Documente Academic
Documente Profesional
Documente Cultură
Foreword
This Course covers the main topics in Radio Propagation, with application to the Engineering of Digital Radio-Relay Links. The aim is to provide the radio engineer with the basic knowledge and understanding of radio propagation phenomena and their impact on the operation and performance of digital radio systems. The Course makes reference to fundamentals in Radio Propagation Physics (without need of complex mathematical tools in Electromagnetics theory). From this, it derives the basic concepts in Radio Link Engineering. A detailed presentation of procedures and computer tools for the Engineering and Planning of Radio Systems is out of the scope of this course. The Course User is assumed to be familiar with elementary notions in Digital Radio Modulations, Equipments, and Systems, as well as in Interference Analysis and Planning and in Regulatory Issues. Some topics in the above areas are discussed, but only in connection with propagation aspects. Hypertextual techniques have been adopted. It is expected that this choice will improve significantly the flexibility and effectiveness of Computer Based Training, both in Class Presentations and in Individual (or Small Group) Use.
Author
USER GUIDE Navigation through the Radio Propagation Tutorial is controlled by clicking on the following symbols : Go to Index Go to the Next Page (in a Chapter Sequence) Go to the Previous Page (in a Chapter Sequence) One Step Back (when this symbol is not available, use Keys : CTRL --) Go to Additional Related Topics Go to Computational Examples
Author
INDEX
1. Introduction Radio Link Equation Radio Link Engineering 2. Refraction through the Atmosphere Ray Curvature - Clearance and Diffraction 3. Refraction through the Atmosphere Multipath Propagation 4. Ground Reflections 5. Atmosphere and Rain Attenuation Rain Scattering and Absorption Absorption without rain 6. Propagation Related Topics Regulatory Background Interference Scenarios
HELP
About ... User Guide
Tx
Rx
The diagram shows the basic elements in any point-to-point radio link. TRANSMITTER ANTENNAS RADIO HOP RECEIVER : : : : Frequency, Tx Power, Capacity (bit/s, tel. chs., ...). Frequency range, Gain. Hop Length. Frequency, Rx Threshold (related to Rx Signal Quality).
Warning : The constant 92.4 is correct only if the frequency is expressed in GHz and the hop length in km. If other units are used, the constant 92.4 must be modified accordingly (e.g. : with hop length in miles, the constant is 96.6).
The FSL increases 6 dB if : the hop length is doubled; or the frequency is doubled. Examples: 1.9 GHz 3.8 GHz 7.6 GHz 15.2 GHz 60 km 60 km 30 km 30 km FSL = 133 dB FSL = 139 dB FSL = 139 dB FSL = 145 dB
Frequency (GHz)
= Antenna Efficiency = 0.55 - 0.65 Parabolic antenna : G = (D/)2 In dB units : G = 20 Log(D) + 20 Log(F) + 18.2 + 0.5 (depending on ).
50
4m
46 42 38 34 30 0 5
antenna diameter is doubled, for a given frequency. frequency is doubled, for a given diameter.
0.5m
10
15
20
Frequency [GHz]
A more detailed Link Budget may include additional losses caused by Tx/Rx components (feeders, branching, etc.) and by propagation impairments.
A Fade Margin is required to compensate for the reduction in Rx power caused by Propagation Anomalies. The Fade Margin guarantees that the link will operate with acceptable quality, even if propagation anomalies causes Additional Losses (AdL), as long as the Additional Loss is lower than the Fade Margin :
Received Power
Fade Margin
Threshold
Time
Assuming the RX Threshold PTH = -77 dBm, then the Fade Margin is : FM = PR - PTH = 43 dB
END OF CHAPTER
Tx
CL
Rx
The Clearance CL is defined as the distance of the radio ray to the ground. A negative Clearance means that a ground obstruction is higher than the ray.
Real Earth
Tx
CL
Rx
R Tx CL
Real Earth Rx
kR
Equivalent Earth
Tx
CL
Rx
R Tx
Real Earth Rx
CL
Flat Earth
Tx
CL
kR
Equivalent Earth Rx
CL
Flat Earth
2.7 - VISIBILITY
Point-to-Point Radio Relay Links are usually designed under the requirement of Visibility between the two Hop Terminals. Two factors to be considered in Defining Visibility Criteria in a Radio Path: Variability in Atmospheric conditions producing different ray curvatures Tool : Statistical distributions of the k-factor. Objective : Find the Typical and Minimum k-factors appropriate for that path. Effects of partial obstructions along the radio path Tool : Fresnel Ellipsoids and Diffraction Analysis. Objective : Identify the Clearance Rules to be applied (minimum distance between the Radio Ray and the Ground).
The ITU-R gives a curve of minimum kEFF values as a function of hop length (temperate climate). This curve can be used in establishing the worst case condition to check Visibility Criteria.
10 20 50 100 200
keff
Rx Tx
F1 P
About half of the Rx signal energy travels through the Fresnel ellipsoid. So, any L1 obstruction within the L2 Fresnel ellipsoid has some impact on the Rx power. F1 is the Fresnel Ellipsoid Radius, depending on distance L1 from the hop terminal.
F = Frequency (GHz)
60 50 40 30 20 10 0 0 20 40 60 80 100
2 GHz 4 GHz 7 GHz 12 GHz
In the figure : Fresnel Ellipsoid Radius at the middle of the path (L1=0.5L). Other Path positions : R (L1=0.3L) = R (L1=0.1L) = R (L1=0.01L) = 0.92 * R (L1=0.5L) 0.6 * R (L1=0.5L) 0.2 * R (L1=0.5L)
0 10 1 20 3 30 40 -1.5 2
In the figure : Loss in Rx Power caused by obstructions with different shapes : 1) Knife-edge; 2) Earth curvature (beyond the horizon link); 3) Intermediate case. The Normalized Clearance (x-axis) is positive for obstacles below the ray, and negative for obstacles above the ray.
