Radio Wave Propagation Lucien Boithias CNET. Translated by David Beeson Foreword by L-J Libois wh 206283 NORTH OXFORD ACADEMIC A division of Kogan Page Detence Acacemg,English translation © 1987 North Oxford Academic Publishers Ltd Original French language edition (Propagation des ondes radiotlectriques: Dans l'environnement terres) © BORDAS and C.N.E.T-E.N.S.T., Paris, 1984 Revised and updated 1987 All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without permission in writing from the publisher. English edition first published 1987 by North Oxford Academic Publishers Ltd, a subsidiary of Kogan Page Ltd, 120 Pentonville Road, London N1 9JN British Library Cataloguing in Publication Data Boithias, Lucien Radio wave propagation, 1. Radio wave propagation I. Title 621.3841'1— TK6SS3 ISBN 0-946536-06-6 ‘Typeset by V & M Graphics Ltd, Aylesbury, Bucks. Printed and bound in Great Britain by Mackays of Chatham Ltd. Produce cesar tote cess sees eee eee teeeeret eee Chapter 1 Wave propagation in free space 1.1. Propagation and Waves 1.2. Ducted propagation and radiation propagation in free space we eieeteereiaai 1.3. Choice of reference antenna: summary of antenna characteristics oo ee ee eee 1.4, Electromagnetic fields and attenuation of propagation SPACE tee ase 1.5. Relations between different quantities characterizing a plane rave sete sees ee eee ae eect ieee 1.6. Propagation attenuation over a link that includes a passi 1.7. Propagation at short distance from the transmitting antennas Chapter 2 Basic propagation phenomena: summary of physical omepis ates Tees eseasaeeanie urease 24, Introduction 6... eee ee eee ce 22. The line of sight propagation model © 6... ss + + 23. Definition and criterion for a line of sight ink... . 24, Reflection and refraction ©... eee eee ee 25, Diffraction... 6 ee ee ee ee eee 2.6 Semie secret eee ie ieteesere ieee ite 2.7. Specular reflection and diffuse reflection: Rayleigh’s criterion 2.8. Wave ducted by a conducting surface: Zenneck waves... ‘Chapter 3 The effects of the Earth’ surface on propagation: line of sight poise eeceaeeesese see eee ea eee "3.1, Electric characteristics of the Earth’s surface... . 2. 32 Surface reflectivity... eee ee B 19 49 52 52 233. Propagation near a smooth spherical surface... . 56 34, Resultantfield 2... eee Bees eee OH 35. Apparent radiation pattern of an antenna taking the Earth’s surface into account inks with aircraft or spacecraft target detection usingradar wo ee ee see eee tag 3.6. The influence of the size and position of a reflecting area of the Banth'ssurface eee eee eee eB 3.7. Effect of irregularities in the Earth's surface... ere 3.8. Wavelength and depth of penetration into the Earth's surface 78 Chapter 4 Influence of the troposphere: refraction effects... 79 4.1. Refractive index of the troposphere... se eee 9, 42. Wave refraction in the troposphere... ee es 86 43, Transpolarization (cross polarization) in clear air 110 Appendix 4.1. Ray paths for a duct close to the Earth's surface Mm ‘Appendix 4.2, Conditions for the existence of duct modes in a tropospheric au eerie Sea 4 Chapter $Tnflence ofthe troposphere: stenuatin doe gash and hhydrometeors . . . - seas ine eeeeeretege ee 116 SL, Introduction... ee eerie 5.2. Attenuation by atmospheric gases 5. ese ee (UB 5.3. Attenuation due to hydrometeors (rain, fog, clouds, snow, hal) eee eerie eesesnanerses 123 5.4, Transpolarization due to hydrometeors 5. + ss + eeeeigT 55. Scatter by hydrometeors . 6s. ss Geese 140 Chapter 6 Problems due to propagation online of sight links 144 6, Introduction 6... ee aeeeeeaead ay ra 62. Major phenomena oo. ee eee ee 144 6.3. Fading due to large scale variation in atmospheric refraction 145 64, Fades and distortion due to tropospheric multi-paths. ... 149 65. Fading due to reflection from the Earths surface eed Se 66. Attenuation due to hydrometeors . «sss 165 ‘Appendix 6.1. Diversity reception... . m1 ‘Chapter 7 Propagation by diffraction... 2.2... es TA. Introduction... oe ee ee ee eee 72 Spherical diffraction ©... .- 000% Tae Bdge ciltraction eineretsiatsteee es sasatice Appendix 7.1, Spherical diffraction : variants Chapter 8 Propagation by scatter 8.1, Tropospheric scatter (or ‘troposcatter’) . 6... ee 82. Ionospheriescatter oo ee ee eee eee Chapter 9 Influence of the ionosphere on propagation =... . 9.1. Structure and general characteristics ofthe ionosphere... 9.2. Wave propagation in the ionosphere... . . 9.3, Links penetrating the ionosphere... ee ee ee 9.4, Links by means of ionospheric reflection... .... ‘Chapter 10 Frequency sharing and frequency re-use, coverage area, Interference 2... ee eee 10.1 Introduction 10.2 Factors to take into account when sharing or reusing frequencies ov vee ee eee eee 10.3. Classification and characteristics of radio communications 10.4 Point to point communication: imerference . .. . . «+ 10.5. Point to zone communication: space service area, interference 10.6. Interference due to ionospheric phenomena»... . . Chapter 11 Noise and interference 6.2. es ALL, Wanted and unwanted signals 6. ee eee a 11.2. Significance of the terms disturbanes, interference and noise 113, Transmission quality 6.6 es eee eae 11.4, Theoretical remarks concerning thermal noise in one-port and tworport networks oe ee ILS, External noisesourees oe ee Appendix 1 Statistical distributions used in radio wave propagation ies estate 176 176 7 193 201 216 218 218 2 235 256‘Appendix 2 Summary of the effects of propagation on the use of different frequency ranges... se ev ee se 9 Appendix 3 Main uses of radio waves for telecommunication 312 Bibliography... 22. +s cence eee aes eeeeeieealy Tndeeetveseeemnneeeeieiy eee sieses teat ad5| Foreword In 1887 Heinrich Hertz, professor of physics at the University of Karlsruhe, wrote to his famous compatriot Helmholtz, who was working on light and electricity: «.., Thave just taken a new step forward which, in my view, establishes an incontrovertible link between light and electricity. I cannot wait to discuss it with you. In the frst place I learnt by a stroke of luck that waves several metres long are not the only ones we can produce, but that we can use far shorter waves which are vastly easier to handle. My tests worked out very wel.” ‘This letter may be regarded as the birt of ‘Hertzian waves’ or radio waves, It is interesting to note that Hertz, who was engaged in a series of experiments using parabolic mirrors, was already thinking in terms of using very short waves; it would however take more than half a century for microwave techniques to be applied to ‘communications using radio links. ‘The work carried out by Hertz provided spectacular confirmation of the theory proposed in 1864 by the brilliant mathematician James Clerk Maxwell that using purely theoretical arguments light and electromagnetic waves were identical phenomena. This famous theory, summarized in four remarkably simple equations called Maxwell's equations, was regarded at the time Maxwel first propounded it sno les revolutionary than Einstein's theory of relativity would be 40 years later. “However, if Hertz could lay claim to being the father of radio waves called for that reason ‘Hertzian waves’ in some countries, it never occurred to him to use them as a means of communication. At that time it was rare for pure scientists to turn their attention to applied science. “This discovery’, wrote Maurice Ponte, ‘would be applied by others. Scientists who were also to some extent engineers, engineers who were masters of science, ‘would some years later create wireless telegraphy.” ‘When mention is made of wireless telegraphy as an invention we think in particular of Edouard Branly in France and Alexandre Stepanovitch Popov in Russia. But it is to Guglielmo Marconi that credit is due for taking wireless telegraphy out ofthe laboratory at the end of the last century. Marconi was avery active technician of immense talent. The experiments he carried out in 1896 were incontestably the first in the area of radio communication. Maxwell, a theoretician of genius, Hertz, a talented experimentalist, and Marconi, a man of firm action, were the founders of radiocommunication. Since the beginning of the twentieth century, progress in our knowledge of phenomena associated with propagation and the use of radio waves has been extraordinary. In December 1901 the first radiotelegraphic link was established between Great Britain and Newfoundland by Marconi. In 1902 Oliver Heaviside,10 FOREWORD in England, and Kennelly, in the United States, proposed that waves could travel round the Earth by reflecting off the ionized layers of the upper atmosphere; this hypothesis, taken up and defended in France by André Blondel and Henri Poincaré, would be confirmed some years later. The discovery of the ‘ionosphere’ formed the foundation for the development of intercontinental telecommunica- tions for half a century. The Second World War, with its marked effect on technical progress, was responsible for new radiocommunication methods: radio relay links in the first place, using techniques developed directly from radar, and then, more recently, satellite tlecommunications. Theoretical and experimental studies of propagation phenomena have been taking on increasing importance forthe construction of telecommunication systems using radio relay links. Teleeommunication engineers have also contributed to the description ofthese phenomena and their applications. Particular mention of the work carried out in France by Pierre David, whose study on wave propagation published in 1955 forms the basis for training radio and telecommunications engineers in this field, must be made. A new edition of this book, updated with the help of Jean Voge, appeared in 1966.’ No other general work on propagation has been published in France since that date. ‘The greatest French contribution to this field since the Second World War has, however, been that made by research workers at the CNET and, in particular, by Lucien Boithias, a scientist of national and international repute. Lutcien Boithias, whose remarkable talents as a researcher I came to value highly at the CNET, began his work on propagation in the troposphere immediately after the Second World War. At that time, radiocommunication and broadcasting used almost exclusively radio waves of frequency less than 30MEz. As wwe have already pointed out, the ionosphere was already playing a crucial part in intercontinental communications, and was being studied by many theoretical and experimental scientists. On the other hand, systematic study of propagation in the troposphere, although it had begun in 1945 in the United States and in the early 1950s in France, had advanced far less. For this reason, Lucien Boithias was asked to et up the ‘troposphere department” at the CNET to meet predicted trends in the development of line of sight radio relay networks. Then, in the mid 1950s the discovery of ‘tropospheric scatter’ propagation phenomenon led to renewed interest in non-line of sight communications. This is a field in the study of propagation to which the group of engineers at the CNET, including most notably Jean Voge, Lucien Boithias and Frangois du Castel, made a major original contribution? However, propagation by reflection from the ionosphere and studies asso- ated with it were still crucial for intercontinental communications. It was only in "David, P; Voge, J Propagation des ondes Editions Eyrolles, Paris, 1966 du Castel, F Propagation troposphérique et faisceaux hertziens trans-horizon Collection ‘Technique et Scientifique du CNET, Editions Chiron, Paris, 1961. FOREWORD 1962, when the frst satelite telecommunications link between Andover and Pleumeur Bodou was established on 11 July 1962, that a new era began for long ‘ange telecommunications. New fields of research appeared for those working on ‘propagation of radio waves. In particular, the ionosphere was no longer seen simply as a medium which reflects waves but also as one through which they pass. Moreover, the available frequencies are becoming saturated, and previously undetected forms of interference are beginning to appear. Consequently, telecom- munications engineers are having to recommend the use of increasingly high frequencies (greater than 10GHz), both for terrestrial radio links and for satellite communications. The use of such frequencies, which may at first appear attractive, poses special problems, such as absorption by the atmosphere, particularly in ‘rainfall, Further research is needed to analyse these phenomena and obtain aclearer understanding of them. ‘We might have thought, a few years ago, that we knew everything there was to know about propagation. Now we realize that nothing could be further from the truth, The propagation of waves, and the phenomena associated with it, such as distortion and noise, form a research field whose value forthe development of radio telecommunication links is obvious, and for which a state of the art review is extremely useful. It is such a review that Lucien Boithias provides in this book. ‘As the author points out in the Preface the book is aimed mainly at users. So, for example, only theories leading to practical results of immediate application are mentioned. On the other hand, this book contains a considerable quantity of information for technicians; generally, the data presented have been obtained by semi-empirical methods based on statistical analysis in particular. {As far as the layout of the book is concerned, Lucien Boithias deliberately separates line of sight from non-line of sight propagation. This a very sensible separation from the point of view of the text's educational nature. This book is therefore divided into 11 chapters. In the Preface he gives a briet history of the field and discusses the different aspects of propagation studies. ‘The first six chapters deal with line of sight propagation, from the traditional starting point of propagation in open space, and the fundamental physical phenomena involved. Note, in particular, the value to users of Chapter 6, devoted to a study of distortion (fades, multi-paths, etc.) which can reduce transmission quality. The author describes prediction and prevention methods which can be applied to alleviate this problem. Having dealt with line of sight propagation, Lucien Boithias describes in Chapters 7, 8 and 9 the three major mechanisms affecting non-line of sight propagation: knife-edge diffraction, tropospheric scatter end reflection from the ionosphere. Chapters 10 and 11 examine phenomena associated with noise, particularly over great distances, and to distortion of all kinds affecting radi ‘communication. Appendix | summarizes the most important probabilistic dstribu- tions for the study of propagation. Lucien Boithias therefore provides study which is both structured and complete, whereas atthe same time being easily accessible. Every chapter contains2 FOREWORD. useful references, as wel as the important graphs and tables needed by the user. The ‘work is undoubtedly an essential reference text for those who are interested in ‘understanding and applying propagation phenomena. Let me finally say that Lucien Boithias’ research has often been used as the basis for official documents ‘drawn up by intemational organizations, in particular the Comité Consultatif International des Radiocommunications (CCIR), which isa leader in this field. ‘The CNET scientific and technical communications series already comprises ‘many essential titles for telecommunications researchers, engineers and technicians. In my view a basic work on propagation, giving a complete, clear, accessible overview of the field was missing. Lucien Boithias's book fills that gap. T take considerable pleasure, as aformer director of the CNET, in congratulating not only the author, but aso the publishers for taking the initiative in making this valuable text available. L-JLIBOIS Conseiller Maitre & la Cour des Comptes Directeur Général Honoraire des Télécommunications October 1986 Preface ‘When I began to take an interest in wave propagation thirty years ago France had not yet established so much as a kilometre of a centimetric radio relay wave network. ‘Radio’ essentially meant broadcasting on long and medium waves or- called short waves (HF) for intercontinental communications. In other words, in most cases frequencies below 30MHz were used. ‘Beyond this frequency, the electromagnetic spectrum was vast and virtually unused. Metric waves (VHF) were beginnining to be used for certain links with few channels, and in France used by the 441 line television transmitter on the Eiffel ‘Tower, operating on 46MHz for picture transmission and 42MHz for sound transmission, Work was beginning on the installation of television transmitters in the 162-216MHz range, and the fist in Paris and Lille had been established on the same frequency, as it was obvious for many people that at this frequency propagation took place along a straight line ‘as for light waves’, and consequently there could be no interference between the two transmitters. This pioneering period, ‘when in order to receive a television picture you generally had to build your own receiver, gave me the opportunity to win a trophy by receiving TV transmissions from Paris 380km away. Reception obviously had certain acrobatic qualities, and not only from the electromagnetic point of view, since the antenna wasset up on top of a church spire. Furthermore, the image received on the oscilloscope tube had a green tinge to it, which was somewhat surprising for black and white picture transmission, Ihave to admit that propagation phenomena were generally kind to me then. ‘Asa result of research into radar during the war, centimetric waves (SHF) were beginning to be better known in certain frequency bands. These bands were ‘generally referred to by code names atthe time (S band, X band), which still survive. ‘The use of radar, particularly over sea, highlighted ducted propagation phe mena, The first known example was the detection of the Arabian coast by radar in Bombay 1500km away. ‘The start of the 1950s saw the introduction of the first microwave radio link in France, and the first studies into direct lin of sight propagation carried out by the CNET, both in the North of France and on the Mediterranean coast. Propagation phenomena were stil new concepts for many people, to the extent that a notable figure, to whom it was being explained that a radio link could not be kept free above 1a forest because the tres were blocking the frst Fresnel ellipsoid asked ifthe tees ‘could not be pruned to the shape of a Fresnel ellipsoid. In 1986 the Comité Consultatif