Chapter 14 Linear Microwave Signal Transportation
14.1 Introduction
Although linear amplitude modulated fiber-optic links are the favored trunking methodology for most applications, there are situations where they are simply not cost-effective. For example, physical barriers, such as waterways, may make cable construction very expensive. Alternatively, the network may need to serve several small outlying communities that are so widely separated that the cost of the interconnecting links affects the economic viability of the project. Finally, network operators sometimes need to transport signals through areas where rights-of-way are either unavailable or very expensive to acquire. In any of these cases, point-to-point or point-to-multipoint broadband amplitude modulated microwave links (AMLs) may offer a superior solution.
This chapter will cover the basic operation of this equipment and the essential calculations and methodology required to engineer AMLs. No attempt is made to present a comprehensive manual on either internal designs of microwave equipment or the many subtleties of path engineering. Several standard reference books on the subject are available for those desiring to pursue the subject in greater depth.1-4
Microwave may be used for relay of individual channels or for transport of the entire FDM spectrum. The use of channelized microwave (which often employs digital or FM modulation) is covered in Chapter 7. The discussion here will be limited to links that transport the entire spectrum and are part of the broadband distribution network.
14.2 U. S. Regulation of Microwave Transmission
Regulation of radio transmission will vary by country, though broad spectrum planning is harmonized worldwide through international conferences. In the United States the essential regulations pertaining to AML systems are contained in the FCC’s rules.5 In particular, Part 17 covers towers, whereas Part 78 covers microwave licensing, channelization, and performance. Operators of microwave systems are required to have current copies of the rules on hand and to comply with detailed requirements regarding licensing, log books, posting of stations, periodic performance tests, and the like. Maintenance of tower lighting, where required, is a matter of critical concern because of aircraft safety issues.
The FCC has set aside the frequencies extending from 12.7 to 13.2 GHz for Cable antenna relay service (CARS). Some additional spectrum is assigned at 18 GHz and 31 GHz, as well. Within the 12-GHz band, several acceptable channelization schemes are given for both FM and AM modulation: A, B, and K channels for FM and C, D, E, and F designations for AML. Each channel in each scheme is given a unique identifier.
In the case of the AML bands, each channelization group represents a direct translation of the FDM cable spectrum to a different portion of the assigned microwave band as follows:
• Group C maps 54 through 300 MHz to 12,700.5 through 12,946.5 MHz.
• Group D maps 54 through 300 MHz to 12,759.7 through 13,005.7 MHz.
• Group E maps 54 through 300 MHz to 12,952.5 through 13,198.5 MHz.
• Group F maps 54 through 240 MHz to 13,012.5 through 13,198.5 MHz.
Transmission of a full 80-channel spectrum is possible through the use of a combination of Groups C, D, and E channels (with some minor shifts from nominal microwave frequencies) occupying the full 500-MHz CARS band. Clearly, without some form of frequency reuse, no combination of available 12-GHz channels is adequate for systems whose bandwidth exceeds 500 MHz.
In general, the FCC limits the microwave power level delivered to the antenna system to 5 watts per channel and the radiated power to +55 dBW EIRP.* Microwave frequencies must be accurate within ±0.005%, and the original cable channel frequencies must not shift by more than ±0.0005% through the link.
14.3 General Operational Principals
The basic principal of operation of an AML link is simplicity itself. At the transmitter, the incoming cable spectrum is simply mixed with an unmodulated signal (the microwave local oscillator, or LO) whose frequency is equal to the difference between the original channel frequencies and the translated microwave frequencies. At the receiver, the microwave spectrum is mixed with another unmodulated carrier at the same LO frequency, resulting in recovery of the original cable spectrum. In order to preserve the frequency accuracy of the VHF spectrum, as well as provide the required microwave frequency accuracy, a crystal reference oscillator operating at 1/171 of the offset frequency is used to phase lock the transmitter local oscillator directly. A sample of the reference frequency (which falls in the VHF spectrum between cable channels 4 and 5) is also up-converted and sent along with the remainder of the cable spectrum through the microwave link. At the receiver, the reference signal is recovered and used to phase lock the microwave LO there. Since both LOs are locked to a common reference, the recovered VHF spectrum is frequency coherent with the transmitted spectrum, while the accuracy of the microwave spectrum depends on the accuracy of the reference oscillator. Figure 14.1 is a basic functional diagram of an AML. Table 14.1 summarizes the frequencies for various bands, and Table 14.2 shows the combination of AML channels used to transport a 54 to 552-MHz cable spectrum.