-1
-0.5
0.
0.5
1.
When a frequency below 2 GHz is used : CL > 60% F1 with Typical k CL > 30% F1 with Minimum k.
END OF CHAPTER
Tx
Rx
n
Under Multipath Propagation conditions, total Rx several "echoes" of the Tx signal arrive at the Rx antenna with random amplitude, delay, and phase shift. The received signal echo 1 can be represented as the addition of multiple Vectors.
Multipath activity depends on environmental conditions and on radio link parameters.Particularly in tropical climates, long multipath events can be observed.
Rx Power
Time
Prob { A > Ao }
10 dB/dec
The Probability of having a fade depth A (dB) greater than a given depth Ao is (Rayleigh formula) : Prob {A>Ao} = Po 10-Ao/10 Po = Multipath Occurrence Factor. It is a measure of the multipath activity in a radio hop.
10
20
30
40
50
Ao [dB]
Operating Hop :
? How to know Po ?
Po is estimated by monitoring the Rx Power and by processing the measured data. Hop under Design : Po is predicted by empirical propagation models.
Po can be predicted using empirical formulas, proposed by ITU-R and by operating companies or research labs. A general formula is : Po = K Q Fa Lb where : F L K Q = = = = Frequency; Hop Length; Geo-climatic Coefficient; Terrain Coefficient;
The Frequency Exponent is close to unity. This means that the fading activity in a given hop is proportional to the frequency (at 11 GHz is approximately twice than at 5.5 GHz). The Distance Exponent is in the range 3 - 3.6. This means that the fading activity , for a given frequency, climate, and terrain, is increased about ten times if the hop length is doubled.
Linear scaling to other frequencies : divide by 2 the Po values computed for the 6 GHz hop 3 GHz 12 GHz multiply for 2 " " " " " " " ? How to use Po predictions ? Fade Depth > 30 dB for 2600 seconds / one month. Any Fade Depth : divide seconds by ten for 10 dB deeper fade (e.g. Po = 1 Fade Depth > 40 dB for 260 seconds / one month). Other Po values : linear scaling (number of seconds proportional to Po). Po = 1
Fade Depth
Notch Slope
Frequency Frequency selectivity is fast varying. This dynamic effect can be observed as a fast-varying slope or even as a fast moving notch trough the signal bandwidth.
The Notch Depth and Frequency are varied (adjusting amplitude and phase of direct and echo signals). In each condition the Bit Error Ratio (BER) is measured. In the Notch Depth / Notch Frequency plane, the Signature gives the region (Notch parameters) with BER > 10-3 (or any other threshold). The area below the Signature gives a measure of the receiver sensitivity to multipath distortions.
-15
-10
-5
10
15
For each Multipath Countermeasure, the Improvement Factor IF is defined : IF = Tout (Unprotected) / Tout (with Countermeasure) where Tout = outage time. The Improvement Factor can be predicted by means of Statistical Models including the effect of each specific countermeasure. When both Equalization and Diversity are used, the overall Improvement Factor is approximately given by the product of the factors relevant to each technique (Synergistic Effect).
With Equalizer
The effectiveness of a signal equalizer can be appreciated by comparing the receiver signatures with and without the equalizer. The reduction in the area below the signature curve gives a measure of the improvement provided by the equalizer.
15
-10
-5
10
END OF CHAPTER
4.1 - GROUND REFLECTIONS Depending on the Path Profile, it may happen that a portion of the Tx radio signal is reflected by the ground toward the Rx antenna. At the receiver, in addition to the direct signal (D), a reflected signal (R) arrives. In most cases, the presence of a ground reflection is rather critical : Fluctuations in the Rx signal level, even for long time periods; Enhancement of Multipath Activity (the reflected signal is not added to a stable direct signal, but to the fast-varying multipath signal); Reduction of Space Diversity effectiveness as a countermeasure to multipath. As far as possible, reflections should be avoided by : Route Planning (in particular over-water paths); Site Selection : Obstruction of the reflected ray can be obtained in some cases, by suitable selection of the radio sites and of antenna heights.
Rx Tx
1 R1
P
Reflection point P; Grazing angle ; Direct path length D; Reflected path length R1+R2; Angles 1, 2 between Direct and Refl. Rays.
2 R2
These parameters are varying with time, because of varying propagation conditions (k-factor).
Tx
Rx
Rx Level
<= Ts : frequency selective attenuation within the signal bandwidth; the max.
in-band dispersion (notch) depends on the reflected signal level; a measure of the Rx sensitivity is given by the "equipment signature".
END OF CHAPTER
A radio wave travelling through rain drops is subject to scattering and absorption phenomena. In this process, part of the signal energy directed to the receiver is lost. The Rain Attenuation : is measured in dB/km; increases with frequency, but can be assumed as uniform (flat) within a radio channel bandwidth; increases with rain intensity; with Horizontal Pol. is higher than with Vertical Pol.; produces fades usually several minutes long.
100
100
500
Frequency [GHz]
% of Time
0.1
From the Time % vs. Rain Attenuation curve, the Unavailability is computed as the time percentage with attenuation greater than the Fade Margin. In the Figure a 38 dB Fade Margin is assumed. Then the Rain Unavailability is about 0.0025% (about 12 minutes / year)..
0.01
0.001
10
20
30
40
50
The prediction method derives from long-term rain rate statistics. Therefore, the Rain Unavailability Prediction must be considered as an average, to be expected during a period of several years.