International des Radiocommunications (CIR) defined standards of quality for microwave radio relay systems, and these standards were much stricter than those that had been considered sufficient for transmission using dekametric wavelengths (HF). Time percentages were no longer4 PREFACE between 50 and 90%, but between 99.9 and 99.99%, while signal: noise ratio requirements shot upwards by several tens of decibels, ‘Consequently, propagation problems took on a new importance and the necessary levels of predicting accuracy grew considerably. A major decision was taken in 1955. On the one hand, asthe French telecommunications agency did not intend any immediate further development of the radio relay network that had already been established, studies into line of sight propagation became less urgent. On the other hand, it had recently been discovered that it was possible to obtain propagation over several hundred kilometres by tropospheric scatter. This discovery immediately stimulated considerable interest both among theoreticians and telecommunication scientists. With hindsight, we can se that theoretical work led to no usable results, and that only simple radiometeorological considerations, based on comparing the results of measurements in various climates, made it possible to find the characteristics of this type of propagation and set up reliable prediction methods. On this subject, it is worth noting the importance of measurements on propagation made by the CNET in France and Africa. In particular, the plans for radio broadcasts to Africa on metric or dekametric waves drawn up by the UIT in 1963 were entirely based on the propagation measurements by the CNET, since no ‘other results were available than those generated by my group at that time and we had also worked for many years in some ten countries in Aftica, under less than laboratory conditions. ‘At around 1970, there was a sudden return to the earlier preoccupation with line of sight communications, but radio relay equipment had in the meantime changed profoundly. On the one hand transistors were replacing traditional valves, while on the other digital modulation was beginning to replace analog modulation ‘There followed a rapid growth of radio relay networks throughout the world, leading to a saturation of frequency bands lower than 10GHz, and to renewed interest in studies into propagation at higher frequencies. At such frequencies rain begins to play an important role. Propagation scientists suddenly began to take an interest in rainfall levels, inthe shape of water droplets, in their speed of descent, etc. ital modulation also had repercussions for the study of propagation because of the bandwidth necessary. On the one hand, the existence of mult propagation paths could lead to unacceptable distortion in propagation delays and to capacity limitations, while on the other it was necessary in order to save on the electromagnetic spectrum used, to transmit two orthogonally polarized signals. As propagation properties were likely to limit the possibility to do ths, it was necessary to carry out new studies. ‘At about the same time, the arrival of telecommunication satelites led to new propagation problems, and a distinction could be drawn between terrestrial links, ‘corresponding to horizontal paths and space links corresponding to slant paths for which the propagation trajectory penetrates the troposphere and ionosphere. Certain propagation anomalies were reduced but others appeared, and in any case more precise knowledge of the phenomena involved was required than for PREFACE 15 terrestrial links, as in general space equipment allows only a very small fading margin, Other satellites, such as those used for geodesic measurements, make it ‘necessary to have precise knowledge of the interference due to the neutral or ionized atmosphere. ‘As for the use ofthe ionosphere for telecommunications, the introduction of satellites led to fundamental changes, as radiowaves in the dekametric (HF) band disappeared almost completely from the intercontinental communications where they had been, for some decades, the only technical means available. They are stil often used, above all for unimportant links over short distances in developing countries, and also for military purposes. The consequence is that in books dealing with radio wave propagation for telecommunications, sections describing reflection from the ionosphere, previously a major aspect of the field, can now be reduced considerably, while it has become necessary, on the other hand, to deal with the ‘question of propagation through the ionosphere. ‘Another consequence of the introduction of satellites is the increased importance of problems associated with interference at long distances, even on super high frequencies. Consequently, propagation phenomena likely to create intense fields at long distances had to be studied in considerable detail. Itis therefore clear that propagation studies have by no means reached a fixed format and, although the basic phenomena are sil the same (tis very unlikely that Maxwell's equations will ever be revised), applications have developed considera~ bly, with certain fields of study losing in importance while many others have appeared or become far more significant. ‘Work on propagation has a number of aspects that have to be considered at the outset since the form a discussion of propagation takes will depend on the relative importance that the author attaches to each of these aspects. ‘One point of view, often qualified as scientific, involves the search for relations between various phenomena. This is an aspect that clearly must not be neglected, ‘even though it can create the illusion that a phenomenon has been ‘explained’ when in fact only its similarity with a known phenomenon has been shown. Itis often also the case that the extent of scientific discussion devoted to a problem depends more cn the formal interest of the computations it makes possible, rather than their practical importance. Another danger to be avoided in theoretical studies is the introduction of parameters which are virtually impossible to measure. Clearly, under such conditions theory cannot be contradicted by experiment, but its value is very limited. The history of propagation studies provides examples of such situations. ‘A more utilitarian viewpoint is also possible, aiming to give practical rules to those who are designing radio telecommunication systems. As the transmission quality of such systems is defined for the different time percentages specified by international organizations such as the CIR statistical analysis of long duration ‘measurements is obviously crucial for drawing up such practical rules. Statistical studies of this kind are therefore extremely useful, and also constitute a powerful analytical tool, of great use for generating theory. It is also known that statistical16 PREFACE studies raise a number of problems, in particular those that concern the quantities to be recorded and the degree of computing to be applied to data. In this computer ‘age it seems obvious that most procedures ought to be done automatically. Infact, however, automated processing requires a level of data collection that is almost impossible to maintain continuously over several years’ study, and for which significant probability levels are sometimes of the order of 107 to 10-. Manual or semi-automated processing makes it possible to eliminate most recording errors. ‘This compromise is regularly made. Inthis book, whichis principally intended for users, we do not study theory for its own sake and only mention those theories which lead to practical results. Instead we provide large amounts of directly usable information, often obtained by semi- empirical methods. should like to draw a parallel between propagation studies and medicine. In medicine we are concerned with discovering the laws of nature and protecting ‘ourselves agninst certain harmful consequences of those laws by the appro ‘treatment, The medical research worker or doctor is often led to carrying out long expensive analysis, although, to a certain degree at least, such analysis can be avoided or at least reduced by the expertise already available in the field. This is indeed also true in propagation studies and the propagation scientist is consulted about an observed anomaly, and asked if possible, to provide a remedy to the situation. Unfortunately, a suitable treatment is not always available. “The analogy could be taken further. There are, for example, spontaneous cures for propagation deficiencies: we just have to wait a sufficiently long time. This isa ‘general law, since propagation defects have the considerable advantage over hardware defects of being able to ‘cure themselves’. However, such defects can sometimes be chronic (eg. they recur every summer) and this may make servicing ‘necessary, generally in the form ofa replacement or modification to an installation, tg. by alding diversity reception faclitice. Thi ‘medical’ aspect of propagation has important consequences. As in medicine, we only attain efficiency and value after many years of work. ‘independence’, which is viewed favourably in some fields, is hharmful here. Iis true that independence has certain advantages, for example, itis just as well not to be around when it becomes