Transmitter designs vary in the way channels are handled and in their power capabilities. Some early AML transmitters used a separate transmitter for each cable channel; most modern units are broadband. High-power transmitters use microwave amplification; low-power units may use the mixer output directly. Finally, early units used klystron tubes; most modern units use solid-state components, though the highest-powered units still use traveling wave tubes (TWTs), which provide high power with both gain and phase stability.
Regardless of the difference in transmitters, all receivers are similar, differing primarily in their mounting, powering, and flexibility in connecting to the downstream coaxial distribution system. In some cases, microwave low-noise amplifiers (LNAs) are used to improve the noise figure of receivers and/or allow the receiver to be mounted at the base of a tower rather than at the back of the antenna.
14.4 Path Design
Microwaves travel approximately along “line-of-sight” paths, meaning that, unlike low-frequency AM broadcast stations, they do not bend significantly to follow the earth’s curvature (though a slight bending in the direction of the earth is typical). Like light waves, however, they can be reflected from multiple intermediate surfaces between the transmitter and receiver. Under some atmospheric conditions, microwave signals can temporarily shift from their normal path, either toward or away from the earth’s surface. Such bending, discussed in Section 14.6, can cause the transmitted signal to completely miss the receiving antenna and is one of the factors limiting link reliability.
Designing a microwave path between two fixed points consists of determining the distance and direction between the points and then examining the path to make sure there is adequate clearance from any objects near the direct line between transmitting and receiving antennas.
14.4.1 Path End Coordinates
The locations of the path ends are described by four parameters: the longitude and latitude, which define the location on the earth’s surface, and the altitude of the base of the antenna tower and the mounting height of the antenna on that tower, which together define the vertical location of the antenna relative to sea level.
The longitude is usually expressed as the sum of the number of degrees, minutes, and seconds that the location is east (negative) or west (positive) of the prime meridian, which passes through Greenwich, England. The latitude is the sum of the number of degrees, minutes, and seconds that the location is north (positive) or south (negative) of the equator.
Since there are 60 minutes in a degree of arc and 60 seconds in a minute of arc, we can convert the longitude and latitude of the transmitter and receiver antenna locations to decimal equivalents as follows:
(14.1)For the remainder of this chapter, we will assume that this conversion has been done and that the transmitting location is at a west longitude of WT and a north latitude of NT, while the receiving antenna is at a west longitude of WR and a north latitude of NR, where all numbers are in decimal degrees. Minor modifications of the formulas will be required for links located in the southern hemisphere and/or eastern longitudes.
14.4.2 Path Length
Path calculations are based on an average earth circumference of 24,857 statute miles. Thus, each degree of arc along that surface represents 24,857/360 = 69.047 statute miles.
We can approximate the path length by calculating the length of the diagonal of a hypothetical rectangle whose corners are located at the transmit and receiver locations, as shown in Figure 14.2. In the north-south direction, the distance in miles is just the difference in the latitude values times 69.047, because latitude lines are all the same length. In the east-west direction, however, one degree represents that distance only at the equator. At any distance away from the equator, the distance is shortened by the cosine of the latitude. Therefore the width of the rectangle in miles is approximately the difference in longitudes times the cosine of the average of the latitudes times 69.047.
The path distance, D, is the length of the diagonal of this rectangle, calculated by taking the square root of the sum of the squares of the height and width of the rectangle, or*
(14.2)Though this gives close results for the path lengths likely to be involved in most AML applications, the precise formula for path length is
(14.3)If the path length is desired in nautical miles for some reason, then use 60 rather than 69.047 to convert degrees of arc to distance (nautical miles are based on an earth circumference of exactly 21,600 miles).