Attenuation [dB]
Attenuation [dB/km]
10
O2
Frequency Bands affected by Atmospheric Absorption Peaks : 22.5 GHz (water vapour) Max Attenuation = 0.18 dB / km
H2 O
0.1 0.01 10
60 GHz (oxygen) Max Attenuation = 16 dB / km. The figure is valid for : sea level temperature : 15 C water vapour : 7.5 g/m3
20
50
100
Frequency [GHz]
END OF CHAPTER
AVAILABILITY OBJECTIVES based on : Definition of Availability : After 10 consecutive SES events , Unavailability is detected. The 10 seconds are part of the Unavailable time. After 10 consecutive non-SES events Availability is detected. The 10 seconds are part of the Available time. Max. Unavailable Time Percentage Note : Error Performance Objectives are checked only during Available Time.
Definition of Errored Second (ES) : One Second with >= 1 Errored Bit (Block). Different definitions for Severely Errored Seconds (SES). Only in G.826 : Background Block Error (BBE) ES / total available seconds : < 0.08 in G.821; depending on bit rate in G.826; SES / total available seconds : < 0.002 both in G.821 and G.826.
< 10 seconds
Not Significant
Criteria followed in the ITU-R Frequency Plan activity : Below 12 GHz : Compatibility of Channel Arrangements in the transition from Analog to Digital systems. Above 12 GHz : Channel Arrangements optimized for Digital systems.
Pol.
GO CHANNELS x 1 3
H(V) V(H) z 2
x/2 x/2
...
4
N-1
...
4
N z
Interferences : two adjacent X-pol. signals, channel spacing F = x/2 two adjacent Co-pol. signals, channel spacing F = x Comment : Recommended in analog systems (frequency reuse not possible) Adopted also for digital systems below 12 GHz.
...
...
F y = Central
guard band z = Edge guard band
Interferences : one X-pol. signal at the same frequency (F = 0) two adjacent Co-pol. signals, channel spacing F = x two adjacent X-pol. signals, channel spacing F = x Comment : Suitable for digital systems, short hops, simple modulations Recommended in frequency bands above 12 GHz (only digital) Recently adopted in some frequency bands below 12 GHz (x-pol. interference canceller required for complex digital modulations)
In te rfe re d R e c e iv e r
C lic k o n fo r d e ta ils
END OF CHAPTER
Propagation issues :
Propagation issues :
50
4m
46 42
3m
2m 1m
38 34 30 0 5 10 15
0.5m
20
Frequency [GHz]
Tx
CL
Rx
Real Earth
Tx
CL
Rx
R Tx CL
Real Earth Rx
kR
Equivalent Earth
Tx
CL
Rx
R Tx CL
Real Earth Rx
Flat Earth
1.1 0.9
keff
1.1 0.9
keff
Rx Tx
F1 P
L1
L2
60 50 40 30 20 10 0 0 20 40 60 80 100
2 GHz 4 GHz 7 GHz 12 GHz
-10
0 10 1 20 3 30 40 -1.5 2
-1
-0.5
0.
0.5
1.
-15
-10
-5
10
15
Without Equalizer
With Equalizer
-15
-10
-5
10
15
Rx Tx
1
D
P
R1
R2
Tx
Rx
Rx Level
100
100
500
Frequency [GHz]
% of Time
0.1
0.01
0.001
10
20
30
40
50
Attenuation [dB]
Attenuation [dB/km]
10
O2
H2O
0.1 0.01 10
20
50
100
Frequency [GHz]
Pol.
GO CHANNELS x 1 3
H(V) V(H) z 2
x/2 x/2
...
4
N-1
...
4
N z
CO-CHANNEL PLAN
RETURN CHANNELS y z 1 3
...
...
N
CO-CHANNEL PLAN
RETURN CHANNELS y z 1 3
...
...
N
Frequency (GHz)
In this Figure : Po = 1
Prob { A > Ao }
10 dB/dec
10
20
30
40
50
Ao [dB]
10
Amplitude [dB]
Frequency
b=0.95
b=0.9 b=0.5
fo
Frequency
fo+1/
To Probe Further : Derivation of the Radio Link Equation Definition of Logarithmic Units
To Probe Further : Loss vs. Distance in Radio and Cables Derivation of the Radio Link Equation
To Probe Further : Atmosphere Structure and Refraction Effects of Vertical Refractivity Gradient
To Probe Further : Atmosphere Structure and Refraction Effects of Vertical Refractivity Gradient
To Probe Further : Effects of Vertical Refractivity Gradient Vectorial Addition of Multiple Signals
To Probe Further : Vectorial Addition of Multiple Signals Rx Signal Level vs. Time and Frequency during Multipath Events
To Probe Further : Rx Signal Level vs. Time and Frequency during Multipath Events
To Probe Further : Selective Fading Prediction Models Signature Measurement De-Polarization due to Multipath
To Probe Further : Selective Fading Prediction Model Signature Measurement De-Polarization due to Multipath Maximum Hop Length (Rain & Multipath)
To Probe Further : Radio Wave Propagation through Rain De-Polarization due to Rain Other Hydrometeor Effects (fog, snow)
To Probe Further : ITU-R Model for Rain Unavailability Prediction De-Polarization due to Rain
To Probe Further : ITU-R Model for Rain Unavailability Prediction Maximum Hop Length (Rain & Multipath)
To Probe Further : De-Polarization due to Rain Other Hydrometeor Effects (fog, snow)
To Probe Further : Angle Diversity Diversity Improvement Selective Fading Prediction Models
Rx Power (actual) : The RF power at the Radio Receiver Input, at a given time, including propagation losses present at that time. Rx Threshold : The minimum RF power at the Rx input, required for the Receiver to operate above a threshold of acceptable quality. Reception Quality : The result of comparison between the original information delivered at the Tx input and that available at the Rx output; it can be expressed by the Bit Error Ratio (BER), or by other (more complex) parameters related to Bit Blocks. Rx Selectivity : The overall effect of the Receiver Filters in discriminating the desired signal from signals received on adjacent radio channels. It is expressed by an overall transfer function, including the contribute of RF, IF, and BB filters. Net Filter Discrimination (NFD) : The attenuation of the Interfering signal level at the Rx decision circuit, as a result of the interfering signal Tx spectral shaping and of the Rx selectivity at the interfered receiver.