clear that a communication link will ‘not work because the predictions of the propagation scientist were wrong, but i nontheless true that even independent workers prefer to consult specialists sometimes. "The way the subjects are distributed throughout this book isa consequence, to ‘a considerable extent, of deliberate decision to separate line of sight propagation from non-line of sight propagation. Many years’ teaching has shown me the advantage of such a distinction. The first six chapters therefore deal with line of sight propagation. Chapter I deals with propagation in open space and chapter 2 with the characteristics ofthe major physical phenomena dealt with inthe rest of the book (reflection, refraction, diffraction, scattering). Chapters 3,4and 5 examine the effect of the ground and the troposphere. Chapter 6 brings together a number of "useful data concerning anomalies met in line of sight links, which are often due to PREFACE, 7 the combined effect of the ground and the atmosphere. Chapters 7, 8 and 9 are devoted to the three main mechanisms allowing links to be established between two points which are not visible to each other, these ‘mechanisms being diffraction by hills, scatter by the Earth's curvature and the ‘troposphere and reflection by the ionosphere, However, in order not to separate the ‘two sections describing the ionosphere, Chapter 9 also deals with links penetrating the ionosphere ‘Chapter 10 gives certain results on propagation mechanisms particularly involved in long distance interference. Not all characteristics of such mechanisms are particularly well known and the number of publications dealing with them isnot always proportional to their actual practical importance. ‘Chapter 11 concerns the problem of electromagnetic disturbance, and mainly interference caused by other transmissions or various forms of radiation. Itisin fact traditional to include a chapter of this kind in a discussion of propagation, since propagation plays an important role in determining the level of disturbance, and also the signal: interference ratio atthe receiver. Appendix 1 gives information concerning the main statistical distributions used in propagation studies. This will perhaps avoid readers having to consult specialist works, which as well as being voluminous are often too theoretical for the intended application. ‘Appendixes 2 and 3 gives both administrative and technical information from Radio Regulations on the various services using electromagnetic waves and for ‘whom, consequently, knowledge of propagation phenomena is L. Thave attached particular importance to the presentation of figures. Most of them, and in particular all the aligned point nomograms have been entirely calculated o verified. Many of the figures inthis work, which also appear in works published by the CIR, were drawn up by the CNET propagation team and supplied to the CCIR. We thank the CCTR for permission to reprociuce these figures. Choosing the bibliography is always difficult. In the present case, the bibliography is relatively concise as there are few recent books in this field, and articles often handle subjects which are too specialized or give results which are likely to be rapidly out of date. A number of titles are given for each chapter and, moreover, I mention Volumes V and VI of the CIR, since these volumes, ‘concerned with propagation, are revised every four years and contain many references. Before concluding this preface I should like to thank my colleagues who inspired some ofthe ideas inthis book, my typist who prepared the manuscript, and the referees who criticized the content. ‘During the time of writing this book I have been associated with many people with the aim of studying propagation for telecommunication purposes. With this in mind, as I explained above, the main tool used was statistical analysis of long time measurements. This is detailed difficult work requiring constant attention, which is difficult to leave to entirely automated devices given the reliability levels of18 PREFACE hardware and to very low time percentages required in statistics. [cannot overstate ‘my gratitude to the people who have generated reliable data and carried out the statistical analysis of those data. I should like to make particular mention of my colleague, J Battesti, who has asa result of our collaboration for so many years and publication of so many articles played a certain role in the preparation of this book. ‘And above all, my thanks are due to him for two extremely precious qualities, total intellectual honesty and considerable patience. ‘My thanks are due to those readers who sent me their comments and criticisms, and [hope that this translation of the new French edition, which includes new formulas and ammendments to bring the text in line with new developments and recent prediction methods presented to the CCIR, will be as wel received as the original French edition. ‘Telecommunication Chief CNET Issy les Moulineaux France October 1986 Chapter 1 Wave propagation in free space 1.1. Propagation and waves ‘The term wave propagation can be divided into two components, propagation ‘and wave, which deserve close analysis. Propagation can be defined in general terms asa transfer of energy without a transfer of matter. The transfer of energy between two points in a certain medium is therefore the result of changes, with time, in the spatial distribution ofa non-statc field in that medium. This concept, which is valid at the macroscopic level, applies to a number of physical phenomena in the electromagnetic, heat, acoustic and other fields. ‘The term wave can only be defined by its relationship with the terms field and propagation. Wave propagation isa particular form of propagation, characteristic of certain fields and such that a given element of the field (c.g. a given phase value, ‘an extremum, ete.) which is at a particular point at a certain moment will at a subsequent moment have undergone a displacement with little or none deforma- tion. This property makes it possible to define a propagation velocity. Its value depends on the field parameter chosen (we can, for example, obtain the phase velocity, the group velocity, etc.) and on the characteristics of the medium. If the variations with time of the quantity which characterizes the fields are periodical, the distance travelled in any one period is known as the wavelength. From the mathematical point of view, this set of properties can be expressed by the fact that the linear second order partial differential equation which represents ‘such afield is @ hyperbolic equation. For such a type of equation there exists two groups of real surfaces on which at least two solutions ofthe equation join up. One of these solutions may be the null solution (absence of the phenomenon). Such a surface is a wave front. Since this surface depends on time, the wave front moves with a definite velocity. ‘The simple harmonic equation, the telegraphist equation and Maxwell's equations are all hyperbolic, and the associated field shows wave propagation. On the other hand, heat conduction along a metal rod obeys a parabolic law; there is ‘no wave front or propagation velocity associated with this phenomenon, and therefore wave propagation isnot involved. ‘Therefore itcan be said thatthe term wave is no more than a label used for the ‘category of fields for which a propagation velocity may be defined. In fact, the ‘equation of afield and that of the corresponding wave cannot be distinguished. ‘As well as these terms the term radiation is introduced. Itis defined according2 RADIO WAVE PROPAGATION to the Radio Regulations as the energy flux produced in the form of waves by a source, or the energy itself. Studying radiation therefore means studying its sources, ‘whereas the study of a wave can be carried out independently of ts source, and does not assume that the wave can actually be produced by a real source (plane waves, cylindrical waves, etc.). The relationship between sources and waves is nothing ‘more than a study of antennas. This is not the purpose of this book. 1.2. Ducted propagation and radiation propagation in free space Approaching the question from the point of view of practical applications there are two major types of electromagnetic waves propagation: 1, propagation ducted along a transmission line; 2. propagation of waves radiated from an antenna. ‘Under the Radio Regulations that part of the electromagnetic wave spectrum below 3000GHz (wavelengths greater than 0.lmm) concerns radio waves. This definition is merely conventional, since the prefix radio (from the Latin to radiate, radior radiari) does not imply any restricted frequency range. Ducting requires material structures and, consequently, it involves attenuation. If we assume that all the elements exhibit linear characteristics and that the line is homogeneous, the percentage energy lost over a section of given length is independent of the energy entering that section. This implies that power along the line decreases exponentially with distance. ‘The same is not true for radiation propagation, since it can exist without a ‘material component. Consider, therefore, the ideal case of an empty unlimited space containing a transmitting antenna and its associated transmitter only. This is termed open space. Assume too that the dimensions ofthese elements are negligible with respect to the transmission distances studied. The energy emitted at any given time will, at any time later, be distributed over the surface of a sphere of radius d ‘The power flux density decreases with respect to 1/d?. From a certain point onwards this relation gives a far smaller reduction than does the exponential function for ducted propagation. For example, no transmission line could allow ‘communication to be set up from the Earth to the Moon, which only requires a relatively low power for radio communication. ‘This difference in behaviour is derived from the fact that, in ducted ‘propagation the decrease in the received level is due to the progressive transforma- tion of electromagnetic energy into heat or another form of energy, whereas in radiation propagation the decrease is the result of the energy being distributed over an increasingly large surface. The 1/d” relation isnot defined atthe origin, since for radiated waves, the transmitting antenna corresponds to a singular point in the: space. Propagation in free space plays a fundamental role in providing a frame of | WAVE PROPAGATION IN FREE SPACE a reference against which to judge all other cases. In particular, the actual received energy level is compared with the level which would have been received in free space. 