG Notes
38
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading Basic Assumptions
Takes into account only Scattering mechanism No direct line-of-sight path considered A fixed Tx with vertical polarization N azimuthal plane waves with arbitrary phases and angle of incidence and equal average amplitude
Fig-01: N Azimuthal plane waves arriving at random angles
T.S. Rappaport Ch 4-5
NDG Notes
39
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading
Mobile Rx is in motion at velocity v in x-direction. The Doppler shift of nth plane wave arriving at angle is given as fn = v cosα n / λ
It can be shown* that random received signal envelop r has a Rayleigh Distribution given by p ( r ) = (r/ σ 2 ) exp ( - r2 / 2 σ 2 ) for 0 <= r <= ∞ =0 for r < 0
* Derivation follows, as given in the TB pp-178-179 T.S. Rappaport Ch 4-5
NDG Notes
40
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading Spectral Shape in Clark’s Model
Gans developed a spectrum analysis for Clark’s Model Total received power at Rx can be expressed as
and instantaneous frequency of the received signal component is given as
And it can be shown* that output spectrum is [see Fig-02]
* Derivation is as given in TB pp-179-180 T.S. Rappaport Ch 4-5
NDG Notes
41
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading Spectral Shape in Clark’s Model
Fig-02: Doppler Power Spectrum for an un-modulated CW carrier
T.S. Rappaport Ch 4-5
NDG Notes
42
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading Simulation of Clark’s Model
Fig-03: Clarke’s Simulator using AM with a [a] Doppler Filter [b] Base-Band Doppler Filter
T.S. Rappaport Ch 4-5
NDG Notes
43
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading Level Crossing and Fading Statistics
Important statistical parameters useful for designing Error Control Codes Level Crossing Rate (LCR) is expected rate at which Rayleigh fading envelop, normalized to the local rms signal level, crosses a specified level in positive-going direction. And this number is given by
Where r is d (r(t))/dt, p(R,r) is the joint density function of r and r at r = R, and fm is max. Doppler shift, and ρ = R/Rrms Average Fade Duration is the average period of time for which the received signal is below a specified level R. For Rayleigh fading signal it is given by
T.S. Rappaport Ch 4-5
NDG Notes
44
Mobile Radio Propagation Statistical Models for Multi-path Fading Channels Clark’s Model for Flat Fading Example-1: For a Raleigh fading signal, calculate N such that ρ = 1 and fm = 20 Hz. If fc = 900 MHz what would be the velocity of the mobile to get this Doppler shift? Example-2: Calculate the Average fade duration for previous example. Repeat it when threshold level is reduced to ρ
T.S. Rappaport Ch 4-5
= 0.01.
NDG Notes
45