GLOSSARY - Antennas
Isotropic Antenna : An ideal source of Electromagnetic Radiation, that radiates uniformly in all directions. Omnidirectional Antenna : A Real Antenna, that approximates an Isotropic Antenna; in most cases radiation is (almost) uniform at all azimuth angles, but within a limited range of elevatuion angles. Directive Antenna : A Real Antenna, that concentrates most of the emitted radiation within a small angle in a given direction. Reflector Antenna : A Directive Antenna, using one or more reflecting surfaces to concentrate the emitted radiation in the desired direction. Parabolic Antenna : A Reflector Antenna using a Parabolic Surface; the feeder position is in the parabola focus. Horn Antenna : A Reflector Antenna using a sector of a Parabolic Surface. Cassegrain Antenna : A double Reflector Antenna: the primary reflector is a parabolic surface, the secondary reflector is a hyperbolic surface.
Glossary - Antennas
Antenna Gain : See Definition. Directivity Diagram : A plot of the antenna gain (usually relative to the maximum gain) as a function of the azimuth or elevation angle. Far Field Region : The region sufficiently distant from the antenna, where the EM field can be represented as a plane wave and the antenna diagram is stabilized. Closer to the antenna, the Near Field Region and the Fresnel (transition) Region are defined. The boundary between the Fresnel and the Far Field Region is approximately at the distance : dFF = 2 D2 / (D = antenna diameter)
Launch and Arrival Angles : The angle of the Radio Ray at the Tx or Rx antenna, with respect to the Horizon; it varies with atmospheric conditions (k-factor).
Logarithmic Units
Logarithmic units are widely used in Radio Link Engineering computations. The Link Budget is put in a very convenient form by using logarithmic units. Gains and Losses are added with positive or negative sign, as in financial budgets. The ratio R between the power P and a reference power PREF (both in the same unit, like W or mW) is expressed in decibels as: R (dB) = 10 Log10 (P / PREF) If R>0, then P > PREF (gain). Otherwise, if R<0, then P < PREF (loss). NOTE : Decibels are used to express ratios only. Absolute power levels can never be expressed in dB. Sometimes the reference power level PREF is implicit in some way. Absolute power levels can be expressed in dBm, by assuming PREF = 1 mW : P (dBm) = 10 Log10 (P[mW] / 1mW) Similarly, with PREF = 1 W, power P in dBW is : P (dBW) = 10 Log10 (P[W] / 1W)
Logarithmic Units
Examples of power levels expressed in different units : Power in mW 0.001 0.1 1.0 2.0 5.0 10.0 40.0 100.0 200.0 1000.0 5000.0 10000.0 Power in dBm -30 -10 0 +3 +7 +10 +16 +20 +23 +30 +37 +40 Power in dBW -60 -40 -30 -27 -23 -20 -14 -10 -7 0 +7 +10
By transforming into Logarithmic Units, we get : PR[dBm] = PT[dBm] + GT[dB] + GR[dB] - 92.44 - 20 Log (F[GHz]) - 20 Log(L[km]) which is the usual Radio Link Equation. If frequency and distance are not expressed, respectively, in GHz and km, then the numerical constant 92.4 must be modified accordingly. Comments : Two equivalent forms of the Radio Link Equation have been derived above: PR = PT GT AR / 4*L2 PR = PT GT GR [c / 4*FL]2 (1) (2)
The antenna parameters used in (1) are the Tx gain GT and the Rx antenna area AR. The Rx power is proportional to GT, to AR and to PT; it is inversely proportional to the distance squared. This probably sounds quite clear from the physical point of view. On the other hand, formula (2) looks attractive for its symmetric form, since both the Tx and Rx antenna gains appear. However, we must remember that, for a given antenna area, the gain increases with frequency (see also Antenna Gain Definition). That's why the frequency term appears in (2).
The Antenna Gain depends on the physical dimensions of the antenna, normalized to the signal wavelength &. For Reflector Antennas we have : G (dB) = 10 Log (4 * A / &2) = 10 Log (4 * AE / &2) where A is the reflector area, is the Antenna Efficiency (generally in the range 0.55 0.65) and AE = A is called Antenna Effective Area.
EM energy interacts with the propagation medium and some phenomena of energy transfer or transformations produce the attenuation in the received signal. This happens, for example, in coaxial cable transmission, where signal loss is mainly due to dissipative phenomena (interaction of EM energy with conductive and dielectric material in the cable). Depending on signal frequency and cable characteristics, some fraction of signal power is lost every kilometre travelled through the cable. So, the loss is usually expressed in dB/km. Also in radio communications through the atmosphere, dissipative phenomena can be observed. Absorption in the atmosphere is caused by water vapour, oxygen molecules or by water in raindrops. Also for these phenomena, the power loss is usually expressed in dB/km. However, in most cases (dry atmosphere, frequencies below 20 GHz), the interaction of the EM radiation with atmosphere components is almost negligible.