1.3. Choice of reference antenna: summary of antenna characteristics ‘The propagation of a wave between two points can only be studied and measured using an antenna at each of these two points. However, the properties of, ‘the antennas used must not influence the results. This means that propagation must bbe defined as occurring between two reference antennas which are also used to define the characteristics ofthe antennas actually used, A clear distinction has to be made between reference antennas used for these calculations, which it is not necessary therefore to build, and comparison antennas used for measurements on antennas, which must be constructed and whose performance levels must be ‘measured. 1.3.1. The most common reference antennas Consider two types of radio links. Each type determines a particular type of reference antenna, 1. Links where the antenna’s height above ground is considerably greater than the wavelength. This is normally the case for the metre range and shorter waves. Reference antennas are then the Hertzian dipole, the half-wave dipole and the isotropic antenna. 4) Hertzian dipole: a current element whose length is considerably shorter than that of the wavelength on which it operates. It can be shown that the radiation pattern of such a dipole ts @ torus whose axis is the dipole and ‘whose cercle de gorge has a zero radius. If this antenna is used as a reference antenna, the direction considered isin the equatorial plane. ') Half-wave dipole: a linear radiating element whose length is half the wavelength on which its operating. Its radiation pattern is very similar to the one above, but the semi-meridian of the torus isnot quite circula ¢) Isotropic antenna: an antenna which, by definition, radiates equally in all directions. Its radiation pattern is therefore a sphere centred on the antenna. Itean be shown that such an antenna cannot be built as isotropic radiation is incompatible with the transverse character of electromagnetic, waves. However, this antenna is increasingly used as a reference antenna because ofits simplicity. 2. Links between antennas whose height above the ground is very much less than the wavelength. This is generally the case for low frequencies. Only vertical polarization is viable and in this case the reference antennas are: 4) Short vertical antenna: above the ground which is assumed to be infinitely conducting. This is half of an Hertzian dipole.2 RADIO WAVE PROPAGATION b) Quarter wave vertical antenna: above the ground whichis assumed to be infinitely conducting. Thisis half ofa half-wave dipole. These two antennas have the same radiation pattern above ground as the symmetrical antennas from which they are derived, but the fact that the power supply to them is radiated in only half the space must be taken into account. 1.3.2. Gain and effective area of an antenna Ifa transmission antenna is used instead ofa reference antenna, the field in ‘general has a different value. The power P ofthe transmitter becomes the effective power P/ and: ” G=5 uy ‘where Gis the infinite gain ofthe transmitting antenna in the direction considered. In cases where the reference antenna isan isotropic antenna the gain is called the absolute gain or isotropic gain ofthe antenna. For example, it can be shown that in free space the Hertzian dipole and the half-wave dipole have absolute gains of 1.5 and 1,64 (1.75 and 2.15dB), respectively. Consequently, if one of these dipoles is taken as the reference antenna, the corresponding relative gain of any antenna is measured as a number that is less than the value for the absolute gain of the same antenna by 1.754B or 2.154B, respectively. ‘A receiving antenna, placed at a point of a plane wave where the surface density of the power is 5, receives a certain power P proportional to S. The coefficient of proportionality, which has the dimensions of an area, is called the effective area of the antenna. This area is not necessarily simply related to the geometrical dimensions of the antenna, Nevertheless, in most cases dealing with ‘waves in the decimetre range (roughly UHF)or shorter, where the antenna presents large opening perpendicular to the direction of propagation, the effective area is generally proportional to the area of this opening, the coefficient of proportionality being between 0.5 and 0.7. Gain is therefore defined for a transmitting antenna and effective area for a reception antenna. Since the same antenna can be used either for transmission or for reception, there is necessarily a relation between these two quantities. Applying the principle of reciprocity, itcan be shown that the gain G of an antenna in a particular direction is proportional tothe effective area in the same direction. The coefficient of proportionality is indepedent of the antenna and it can be shown that for a wavelength 2. in free space: (1.2) 1 WAVE PROPAGATION IN FREE SPACE 2B For anisotropic antenna gain equals 1, and consequently the effective area A, of an isotropic antenna equals: 2 A= aR" a3) ‘Under non-free space conditions these equations no longer apply. In particular, if a vertical antenna, on an infinitely conducting surface, is compared with a dipole antenna obtained by adding a portion symmetric to it with respect to the surface, With respect tothe dipole itis found that: the gain is twice as great; the effective area is halved; the radiation resistance is halved; the power necessary to create a given field is halved; the G/A ratio is four times greater. 1.4 Electromagnetic fields and attenuation of propagation in free space ‘A transmitting antenna supplied with power P radiates an electromagnetic ficld Ein the entire space, and a receiving antenna in this field will receive a certain power p. There are therefore two different ways of handling a propagation problem: 1. Calculate the electromagnetic field created at a given point by an antenna fed from a transmitter. 2. Calculate the ratio between the transmitted and received power. This ratio, called the transmission loss, is usually expressed in decibels. ‘The choice of method depends on the problem being studied. Ifa transmitter must serve a large number of receivers whose positions are not necessarily known (radio broadcasting), the first method is to be preferred. On the other hand, in a point to point communication (radio relay link) the link forms a whole in which the transmitting and receiving antennas play symmetrical roles; i is then natural to use the second method. 1.4.1, Field in free space created by an isotropic antenna A transmitter of power P supplying an isotropic antenna creates at distance d ‘a spherical wave transporting a power flux density S such that: P Sta aa ‘This power flux density is equal to the amplitude of the Poynting vector. S=EXH. asm4 RADIO WAVE PROPAGATION ‘At great distances from the antenna the vectors E, H, and § form a direct tri- rectangular trihedron and the scalar relation is obtained: E_ [te 5 - |i ootm 00 The modulus of § therefore takes the valu: S=E ic a7) from which the electric field can be deduced: LP E= T: (1.8) vihich i often expressed (Fig. 11): Pam Vimy = 173 : As Feat = BYTOM ay ‘The product Ed, which is a voltage and is generally expressed in volts, is known as the cymomotive force (c.m..)