14.4.3 Path Azimuth
The next required value is the compass heading (azimuth) for the antennas, in order to align the path. Compass headings are measured in degrees clockwise from true north. The first step is to calculate the angle θusing
(14.4)If WR is less than WT, then the transmit antenna azimuth θT =θ. If WR is greater than WT, then θT = 360° -θ.
The azimuth for the antenna at the receive site is the opposite of that at the transmit site, that is, for θT between 0° and 180°, θR = θT + 180 degrees, whereas for θT between 180° and 360°, θR = θT − 180°.
14.4.4 Path Clearances
Having determined the horizontal distance and azimuth, the next step is to check for adequate clearances along the path. In the horizontal direction, this requires a calculation of a factor known as the Fresnel zone; in the vertical direction, it additionally requires consideration of the topology of the land, the height of such features as trees and buildings, and the effective curvature of the earth.
Fresnel Zone Calculations
The primary path from transmitting to receiving antenna is a nominally straight line. It is possible, however, for the signal to be reflected from objects that, though not in the direct path, are close enough that they are within the beamwidth of the antennas. If that happens, whatever portion of the transmitted signal is reflected toward the receiving antenna will arrive there after the direct signal because of the longer path length involved. The impact of this reflected wave will depend on the nature of the surface from which the signal is reflected, on the difference in path length, and on the pattern of the transmitting and receiving antennas.
The first Fresnel zone is defined as an imaginary ring surrounding the center of the direct path such that the distance from the transmitting antenna to this ring plus the distance from the ring to the receiving antenna is equal to one-half wavelength more than the direct path between the antennas. Subsequent Fresnel zones are defined as larger imaginary rings, where the difference in path lengths is nλ/2, where n is an integer larger than 1 and λis the free-space wavelength of the signal.
The radius of the first Fresnel zone at a point Dl miles from the transmitting antenna is
(14.5)where
F1 = the radius of the first Fresnel zone in feet
D = the total direct path length in miles
D1 = the distance from the transmitting antenna in miles
D2 = the distance from the receiving antenna in miles = D – D1
Figure 14.2 shows how the width of the first Fresnel zone varies along the transmission path.
Although higher-order Fresnel reflections can cause problems with highly reflective surfaces that are oriented optimally (such as smooth, flat terrain or calm water), the general rule of microwave design is that it is sufficient to clear objects by 0.6F1.
Additional Vertical Clearances
In the vertical plane, it is necessary to add the Fresnel zone clearance to the effective height of objects that lie under the direct path. Factors to be considered include the curvature of the earth, peaks in the terrain, and/or any objects, such as buildings and trees. Traditionally, such analyses were done graphically, although they are more likely to be done with a computer today. Figure 14.3 illustrates how the terms add up.
The flat line at the bottom of the Figure represents a straight line drawn through the earth from mean sea level at the transmit location to mean sea level at the receive location. At the transmit end, a vertical line represents the elevation of the ground at the base of the antenna tower, as surveyed or shown on a topological map. The units are feet above mean sea level (AMSL), we have designated this elevation as AMSLT to identify which site is referenced. The elevation of the center of the transmit antenna relative to the ground is above ground level, or AGLT. The sum of the ground height and antenna mounting height we have designated as HT. At the other end of the path, the equivalent values are AMSLR, AGLR, and HR.
The curved line that intersects the ends of the flat line represents the effective curvature of the earth between the endpoints. The effective curvature may not equal the actual curvature, however, because the microwave beam may be slightly bent due to atmospheric conditions. While the beam most commonly bends toward the earth (thereby allowing greater clearance over obstacles in the path), occasionally the reverse is true. These effects are taken into account using a factor called the K factor. K is the ratio between the effective and actual earth radius. The effect of earth curvature along a path of length D miles can be calculated using
(14.6)h = the virtual height, in feet, due to effective earth curvature
D1 = the distance to the transmitting antenna in statute miles
D2 = the distance to the receiving antenna in statute miles = D – D1
Under normal atmospheric conditions, K is approximately 4/3. Conservative path design, however, frequently calls for calculating clearances using K = 2/3. K for any specific location is determined from a factor known as the August mean radio refractivity and the path elevation above mean sea level.