The average value of N at sea level is about No = 315 The ITU-R gives world maps with the mean values of No in February and August. The Vertical Refractivity Gradient G (measured in N-units per km, N/km) is defined as: G = (N1 - N2) / (H1 - H2) where N1 and N2 are the refractivity values at elevations H1 and H2, respectively. In a Standard Atmosphere model, the Vertical Refractivity Gradient is assumed as constant in the first kilometre of the atmosphere : G = - 40 N/km. This corresponds to the Standard Propagation conditions. Deviation from the Standard Atmosphere condition leads to Anomalous Propagation. Such anomalies are usually associated with particular meteorological conditions, like temperature inversion, very high evaporation and humidity, passage of cold air over warm surfaces or vice versa. In this conditions, the Vertical Refractivity Gradient is no longer constant. A number of different profiles have been observed and measured. It is worth noting that, at greater altitude, the Refractive Index is, in any case, closer and closer to 1; so the Refractivity N goes to zero, according to an exponential function.
For the Standard Atmosphere (G = -40 N units/km), this gives k = 4/3 ( = 1.33).
Anomalous propagation conditions : Constant-G profiles : deviation from the Standard Atmosphere (k = 4/3), can result in : Sub-Refractive propagation (k < 4/3, G > -40 N-units/km) : usually associated to atmosphere density increasing with height (warm air over cool air or moist surface). The ray curvature is reduced or even is bent upward ( k < 1, G > 0 ); the ray path is closer to the ground (maximum obstruction probability when k is minimum). Super-Refractive propagation (k > 4/3, G < - 40 N-units/km) : observed when temperature inversion happens or other phenomena makes atmosphere density decreasing with height. (cool dry air over a warm body of water). The equivalent earth reduces its curvature; for k the ray is parallel to the earth and propagation may extend its range (unexpected interference may appear). In extreme super-refraction conditions (G < -157 N-units/km, negative k) the ray is bent toward ground and no signal arrives at the Rx antenna (black-out). Variable-G profiles : with better approximation, the Refractivity Gradient can be assumed as constant only in limited height ranges (layered atmosphere). Under this model, the ray curvature changes when passing from one atmosphere layer (G = G1)
to the higher (or lower) one (G = G2). Even if, in the real case, the transition from one layer to another is smoothed in some way, a layered atmosphere model is useful in explaining : Multipath propagation : the different ray curvature in atmospheric layers may produce a number of separate propagation paths from the transmitter to the receiver. Duct formation : the atmosphere layers are such that the rays at the duct lower boundary tend to be bent upward, while rays close to the upper boundary tend to be bent downward. The result is that propagation is confined within a limited height ranges, with attenuation much lower than in well mixed atmosphere. If the Rx antenna is within the duct, a stronger signal will be received. If the Rx antenna is out of the duct, rather long signal fadings are observed.
Diffraction Analysis
Generally speaking, Diffraction effects can be observed when the wave propagation is altered by an obstacle which has dimensions comparable to the wavelength in the plane normal to the propagation direction. Usually, this means that the obstacle is close to the Fresnel ellipsoid (possibly with partial or total obstruction). The theoretical computation of diffraction loss is rather complex. Usually, reference is made to two obstacle models : the smooth spherical earth ; the knife-edge obstruction. They represent extreme and opposite conditions and most practical cases can be assumed as intermediate between these two models. The ITU-R Recs. 368 and 526 include the analysis of the two models. In Rec. 526 the Knife-edge model is generalized to rounded obstacles and the case of multiple obstructions is also dealt with. An approximation for the knife-edge case is given by : Diffraction Loss (dB) = 6.9 + 20 Log [ ( y2 +1 ) + y ] where y = 2 CL/F1 - 0.1 is a function of the clearance CL normalized to the Fresnel radius F1, and the approximation is valid for y > - 0.8 (Note : CL is negative when a ground obstruction is higher than the radio ray).
Amplitude [dB]
Frequency
The Group Delay is represented for different values of the echo amplitude b. In the figure it is assumed b < 1(delayed signal smaller than the direct one). This gives a Minimum-Phase Transfer Function. If the delayed signal is greater than the direct one, then a NonMinimum Phase Transfer Function is produced. This case is usually dealt with by assuming again the echo as the smaller signal, with a negative . The Group Delay diagram is given by the same figure, but reversing the ordinate axis.
b=0.95
b=0.9 b=0.5
fo
Frequency
fo+1/
With N (number of signals) growing to infinity, reference can be made to the Central Limit Theorem : X and Y are independent Gaussian Random Variables, whose average value is zero and variance is 2 (we are not interested in investigating the precise value of ). The amplitude R of the {X,Y} vector is given by : R2 = X2 + Y2 (2)
Given the statistical properties of X and Y (mentioned above), formula (2) defines R as a Rayleigh Random Variable, with average value (/2). The Rayleigh Cumulative Function gives : Prob { R < c } = 1 - exp [ - c2 / 22 ] c2 / 22 (3)
where the approximation holds when c << , or c << (Rayleigh average value). In the application to Radio Propagation Models, Logarithmic Units are usual. So, we are interested in evaluating probilities like ( C = 20 Log [c] ) : Prob [ R(dB) < C(dB) ] c2 / 22 = 10 C/10 / 22 which is the Rayleigh asymptotic formula in Logarithmic Units. If we define C(dB) = M(dB) - A(dB) , where A is the attenuation below the average value M in dB, then (4) becomes : Prob [ R(dB) < M(dB) - A(dB) ] 10 (M-A)/10 / 22 = W 10 -A/10 (5) (4)
where W is a constant depending only on the Rayleigh average value. Formula (5) is the usual asymptotic expression for Rayleigh distributed attenuation in dB, holding for R values much lower than the Rayleigh average value (A > 15 - 20 dB). Formula (5) is known as the 10dB / decade formula. If attenuation A is made 10 dB deeper, then the probablity is one decade smaller.