- If instead of the isotropic antenna, the Hertzian dipole or the half-wave dipole is taken as the reference antenna, the constant 173 in Equation (1.9) must be replaced by 212 or 222, respectively. In other words, for IkW of radiated power, the ceymomotive force (c.m.) is: 173 volts for an isotropic antenna; 212 volts for a Hertzian dipole; 222 volts for a half-wave dipole antenna. Equations (1.8) and (1.9) apply to waves with rectilinear polarization. If polarization is elliptical and if £, and E, are the components of the electric field along two orthogonal axes, the first elements in these equations must be replaced by VELTET, From this, and E, can be deduced only ithe elipiity ratio is known. For example, for circular polarization E must be replaced by 4.4.2. Propagation in the free space between two isotropic antennas: free space basic transmission loss ‘This is the term used for the ratio between the power P, transmitted by an isotropic antenna and the power P, received by another isotropic antenna at a distance d away and in free space. Now, for any two antennas: P_ GA Pane where G is the gain for the transmitting antenna and A is the effective area of the receiving antenna. For isotropic antennas from Equation (1.3: (1.10) re0-f Distance in kilometres |. WAVE PROPAGATION IN FREE SPACE. Vm een ta athe Hiner fw ste Pils ong-ao Hin do ot-~ « 4B with respect to InV/m, yew wd eo thw wodhw 4. z = Effecive isotropic radiated power (EIRP)6 RADIO WAVE PROPAGATION (#),- (2) Free space basic transmission loss is given by the reciprocal of this ratio, which is ‘generally a very large number, particularly for decimetre or shorter wavelengths; it is usually expressed in decibels (Fig. 1.2). If transmitting and receiving antennas with respective gains G, and Gare used, aay (12) Pa laa 2 06i(g4g) 6 ay ‘The quantity P,G, is called the effective isotropic radiated power (EIRP) of the transmitting station in the direction considered. This concept has replaced the older idea of effective radiated power (ERP) where gain was measured not with respect, to an isotropic antenna, but with respect to a half-wave dipole. ERP is therefore 2.15AB less than EIRP. For long waves the effective radiated power from a short vertical antenna (ERPV), is also used. Example: Consider a link over SOkm at 6GHz with 40 dB gain antennas and a 2W transmitter; if losses are ignored: free space attenuation between isotropic antennas: 1424B; free space attenuation between antennas used: 624B;, EIRP: 43dbW; received power: -59dBW. NB. Use is sometimes made of the concept of “free space’ attenuation above a ‘conducting surface: To do this certain precautions are necessary. See in particular Section 13.2 . Relations between different quantities characterizing a plane wave [A plane wave may be characterized at a point by various quanti, selected according othe problem being considered They are 35 follows: E, the electric field; A, the magnetic field; S, the surface power, p the power received by an isotropic antenna; 1. WAVE PROPAGATION IN FREE SPACE Distance in Kilometres EE ad Foon E-s000 EE ‘Avenuation between eotropic antennas (4B) feedeh a & i Frequency eka Pree fee SPE SEETS ? Wordleagin rs oe Fig, 1.2. Free space attenuation between isotropic antennas.B RADIO WAVE PROPAGATION +, the voltage induced at the ends of an isotropic antenna of given impedance, eg. $0 ohms. ‘Since itis practically impossible to construct an isotropic point source, the last two ‘quantities will be used as intermediate values for applications concerned with any antenna whose characteristics are defined with respect to an isotropic antenna, ‘These quantities are related as follows: so Eo = 00 =P ss ‘The nomographs in Fig. 1.3 may be used to work out all these relations except the ‘magnetic field which is not often used. a4) 1.6. Propagation attenuation over a link that includes a pastve reflector 7 In certain cases a link is the result of reflection from an obstacle, whether intended for that purpose or not. This isthe case, for example, in a radio relay link using a plane passive reflector, and is also the case for radars used to detect specific shapes (G0 itis not true for meterological radars). Assume that the obstacle isin the far field of the antennas. Periscopic antennas which include a reflector in the near field are different, and have to be dealt with when studying the antenna itself. ‘We use Eto denote the effective area of the obstacle in the direction of the transmitter at a distance d, and G for its direction with respect to the receiver's direction, which is at a distance dj. Assuming that the transmitting and receiving antennas are isotropic, the ratio of the received power to the transmitted power is ‘equal to the product of the following factors. Perens B > Fndh \andy (1s) 1. Incases involving a plane reflector, used for example, for a telecommuni- cations radio relay link (see Fig. 1.4), the reflected wave has the same phase structure as the incident wave and the effective area of the reflector is the same for both antennas. Consequently: o-*8, (19 trom which we conta! Pp, zy (wea) ie note that attenuation does not depend on wavelength. 2. In the case of an obstacle, of any given shape, placed in front of standard ‘monostatic radar (see Fig. 1.4), directions and distances are the same for |. WAVE PROPAGATION IN FREE SPACE. e558 Field in pV Power density per square mere in 4BW “ Received voltage over $0 ohms ig. 13. Characteristics ofa plane wave, ages e838 8 eee Wavelength30 RADIO WAVE PROPAGATION Radar Fig, 14. A comparison between telecommunications and radar both paths. On the other hand, itis not generally possible to determine the structure of the seatered wave, ie. the relation between G and is tunknown. Under these conditions the product GEis taken asthe effective radar cross-section Ex of the obstacle in the direction in question. We can therefore conclude that: Po dye Pape ‘The inclusion of factor G means that the effective radar cross-section can be ‘very different from the effective geometrical cross-section of the obstacle, dnd also tha it depends on is vrientation with respect tothe radar beam. 1.7 Propagation at short distances from the transmitting antenna (ts) In the discussion above it was assumed that the distance between transmitting and receiving antennas was very large compared with the wavelength and the size of the antenna. This is not necessarily true under all conditions. In Section 1.7 we consider the two most common situations in which these conditions are not met. 1.7.1. Very high gain antennas If the transmitting antenna has very high gain in one direction, itis necessarily very large. with respect to’ the wavelength. The most common case is the ‘antenna with a parabolic reflector used in point to point links. Results obtained above concerning the field or transmission losses only apply beyond a distance where the differences in path lengths from different points on the antenna are neglible with respect to the wavelength, If D denotes the diameter of the antenna | WAVE PROPAGATION IN FREE SPACE 31 and 3. denotes the wavelength, distance d must satisfy the following condition: 2D? ‘D\* a> 2D, ot $>2() (119) (Or a condition which can be written in a still more general form as follows: a f>04G, (1.20) ‘where G is the antenna gain expressed as a power ratio (not in decibels) in the direction in question. ‘The minimum distance obtained in this way may be very great in certain cases. For example, for an antenna used in satellite telecommunications, with a diameter ‘of 10m and operating at 14GHz, the minimum distance along the axis ofthe beam is 9.3km. At a closer distance from the antenna the field along the axis of the beam no longer follows a 1/d relation, but oscillates around a nearly constant value. 1.7.2. Very low gain antennas If the antenna is very small with respect to the wavelength, its radiation is, indistinguishable for practical purposes from that ofan electric or magnetic dipole with maximum gain of 1.5, and the actual size of the antenna is no longer a contributing factor. This case is met in particular when studying unwanted radiation from certain equipment. ‘At distances less than the wavelength, radiation no longer has a plane wave structure, i. the ratio between the electric field and the magnetic field is no longer siven by the equation: E_ fie ge fe = Zy = 120 nohms a2 In particular, if the wave is treated as though it was radiated by an electric dipole, the ratio between the electric field and the magnetic field at a short distance d with ‘respect to the wavelength A, in the equatorial plane of the dipole diameter is given by: Elz Hina (1.22) In this ase, « measuring recive with an antngasenstve tthe ec fd (Get ape wil e's higher eading han it eulped wi an antenna sensitive to the magnetic field (frame). The opposite is true if the wave is radiated by an element similar to a magnetic dipole, i.e. a ring or a coil, for then at short Soancee E a Fn 280s. (1.23)Chapter 2 Fundamental propagation phenomena: summary of physical concepts 2.1. Introduction For the purposes of studying propagation in free space as described in ‘Chapter 1 itis assumed that the propagation medium isa vacuum housing only the ‘transmitting and receiving antennas of negligible size. Real conditions generally differ significantly from this ideal case. On the one hand, the characteristics ofthe propagation medium are different from those ofa vacuum and can vary from point to point, while on the other hand the propagation medium may contain bodies with very varied electrical characteristics, producing discontinuity surfaces. Here antenna sizes are still considered to be negligible with respect to the distances in ‘question, From the theoretical point of view, the study of a propagation problem involves solving Maxwell's equations, which can be theoretically determined as soon as we know 1. the position and the nature of the sources; 2. the electromagnetic characteristics of the propagation medium or media (e.g. the refractive index); 3. the boundary conditions on surfaces delimiting or separating media; 4, initial conditions (these generally vanish as we are usually only concerned with solutions which are sine functions of time). In practice, this method can only be used at the price of introducing major simplifications, which still leave the problem too complex to handle. Iti therefore ‘usually reduced still further into individual models for which far simpler specific ‘methods are already available. Clearly, this leads to problems of reconciling different solutions with each other, and these difficulties have to be overcome for ‘each individual case. Introducing these problems also means considering ‘propaga- tion phenomena’ or ‘propagation mechanisms’, such as line of sight propagation, diffraction, et., which also means interpreting Maxwel’s equations, but in theory ro such interpretation should be necessary. 