In the special case of equal-height antennas transmitting over a smooth surface (so that the maximum interference point is midway along the path), Equation (14.6) reduces to
(14.7)where
The next step is to examine the vertical path profile. Generally this is done using topological maps, such as those available from the United States Geological Survey (USGS). Rather than plot the entire profile, it is sufficient to plot selected points that represent elevation maxima. For simplicity, we have shown a single peak, located at distance Dl from the transmit antenna, whose effective height is AMSLP plus the effective curvature of the earth at that distance along the path.
Added to this effective peak height must be anything that projects above the ground level at the peak. In wooded areas, for instance, it is common to add a number of feet that represents the maximum height of trees in that area; in a more urban setting, it might be a specific building. Finally, we must add 0.6F1, the required clearance to the first Fresnel zone, to get HP, the total effective height of the peak.
Using the height of the transmit antenna, the effective height of the peak, and its distance from the transmit antenna, we can calculate the minimum mounting height for the receive antenna using
(14.8)So long as the receive antenna mounting height above ground is greater than this value, the path should be acceptable from the standpoint of vertical clearances.
Note that in the special case of a flat landscape covering much of the distance between antennas, it may not be obvious where along the path the greatest potential interference occurs. A reasonable approximation requires first calculating the effective difference, δH, in elevation between the transmit antenna and the flat ground along the path:
(14.9)where
AMSLT = the elevation at the base of the transmit tower
AGLT = the mounting height on the tower
AMSLP = the elevation of the flat land between the antennas
Trees = the allowance for foliage and objects on the flat land
The only approximation in Equation (14.9) is the last term, which calculates the Fresnel clearance on the assumption that the highest point will be roughly in the center of the path. Obviously, to the degree that the total heights of the transmitting antenna and of the receiving antenna are different, this will be in error.
Given the effective difference in elevation between the transmit antenna and the flat area, the maximum distance that a signal can travel from the transmitter and still clear the ground is
(14.10)where
D1 = the distance from the transmitter, in miles, to the point where the signal just clears the ground by the required amount
D1 can now be plugged into Equation (14.8), along with the other factors, to determine the required mounting height of the receive antenna.
14.5 Performance Calculation
In order to predict the performance of an AML, we first need to define the details of the complete signal path from the transmitter to each receiver. Then we will start with the transmitter output RF power per channel and add up all the signal losses and gains (in dB) from the path elements to determine the nominal power level at each significant point. Knowing the performance of each circuit element, we will then calculate the cascaded noise and distortion of the link under normal weather conditions. Finally, using charts of expected weather disturbances, we will estimate the reliability of the circuit.
14.5.1 Signal Path Definition
The first step in performance calculation is to carefully list every circuit element in the path between transmitter and receiver. Since a single transmitter may feed more than one receiver, the first element may well be a signal splitter. Following that will usually be sections of waveguide. Since elliptical waveguide is more lossy than circular, it is not uncommon to use circular waveguide to feed up tall towers and then to transition to elliptical guide for the last few feet at both ends of the transmission line to allow flexibility for dish adjustments and equipment placement.
The dish will have a defined gain, followed by the path loss through the air, followed by the gain of the receiving antenna. Finally, there may be a low-noise amplifier at the output of the receive antenna, followed by some combination of waveguides down the tower and finally terminating at the receiver input. All the applicable circuit elements should be listed separately for each section of the path and a separate list made up for each path fed from a transmitter. The left two columns of Table 14.3 show a sample list of circuit elements for a typical AML path at 13 GHz.
14.5.2 Power Budget Calculation
In the first row of the fourth column of the worksheet is the transmitter power output per channel in dBm. Unlike coaxial cable distribution networks, where the reference for power measurements is 1 millivolt in a 75-ohm impedance (dBmV), the reference for microwave power is the milliwatt = 10−3 watts. Thus the power in dBm = 10 log (P), where P is the power in milliwatts. AML transmitters are available with rated power output levels per channel varying from about −12 dBm to +33 dBm.