Reflection Coefficients
In many cases it is advisable to adopt the conservative assumption that the modulus of the Reflection Coefficient be equal to 1. For very small grazing angles and/or for Horizontal Polarization this is a realistic approximation. For Vertical Polarization, the table gives the Reflection Coefficient modulus in the case of sea reflection : Grazing Angle Frequency 0.5 1 2 4 1 GHz 3 GHz 10 GHz -1.4 dB -1.2 dB -0.8 dB -2.8 dB -2.3 dB -1.7 dB -6.2 dB -5.3 dB -4.3 dB -12.0 dB -10.9 dB -8.4 dB
Similarly, the phase of the Reflection Coefficient is very close to 180 with Horizontal Polarization (any grazing angle) and also with Vertical Polarization, for grazing angles below 0.5. With Vertical Polarization and larger grazing angles, the phase is smaller than 180, depending also on the frequency. However, in radio link design, the prediction of the Reflection Coefficient phase is usually not required.
Anti-Reflection System
Space Diversity, as a Reflection Countermeasure, is usually implemented to maximize the Rx power level, by switching to the best signal or combining in IF or RF. However, the maximized Rx signal is in some way a combination of the direct and the reflected signal. In some cases it is required that the reflected signal be cancelled, to avoid signal distortion (particularly with very long reflection delays). The Anti-Reflection system is an RF Space Diversity implementation with the objective of cancelling (or minimizing) the reflected signal component in the overall Rx signal. . The Antenna Spacing is such that the reflected signal arrives in phase opposition at the two antennas. The phase shifter is used for fine Hybrid Rx phase adjustment. The result at the hybrid is a co-phase addition of the direct signal components from the two antennas, and phase opposition cancelling of the reflected signal components.
Anti-Reflection System
The Space Diversity configuration can also be seen as an Antenna Array, with an overall directivity pattern having a null toward the reflection point. Also the Anti-Reflecting systems, as other Space Diversity implementations, are usually optimized for standard propagation conditions. Then the solution is checked in the whole range of expected k-factor values.
Frequency
n 4 3 Rx 2 1 n 4 n 4 3 2 1 1 Rx 3 2 1 2 n Rx 4 3
f2
f1
Rx
Time t2
.
t1
Multipath propagation is a dynamic event, both in the time dimension and in the frequency dimension.
Rx SIGNAL vs. TIME : Phase shifts are varying with time because of movements in atmospheric layers. Since the wavelength is rather small (e.g. 10 cm at 3 GHz), even a small change in the path length produces a significant phase rotation. So the relative phases among multipath components are fast varying, as well as the reultant vector (Rx amplitude). The speed of change of the multipath channel is expressed in two ways. 1) Rx signal power variation in a given time interval : up to 10 dB in 100 ms (100 dB/s) has been observed, according to several authors. 2) Fade Notch (deepest attenuation) movement through the signal bandwidth : up to 10 MHz in 100 ms (100 MHz/s) has been observed, according to several authors. The above estimates affect the design of multipath countermeasures, like adaptive equalizers, combiners, and switches to the protection channel (frequency diversity).
Rx SIGNAL vs. FREQUENCY : In a modulated signal, let us consider the spectral components at the edge of the signal bandwidth F : F1 = Fc - F/2 and F2 = Fc + F/2 (Fc = carrier frequency). Moreover, let us simplify the multipath propagation as the composition of a direct and a delayed (echo) signal. The phase shift between the F1 signal component and the delayed echo is : 1 = 2F1 where is the echo delay. Similarly, at F2, we have 2 = 2F2. The multipath propagation produces an (almost) flat attenuation through the whole signal bandwidth F, if the phase shifts 1 and 2 are equal (or very close) (same composition of the direct and echo signals at F1 and F2 ) : = 2 - 1 = 2 (F2 - F1) = 2 F 0 that is 2 F << 1 /
Under this condition, the frequency selective effect of multipath propagation can be neglected, Otherwise, for frequencies spaced more than F above, the composition of multipath components, at the same time instant, gives different Rx amplitudes (frequency selective multipath).
Formula for NW. Europe : This formula was derived from measurements performed by European PTT and reported by CCIR documents in 1978-1986. The parameters in the general Po formula are specialized as follows : Frequency Exponent a = 1; Distance Exponent b = 3.5; C = 1.4 10-8 (continental temperate climate, to be modified for very dry or humid or hot climates and/or wet terrain); Q = 1 (for average rolling terrain, to be increased for completely flat profile). Formula from ITU-R Rec. 530 : The present issue of ITU-R Rec. 530 reports a formula for predicting the Po factor in any radio link. The general Po formula is specialized as follows : Frequency Exponent a = 0.89; Distance Exponent b = 3.6;
coastal links, fraction of the path profile near large or medium bodies of water; plain, hilly or mountainous terrain; latitude and longitude). As an example, for inland links : C = 0.01 E pL1.5 where E = environment factor (in the range 2.2 10-8 to 5 10-6) pL = % time with average refractivity gradient in the lowest 100m of the atmosphere less than -100 N-units/km (from maps in ITU-R Rec. 453); where = slope (mrad) from Tx to Rx antenna.
Q = ( 1 + ) -1.4
The specific procedure for the computation of E is reported in ITU-R Rec. 530. Note : In the above formulas C has the same role as K in Rec. 530, but C = 0.01 K, since in Rec. 530 the Rayleigh formula is expressed in %.