2. FUNDAMENTAL PROPAGATION PHENOMENA 3 2.2. The line of sight propagation model ‘The simplest model is drawn from optics. The concept of a wave is replaced by that of a trajectory or a ‘ray’. Therefore, phenomena such as reflection and refraction are involved. The concept of a wavelength no longer applies and Maxwel’s equations are replaced by far simpler relations, such as the equality of angles of incidence and reflection or Snell's law of refraction. ‘The crucial problem is therefore to determine cases suited to this purely ‘geometrical method. In optics the concept of line of sight is used, and that concept has to be re-analysed to make it applicable to radio waves. We start with the following comment: an electromagnetic wave is characterized not only by electric and magnetic parameters, but also by a geometric parameter, its wavelength, Optics ignores this parameter, which reduces to the wavelength being treated as though it ‘was infinitesimal with respect to the size of all objects lying in the path ofthe waves. Given the shortness of their wavelengths, this condition is easy to apply to light waves. It is, however, far more difficult to apply to radio waves, where the wavelengths of the shortest waves commonly used are some 10* times larger than for light waves, and this factor can be increased to 10" if the whole radio spectrum is covered, 2.3. Definition and criterion for a line of sight link Great care is therefore needed ifthe concept of line of sight is applied to radio waves, and infact it can only be applied negatively by defining a ‘non-line of sight” link. Tn optics such a link would be associated with a zero field strength. In fact, however, the field strength is not ero. To explain this fact, the phenomenon known as diffraction is introduced. In this way, a line of sight link can be defined as a link for which diffraction is negligible, and therefore one for which only reflection, refraction or scattering phenomena apply For this definition to be usable a simple criterion to determine whether of not diffraction can be neglected must be established. ‘Consider the situation in Fig 2.1 Signals are transmitted from a transmitter T and received at R; Pisa plane perpendicular tothe straight line TR at point 4. Using Maxwells equations to calculate the field strength at R we only need to know the field strength created by T in all points of plane P. Moreover, if polarization, is ignored, all points on the circle within plane P about a centre at A make the same contribution to the field at R, because they are in phase with each other. Therefore, plane P can be divided into concentric rings and the contribution to the field at R made by each ring, taking into account the phase variation from one to the other, ‘an be calculated. Phase differences are calculated with respect to the origin defined by ray B following the shortest path, TAR. The path through B is longer than the straight line path by a quantity A Z which can easily be calculated. All points with thesame value for AL make the same contribution to the field strength at R.Ifplane Pis displaced parallel to itself and AL has the same value, point B will generate an ellipsoid of revolution with E and R as foci. Therefore, the space is divided into a4 [RADIO WAVE PROPAGATION Fig. 21. Determination of Fresnel lipids. series of homofocal ellipsoids. Furthermore, itis clear, and can be shown by the stationary phase theorem, that most of the energy at R has travelled along TR. As long as: AL = TBR ~ TAR @uay is less than half the wavelength, the points within a circle of radius AB, summed at R, have the same sign. Ifthe difference is greater than half the wavelength, some of the points will supply a negative contribution, and the field at R will decrease. It increases once the difference is greater than the wavelength, and so on. Let AL 2.18) ‘The integer values of n define a group of ellipsoids called Fresnel ellipsoids, and the first of them (n= 1) plays a special role, in that it defines the region of space through which most ofthe energy travels, and therefore the region which needs to be free of obstacles if propagation is not to be ‘disturbed’. This makes it possible to define conditions for line of sight propagation. Closer study shows that only about 60% of the first ellipsoid needs to be free of obstacles for the received level to be equal to that in free space (see Chapter 7). From Fig, 2.1 itis immediately clear that: 2. FUNDAMENTAL PROPAGATION PHENOMENA 35 aL = JF + JF pad. @2) Since p is generally small with respect to a and 6, a limited expansion can be used: ): es Fora given path length d= a+ b, AL is maximum when a= 6, ie. at the mid-point of the path and: (AD nus ‘The first ellipsoid is obtained when the value of ALis equal to half the wavelength. ‘The equatorial radius of the ellipsoid is therefore equal to: Sid. @5) To give an idea of the orders of magnitude involved consider a SOkm link operated ata wavelength 2. The following values for pe are given in Table 2.1. 24) Pe a Pe o.5um 0.08m (ight waves) 50cm 25.00m (centimetre wave range) 5.0m 250.00m (metre wave range) 500.0m 2500.00m (100 metre wave Fange) ‘able 21, Relationship between and py. Table 2.1 shows that the concept ofa “ray” has practical significance for light waves because even on a SOkm lnk all the energy is concentrated within a cylinder a few centimetres in diameter. On the other hand, for the centimetre wave range the diameter ofthe cylinder is some tens of metres, making itnecessary to build antenna towers (fig. 2.2), and for the metre wave range or higher ranges itis virtually impossible to keep the first Fresnel ellipsoid free. In other words, for the latter ‘wavelengths ‘line of sight’ conditions are virtually impossible to create over a link atthe Earth’s surface, and diffraction is always involved (see Chapter 7).36 RADIO WAVE PROPAGATION yr Fig. 22, Debrition of line of sight lnk atthe Earth's surface, 2.4, Reflection and refraction ‘A major role in line of sight propagation is played by reflection and refraction, assume that the reader is familiar with these phenomena and only a few brief remarks are given here. Reflection and refraction take place when two different types of medium are separated by a surface which is very large and has very small irregularities with respect to the wavelength of the radiation considered. In particular, the radius of ‘curvature at any point on the surface must be very great with respect to the wavelength Its for this reason that these phenomena are stucied inthe first place fora flat and infinitely large separation surface. Surface curvature is then taken into account and the concepts of convergence, divergence and focussing are introduced. Finally, we study perturbations in these phenomena due to the fact that the surface is not extremely large, or irregularities extremely small, with respect to the ‘wavelength. In the same way, with the aim of simplifying analysis, these phenomena ae defined in terms of plane waves. It is well known that such @ wave is a mathematical solution of Maxwell's equations, but is not a physical solution, a it would transport infinite power. Its value stems from its simplicity, as any wave may bebroken down into a summation of plane waves, and finally because, iffar enough from its source, any wave has the structure of a plane wave locally nll cases, the formulas which express the laws of refraction and diffraction are derived from the following continuity properties which are themselves consequences of Maxwell's equations. At the separation surface between the two ‘media the following four properties apply: 1. the tangential components of the electric and magnetic fields are continuous, 2. the normal components of electric and magnetic induction are continuous. When considering wave propagation one medium is very likely to be the atmosphere, which can generally be treated as though it were a vacuum for the purposes of reflection or refraction (thisis not truefor other phenomena). The types ‘of medium in contact with the atmosphere (ground, sea) are characterized by 2. FUNDAMENTAL PROPAGATION PHENOMENA cy relative permittivity « (to simplify the notation ¢ is used to denote permittivity instead of ¢, which is the standard notation) and conductivity o, but their magnetic ‘permeability can generally be regarded as equal to that of the vacuum. Therefore, it is assumed throughout that relative magnetic permeability is equal to I Moreover, it