Next come any splitters used to feed multiple receive sites. Manufacturer’s data sheets will give the loss of these. As a first approximation, the loss will typically be a few tenths of a decibel higher than the theoretical splitter ratio loss (e.g., a four-way splitter has a theoretical loss of 6 dB and an actual loss of about 6.5 dB to each port). Enter the appropriate loss in the third column of the worksheet as a negative number (representing “negative gain” in that path element).
After the splitter will be the waveguides connecting the splitter output port to the antenna. As a general rule, circular waveguide has a loss of about 0.014 dB/ft at 13 GHz, while elliptical waveguide has a loss of about 0.038 dB/ft at the same frequency. The losses of transmission lines used at other frequencies will be given in the manufacturer’s literature or in commonly available design tables. For each type of transmission line, multiply the line length by the loss per unit length and enter the net loss into the third column, again as a negative gain number.
Next enter the transmitting antenna gain, which will be positive relative to an isotropic radiator. Parabolic antenna design and performance are treated in detail in Chapter 8. The antennas used for CARS band are almost always constructed with prime focus feeds. Diameters commonly range from 4 to 10 feet. While actual antennas will vary slightly, a typical gain for a circular antenna with a parabolic cross section and a prime focus “button hook” feed is
(14.11)where
The loss suffered as a result of sending the signal through the air, the free-space loss, is dependent on both path length and operating frequency. Under normal atmospheric conditions, free-space loss is
(14.12)where
Next, it is common to add an additional loss allowance of about 2 dB to account for aging, imperfect alignment of equipment, and so on (sometimes known as the field factor).
The receive antenna will have a gain that will also be calculated in accordance with Equation (14.11). If an LNA is mounted at the back of the receive antenna feedhorn, as is common, enter its gain at this point. Now, as at the transmitter end, enter all the feedline losses between the antenna or LNA and receiver input.
In the fourth column, calculate the power at each point in the circuit, adding power gain values and subtracting losses in dB. The final number is the receiver input level, in dBm, under normal atmospheric conditions.
14.5.3 Carrier-to-Noise Calculation
As discussed in Chapter 11, the room-temperature thermal noise power in a 4-MHz bandwidth is about −59 dBmV. 0 dBmV, however, is approximately equal to −49 dBm, so the noise level referenced to 1 mW is −108 dBm.
Thus, the equivalent input noise floor of any device is equal to −108 dBm + FA, where FA is its noise Figure. Its C/N contribution (the difference between the driving signal level and the input noise floor) can be calculated using
(14.13)where
C/NDevice = the carrier-to-noise ratio contribution of the device
There are three principal contributors to link C/N: the transmitter, the LNA (if used), and the receiver. The transmitter C/N is generally given by the manufacturer for various channel-loading conditions. Since the input levels and noise Figures of both LNA and receiver are known, their C/N contributions are readily calculable. All three C/N values are entered into column five of the worksheet.
The link C/N is calculated using the methodology of Equation (11.2):
(14.14)where C/Nttl is the carrier-to-noise ratio of the entire link and C/NT, C/NL, and C/NR are the carrier-to-noise contributions of the transmitter, LNA, and receiver, respectively.
14.5.4 Distortion Calculation
The CTB and CSO distortion levels are given by the manufacturers for various channel loadings and power levels. In most cases, the LNA will contribute only slightly to the overall distortion levels.
Addition of second- and third-order distortion levels is discussed in Chapter 11. The link performance numbers in the worksheet were calculated using the most conservative assumption, which is
(14.15)where
A similar formula is used to calculate link CSO performance. Manufacturers should be consulted for the specific microwave equipment chosen to determine whether Equation (14.15) is appropriate for calculating cascaded distortion.
14.6 Link Availability Factors
14.6.1 Multipath
The preceding calculation predicts the performance of the microwave link under normal weather conditions. Two significant atmospheric effects can degrade the signal, however: multipath and rain fade. An understanding of the general causes and methods for calculating their effects on link performance is important.
It will be recalled that generally a microwave beam bends slightly toward the earth (with the result that the earth looks “flatter” than it actually is). The principal cause of this bending is that signals are not transmitted through a vacuum, but rather through the air. Since the density of air usually decreases with altitude, there will be a slight gradient across the width of the microwave beam, causing a slight decrease in velocity of propagation at the bottom of the beam. This causes the beam to bend slowly toward the earth.