Log(BER)
-2 -3 -4 -5 -6 -7 -8 60 50 40 30 20 10
When a digital signal is distorted by frequency selective fading, the Bit Error Ratio (BER) may be greater than the threshold (10-3) even if the Rx power is higher than the Rx threshold. In the figure, the continuous line (BER vs. Fade Depth) is measured in the laboratory with a flat fading simulator (no signal distortion). The Flat Fade Margin is 41 dB (fade depth corresponding to BER = 10-3)
The points at the right of the BER line are the result of field measurements, Fade Depth [dB] during selective fading. Each point is a measured BER, with the corresponding fade depth. The BER threshold is reached even with 20 dB fade depth only. Continuous Line : Flat Fading Bit errors caused by noise only. Points on the right : Multipath Selective Fading Bit errors caused by the joint effect of noise and signal distortion.
Signature Measurement
In order to simulate in the laboratory the distortions produced during multipath fading events, a Two-Ray Channel Model is usually adopted. Measurement Set-up :
Tx MOD
Rx
= echo signal
delay;
Bit Sequence
= echo signal
BER
DEM
Error Counter
Signature Measurement
Two-Ray Channel Transfer Function : H(f) = 1 - b exp [ - j (2f - ) ] = 1 - b exp [ - j 2(f - fo) ]
1/ Amplitude [dB]
0 -10
Echo delay : assumed as constant ( = 6.3 ns in the original Bell Labs / Rummler model ). Notch Frequency fo = / 2 : corresponds to the minimum of H(f) ; it is controlled by varying the echo phase ; the notch moves through (or out of) the signal bandwidth.
-20
Notch Depth B = - 20 Log10 (1 - b) : minimum of H(f) expressed in dB. It fo fo+1/ Frequency is controlled by the value of the echo amplitude b. The b-range is 0 - 1 : this corresponds to a Minimum-Phase Transfer Function (first signal larger than the delayed one). In the other case (first signal smaller) a Non-Minimum-Phase Transfer Function is produced. Assuming the smaller signal as the echo signal, this is referred as a "negative " condition.
-20log(1-b)
Signature Measurement
Measurement Procedure : The Bit Error Rate (BER) is measured by comparing the bit stream at the Tx input with the one estimated at the receiver. The following steps must be performed : a) Set the echo delay to a positive value (to get a minimum phase signature). b) Set the echo phase to the value corresponding to Notch Frequency fo = Fc - F (Fc = carrier frequency, 2F = bandwidth to be explored). c) Starting with b=0, increase the Notch Depth B; stop when the BER reaches a given threshold (usually 10-3). This is the Critical Notch Depth Bc for that BER value. d) The point [ Bc, fo] is a Signature point, to be plotted in the Notch Depth vs. Notch Frequency plane. e) Move the Notch Frequency fo of a given frequency step. Repeat steps c), and d) until fo = Fc + F (the band to be explored is completed). f) Repeat steps b) to e) with a negative delay (to get a non-minimum phase signature).
Angle Diversity
Some Space Diversity experiments have led to the conclusion that the diversity improvement in signal distortion seems to be independent of the antenna spacing. Even if this point is, in some way, controversial, it suggested the possibility of getting a diversity improvement also with Angle Diversity. Two implementations of Angle Diversity are considered : Two antennas (of the same type or of different types) side-byAntenna Diversity : side with slightly different pointing angles (used in preliminary experiments). One antenna with two feeders, producing beams with different Beam Diversity : shapes and/or pointing. In both cases, two beams operate at the receiver, closely spaced, but with different shapes. The multipath components are subject to different weighting at the two beams and the two composed Rx signals are in some measure uncorrelated. No need of high, complex tower structures; only one antenna with Beam Diversity; lower costs. Disadvantages : Less diversity improvement; less experience in Outage Prediction models. Advantages :
Diversity Receiver The Outage Probability due to Selective Fading only (PDS) is given by : PDS = PS2 / (1 - k2s) where : PS = Selective Fading Outage Probability with single receiver; ks = Selective Correlation Coefficient ; formulas for the computation of ks for space and frequency diversity are rather complex and are given in Rec. 530. Total Outage Probability The Selective Outage Probability (PS or PDS) and the Non-Selective (Thermal Noise) Outage Probability (PT or PDT) are combined to give the Total Outage Probability : POUT = PT + PS for Single Receiver POUT = [ PDT0.75 + PDS0.75 ]1.33 for Diversity Receiver
Diversity Improvement
The Diversity Outage Probability for Narrow-Band Signals (PDT, Non-Selective
Fading Model) is given by : PDT = PT / IFD where : PT = Non-Selective Outage Probability, Single Receiver = Po 10 -FM/10 ; Po = multipath fading occurrence factor ; FM = Fade margin ; IFD = Diversity Improvement factor.
Space Diversity : The ITU-R Rec. 530 gives the Improvement Factor ISD at Fade Depth A (dB) as : IFSD = [ 1 - exp ( -3.34 10-4 S0.87 F-0.12 L0.48 Po-1.04) ] 10(A-V)/10 where : S = Vertical Spacing (m) of Rx Antennas (centre-to-centre); F = Frequency (GHz) ; L = Hop Length (km) ; V = Absolute difference of the two antenna gains (dB).
With the same symbols, the Bell Labs (Vigants) formula gives (S < 15 m) : . IFSD = 1.2 10-3 S2 F L-1 10(A-V)/10
Diversity Improvement
1 + 1 Frequency Diversity : Both the ITU-R Rec. 530 and the Bell Labs (Barnett, Vigants) give the Improvement Factor IFFD at Fade Depth A (dB) as : IFFD = (80 / F L) (F / F) 10A/10 where : F = Frequency Spacing (GHz) .