Unfortunately, the density and change in density of the air is not a constant, with the result that sometimes the beam bends more and sometimes less. Under extreme conditions, the beam can miss the receiving antenna altogether. Under other conditions, there may also be more than one path between transmitter and receiver, with slightly different transit times and, therefore, phase and amplitude variations at the receiver. These effects are called multipath.
Since multipath degradation is a statistically random effect, the usual way of quantifying its effect is to define a minimum usable condition for the link and then to predict how much of the time the link performance exceeds this condition (defined as the availability of the path). For analog video links, a common definition of the minimum usable C/N is 35 dB.
The first step in calculating path availability is to determine how much the path loss can deteriorate before the C/N reaches 35 dB (the fade margin). Realizing that the C/N of both LNA and receiver will degrade linearly with increasing path loss while the transmitter’s contribution will stay constant, we can write a single equation for fade margin:
(14.16)where
C/Nmin = the defined minimum usable link C/N, usually 35 dB
In the example given, the fade margin is 18.35 dB. Since the transmitter typically contributes very little to the C/N at threshold, M can be approximated by simply calculating the difference between the link C/N under normal signal conditions and the defined minimum link C/N. In the example given, M determined by this method would be approximately 52.6 − 35 = 17.6 dB.
Based on extensive field testing, the number of hours the path will be below the threshold can be estimated using
(14.17)where
Um = the average probability that the excess path attenuation will be greater than the fade margin due to multipath effects
a = a “terrain factor,” varying from 4 for very smooth terrain or water to 1 for average terrain to 0.25 for very mountainous terrain
b = a “temperature/humidity factor,” varying from 0.5 for very humid areas to 0.25 for normal interior or northern climates to 0.125 for very dry desert or high-altitude areas
f = the operating frequency in GHz
D = the path length in statute miles
M = the fade margin in dB, calculated using Equation (14.16)
When the terrain and temperature/humidity factors are not known, it is common to use 0.25 for the product of a and b for estimation purposes.
14.6.2 Rain Fade
The second factor affecting path availability is rain. Water has a very high dielectric constant as compared with air (about 80 times as high); therefore, when heavy rain cells move through the microwave path, the transmission can be seriously affected. The important parameters are the intensity of the rain, the frequency of rain, the distribution of heavy rain peaks (that is, how big the cells of downpour are), and how fast the cells of intense rain move through the microwave path.
As with multipath, these factors do not lend themselves to theoretical analysis as much as to field experience. Extensive research into rain fade effects at different frequencies and in different geographic areas has been carried out by many researchers. One approach to estimating rain fade uses the methodology outlined next.6
First, the available fade margin per kilometer, M′, must be calculated by dividing the total fade margin M as determined in Equation (14.15) by the path length in kilometers. Since the path length, D, calculated in Equation (14.3), is in statute miles,
(14.18)Second, it is necessary to determine the rainfall rate that will result in an attenuation per kilometer greater than M′. The relationship between signal attenuation and rainfall rate can be approximated7 using
(14.19)where
YR = the attenuation along the signal path in dB/km
R = the rainfall rate in mm/hr
k and α = “constants” that are a function of frequency, polarization, and vertical path angle
k and α, in turn, are determined using the following formulas:
(14.20)
(14.21)where kh, kv, αh, and αv are determined from Table 14.4 and
θ= the path elevation angle in degrees (zero for a horizontal path)
T = the path polarization tilt angle in degrees: 0° for horizontal polarization, 90° for vertical polarization, 45° for circular
For relatively flat paths and horizontal polarization, k = kh and α = α h; for relatively flat paths with vertical polarization, k = kv and α = α v.
The maximum allowable rainfall rate can be calculated by solving Equation (14.19) for R and substituting M′ for YR:
(14.22)The probability that the rain rate exceeds this value in any given climatic region is given by8
(14.23)where u depends on the climate region and is given in Table 14.5 and
UR = the time-averaged probability that the rainfall rate exceeds Rmax
R0.01 = the rainfall rate, in mm/hr, that is exceeded in any given climate zone for 0.01% of the time (0.86 hr/yr), with an integration time of 1 min. As with u, it is given in Table 14.5.
Table 14.5 Rainfall Variable by Climatic Region
| Climate region | u | R0.01 |
|---|---|---|
| A | -0.2391 | 8 |
| B | -0.0050 | 12 |
| C | 0.05669 | 15 |
| D | 0.10252 | 19 |
| E | 0.03726 | 22 |
| F | 0.04652 | 28 |
| G | 0.07652 | 30 |
| H | 0.04081 | 32 |
| J | 0.1082 | 35 |
| K | 0.0444 | 42 |
| L | 0.03368 | 60 |
| M | 0.04087 | 63 |
| N | 0.02924 | 95 |
| P | 0.02477 | 145 |
Note: Figures 14.4, 14.5, and 14.6 delineate worldwide climate regions.9
(14.24)Once the probability of rainfall exceeding the value needed to attenuate the signal below the specified performance threshold is determined, the annual number of hours of rainfall-caused outage can be determined simply by multiplying UR by 8,760.
Finally, once the rainfall and multipath unavailability probabilities are determined, the total unavailability can be expressed as yearly hours of outage using
(14.25)Alternatively, the overall path availability can be calculated using
(14.26)
Typical link specifications vary from 0.999 availability for noncritical video applications to 0.99999 or even higher for some telephony applications. 0.9999 availability (equivalent to 53 minutes per year average outage time) is a typical goal for a cable television trunking application.
14.7 Summary
For most applications, fiber optics is the preferred choice for broadband signal transportation. Microwave links, however, are still a viable and cost-effective way to transport signals where construction costs or physical or regulatory barriers prevent cable construction.
Broadband amplitude modulated microwave links are simple in concept and straightforward to design. The performance of typical AMLs is comparable in noise and distortion to links constructed using directly modulated DFB laser transmitters but suffer from occasional outages due to atmospheric conditions. More important, though fiber-optic links have almost unlimited potential bandwidth, current regulations limit 12-GHz AML links to about 500 MHz of spectrum. The use of higher frequencies, though acceptable for short paths, is limited by the dramatic increase in rainfall and atmospheric attenuation.
The analysis of AML link performance involves a combination of trigonometry and a lot of empirical data on climatic factors, but it is amenable to straightforward computation using simple spreadsheet templates.
* See Chapter 7 for a discussion of the relationship among antenna gain, transmitter power, and effective isotropic radiated power (EIRP)
* Unless otherwise specified, the arguments for trigonometric functions are in degrees, not radians.
1. Engineering Considerations for Microwave Communications Systems, G.T.E. Lenkurt, San Carlos, CA, 1975.
2. Engineering Information, Rockwell International, Dallas TX, 1979.
3. The Handbook of Digital Communications, Microwave Systems News, Vol. 9, No. 11, 1979.
4. Transmission Applications Guide, Harris Farinon Division, San Carlos, CA, 1987.
5. C.F.R. 47$17 and C.F.R. 47$78, published by the Office of the Federal Register National Archives and Records Administration and available from the U. S. Government Printing Office.
6. The authors are indebted to Dr. Francisco Bernues of CableAML, Inc., for providing this approach to estimating rain outage probability. Private correspondence, September 1997.
7. From CCIR Report 721-3, p. 229.
8. From CCIR Report 563-3, p. 132.
9. Figures 14.4, 14.5, and 14.6 are extracted from International Telecommunications Union Recommendation ITU-R PN. 837-1, Characteristics of Precipitation for Propagation Modeling, Figures 1, 2, and 3, and are reprinted with permission of the ITU, which is the copyright holder. The sole responsibility for selecting these extracts lies with the author and can in no way be attributed to the ITU. The complete volume(s) of the ITU material from which the figures were extracted can be obtained from: International Telecommunications Union, General Secretariat — Sales and Marketing Service, Place des Nations, CH-1211 Geneva 20, Switzerland. They can be contacted on the Internet at: [email protected].