The Diversity Outage Probability due to Selective Fading only (PDS) is given by : PDS = PS2 / (1 - k2s) where : PS = Selective Fading Outage Probability with single receiver; ks = Selective Correlation Coefficient ; formulas for the computation of ks with space and frequency diversity are rather complex (see Rec. 530).
Equalization Improvement
The reduction in Multipath Outage Time provided by equalization is estimated by comparing the Rx signatures with and without the equalizer. The Equalizer Improvement Factor (IFE) is defined as the ratio of the Selective Outage Probability PS without and with the equalizer : IFE = PS(unequalized) / PS(equalized) According to the ITU-R Rec. 530 model, the Selective Outage Probability PS is proportional to the Signature Area AS , defined as : AS = W 10-B/20 where : W = signature width (GHz) and B = signature depth (dB).
So, the Equalizer Improvement Factor (IFE) can be estimated as : IFE = AS(unequalized) / AS(equalized)
has been derived, giving the specific rain attenuation (dB/km), as a function of the rain rate R (mm/h). The factor k and the exponent are given in ITU-R Rec. 838, as functions of frequency and of wave polarization (horizontal or vertical). Formulas in the same Recommendation also include the cases of any linear or circular polarization and the effect of path elevation angle. The above discussion applies to rain attenuation within a rain cell, with a uniform rain intensity. The real situation is that of a radio path, whose length is usually (much) greater than the average cell size. In order to represent the effect of spatial variation in rain intensity, most prediction models make reference to an Effective Path Length LEFF, depending on the real path length and on assumptions on the rain distribution along the path. For very short paths, LEFF becomes equal to the real path length. The ITU-R Model for Rain Unavailability Prediction is based on these concepts.
A Table with k and values from 1 to 400 GHz is reported in ITU-R Rec. 838, which includes formulas for taking into account also path elevation angles and circular polarization. Effective Hop Length : LEFF = r L where L = actual hop length ; r = distance factor. r = 1 / (1 + L / {35 exp[-0.015 R)]} r = 1 / (1 + L / 7.81) if R < 100 mm/h if R >= 100 mm/h
Attenuation for any time percentage p (in the range 1% to 0.001%) : A (dB) = 0.12 A0.01 p -(0.546 + 0.043 Log10 p) The ITU-R prediction method is considered to be valid for frequencies up to 40 GHz and hop lengths up to 60 km.
The uncertainty in the Uo value makes not significant the difference between H and V pol. attenuation. In (1) CPA can be assumed as an average value. Other validity limitations : Frequency in the range 8 < F < 35 GHz; Small path elevation angle. Formula (1) (equiprobability relation) means that the probability of attenuation higher than CPA value used in (1) is roughly equal to the probability of Cross-Polar Discrimination lower than the XPD computed by (1).
Predicted Predicted
Objective Objective
Percentage of Time
7 GHz
Single Receiver
10-3
Rain Region P
Diversity Receiver
10-4 0 20 40 60 80 100
Predicted Predicted
Objective Objective
Percentage of Time
13 GHz
Rain Region: P N
10-3
K Single Receiver
10-4 0 6 12 18 24 30 36
Rain Unavailability:
Predicted
Objective
Percentage of Time
10-2
Rain Region: P
L 10 min/year
10-3
30 min/year
10-4 0 2 4 6
23 GHz 8
Status Post G.826 / G.827 (Objectives for Bit Rates at or above Primary Rate) Availability : New ITU-R Rec. under study. Error performance : Rec. 1092 (Digital Radio System in International Portion of a 27500 HRP) Rec. 1189 (Digital Radio System in National Portion of a 27500 HRP) Draft New Rec. (Real Digital Radio Links in International Portion of a 27500 HRP) Basic Objective for a 50 km hop in the International portion of HRDP : SES (Severely Errored Seconds) percentage not greater than 2 - 4 10-4 % ( 5 - 10 seconds / month, depending on overall objective allocation, block allowance, etc. ).
Definitions : Errored Second (ES) : a one-second period with one or more errored bits; Severely Errored Second (SES) : a one-second period with BER >= 1.10-3. Objectives : Number of ES / Seconds in total available time < 0.08. Number of SES / Seconds in total available time < 0.002. Objective Allocation : The total objectives are allocated to High, Medium, and Local Grade portions of the HRX.
G.826 : The Error Performance is based on Errored Block measurements at a Bit Rate at or above Primary Rate. Scope : 27,500 km IHypothetical Reference Path (HRP); PDH and SDH transport networks, Cell-based (ATM) connections. Definitions : Block : Set of consecutive bits, with size depending on bit rate. Errored Second (ES) : a one-second period with one or more errored blocks; Severely Errored Second (SES) : a one-second period which contains >= 30% errored blocks or at least one defect. Defect : Loss of signal, AIS, LOF for PDH; more complex for SDH and ATM. Background Block Error (BBE) : an errored block not as part of a SES.
of
Objectives : Number of ES / Seconds in total available time < (0.04 - 0.16) (depends on bit rate); Number of SES / Seconds in total available time < 0.002; Number of BBE / Number of blocks in available time (excluding SES) < 0.0004. Objective Allocation : The total objectives are allocated to the International and National portions (intermediate and terminating countries) of the HRP.
Many of the above bands are furtherly sub-divided among different classes of users, on a National or International basis. The ITU-R gives detailed channel arrangements for most of these bands. Additional indications are provided for particular conditions or exceptions relevant to some countries or group of countries. Therefore. the above list is to be intended just for general information.