is worth considering first of all the case where o is zero (dielectric ‘medium), since this simplifies the situation, 2.4.1. Reflection and refraction in dielectric media Ifmedia are purely ieletrc, there is no attenuation and te incident, reflected and refracted plane waves are each defined by a single vector perpendicular tothe corresponding equal phase planes. A first relation independent of the incident ‘wave's polarization can be obtained by assuming that the components u,v,,viof the phase velocities of the three waves (incident, reflected and transmitted) along the separation surface are the same. Using the notation given in Fig. 2.3: 7 26) en Which therefore gives: = 6, 8) law of reflection 1 sin 8, = my sin 8: 9) law of refraction (Snell's law) Other relations independent of polarization can be drawn from the fact that the waves are plane. For example, for each medium: Ez ar fe where Zp is the intrinsic impedance of a vacuum (120:r ohms) and ¢ and o are the relative permittivity and permeability of the medium in question. As jis generally equal to 1, ¢is equal to the square of the refractive index n of the medium: wel Jenn. a1) The relative amplitudes of the incident, reflected and transmitted waves can be derived from these continuity equations. As they depend on polarization, 2.10)8 RADIO WAVE PROPAGATION ig. 23. Reflection and refraction ofa plane wave. traditional to study two particular cases which cover most real cases, so-called horizontal and so-called vertical polarization. The plane separating the two media is assumed to be horizontal. 24.1.1 Horizontal polarization ‘The electric field is horizontal, i.e. parallel to the plane of separation. From the continuity relations: Et+BnE, 2.12) (H, — H,) cos 8, = H, cos 8, , (2.13) wa(H, +H) sin 8, = ay Hsin Op 14) If the relations, Equations (2.9) and (2.10), which are independent of polarization are taken into account the reflection factor (factor is better than cofficient for dimensionless quantities) for horizontal polarization can be calculated: E,_H, ROH Asin general n, and y, are equal, we can use the refractive indices m, and m, Then: Ry Qs) 21 FUNDAMENTAL PROPAGATION PHENOMENA » 1m £08 8, ~ My 608 83 t= i e080, + My C08 Oy (2.16) “which can be rewritten taking Equation (2.9) into account: Ry = 80 = 8) ani sn @, + 0) or again, using the angle of reflection @ which is the complement of the angle of incidence 6: Ry 18) 2.4.1.2. Vertical polarization ‘The magnetic field is horizontal, i. parallel to the plane of separation (the clectric field is not vertical although this might be the impression given by the term used). Continuity relations give: , - B) cos 6, = E,008 0; @19) Hy, +H, =H, 2.20) 2(E, + E) sin 8, = e, E, sin 8, (2.21) If Equations (2.9) and (2.10) are used, the following equations are obtained: 2 223) 2 ‘Vertical polarization has a particular property derived from Equation (2.23): the reflection factor vanishes for a value of the angle of incidence given by: +6 <3 0.25)0 [RADIO WAVE PROPAGATION If in addition, we take into account Snel’ law, Equation (29): tan 9, = 226) Which defines Brewster's angle. We then talk of Brewster's angle of incidence. In the case of Brewster's angle of incidence the refracted ray is perpendicular to the reflected ray (which has a zero amplitude). NB. In Equations (2.18) and (2.24), which give the reflection factors, itcan be seen that the refractive indices of the two media only appear in the square of their uotient, which can therefore be replaced by the quotient of permittivities. In cases ‘where the relative permittivity ofthe first medium is equal to I the classic equations can be obtained: sing — Je — cos? @ Ru sin +e — cos? 9" em) 224) sing ~ z sng Ee Ry 2.4.1.3 Case of grazing incidence If @ is close to zer0, Equations (2.18) and (2.24) can be replaced by a expansion. Ifthe ratio n/n, (the ratio ofthe refractive index of the second medium to the refractive of the first) is by denoted n: 29 Ry = 1432 27 ty = 14+ 75h @2n ate a2 Ry= 1+ If furthermore n is very much greater than 1, which is frequently the case: Ry = -1 422, 229) Ry = -1+2n9. (2.30) 2.4.14, Case of two media with ony slightly different refractive indices ‘Under certain conditions an approximately plane horizontal surface may form atmosphere where its refractive index undergoes a slight discontinuity, by a positive or negative amount Av. The reflection factor associated with such a change is now considered. First, assume vertical polarization. The reflection factor is given by Equation 2. FUNDAMENTAL PROPAGATION PHENOMENA 4 (2.24), As m and n, are close to 1: : (By 21+ 2a. ean Equation (2.24) then becomes: py wht 2AN sing = tao ETI eo” (1 +240) sing + /sin? 9 + 2An Figure 2.4 gives the modulus of the reflection factor R, as a function of angle @ for ‘various values of Av (in fact R, is negative). If Am is positive (movement into a more highly refractive medium), the curves show no discontinuity. On the other hand, if ‘Anis negative (movement into a less refractive medium), there is total reflection for any angle less than the li = J= Thin. 233) For values of R, less than 0.1 the sign of An is not very important and: Lan Raat 234) ‘With horizontal polarization and using the same assumptions: Ry = Sit aia FB : e35) sing +,/sin? 9 + 2An For the very small values of A, and @ considered here we can therefore write, forall practical purposes: Ry = Ry, 236 ‘and the curves in Fig. 2.4 can be used for either type of polarization. 2.4.1.5. Gradual variation of the refractive index within a thin layer Assume that the refractive index varies over a small layer of atmosphere of depth Ah between altitudes A, and , with a variation of n given by: an Rnfn). 237 Let us also assume that the reflection factors ess than 0.1, making t possible to use the approximation given in Equation (24). The fact that the phase difference between the incident ray and the reflected ray will vary within depth AA must be taken into account, which gives:2 IR | Modus ofthe reflection factor RADIO WAVE PROPAGATION ww Fig. 24 Reflection due to reftactive index discontinuity. ‘0 ‘Angle in radians 2. FUNDAMENTAL PROPAGATION PHENOMENA a a-5f sttyexp(— 54x sing) ah 23 If nis a linear function of A: Sy = Mt 239) and therefore: a ain( Ras 240) z 2 It is clear that a gradual change in the refractive index throughout a layer gives a reflection factor whose modulus is less than it would be in the case of a sudden transition, which may even vanish under certain conditions or for particular frequencies. 2.4.1.6. Media in which the refractive index varies continuously If the propagation medium shows no discontinuity but only a continuous variation in refractive index, the paths followed are not straight lines and once again ‘the phenomenon is called refraction. Ifthe equal index surfaces are parallel planes, Snell's law still applies but it has to be expressed as a differential equation. The ‘quantity 7 cos ¢ remains constant along the whole ofthe curved path in question. If the equal index surfaces are not parallel planes, Snell's law has to be replaced by another which can be derived from Fermat's principle by writing thatthe value of the integral: [ow @aiay is stationary. For example, in cases where the equal index surfaces are concentric spheres, if we use polar coordinates r, 8, the problem becomes one of making the integral: [i T+? 0? dr. ip) stationary. As the function under the integral sign does not depend on 6, Euler's condition immediately gives: Po as Tire which can also be written as: nr) Cte (242A)“ RADIO WAVE PROPAGATION nr sin V = Cte (2.428) where His the angle between the tangent to the path and the radial vector (a more clementary demonstration is given in Chapter 4). 2.4.2. Reflection and refraction between a dielectric medium and an ‘imperfect’ conducting medium This isa very common case in propagation, since the Earth's surface separates two types of medium. The atmosphere can be regarded forall practical purposes as a vacuum, and the Earth is characterized by its relative permittivity © and its conductivity 6, It is well known that if we introduce a complex permittivity: n=6-j 600d 2.43) the equations for reflection and refraction take the same form as in the previous case, but the reflection factor and the angle of refraction take complex values which hhave tobe interpreted physically. “The complex nature ofthe reflection factor means that reflection introduces a certain phase difference other than 0° or 180°, Calculation of ths factor therefore ives oth the amplitude and the relative phase ofthe reflected waves. Ifthe incident wave shows rectilinear polarization, so will the reflected wave. In vertical polarization the modulus ofthe reflection factor goes through a non-zero minimum fora certain angle of incidence. By analogy with the dielectric medium case, this is called the pseudo-Brewster angle of incidence. ‘The fact that the angle of refraction is complex may be interpreted as follows. It is known that a plane wave in a dielectric medium may be characterized by @ vector q perpendicular to the equal phase surfaces (called the wave fronts), and 6 then represents the angle that this vector makes with the normal to the separation surface between the two media. Ifthe conductivity ofthe medium is non-zero, there also exists a second vector p perpendicular to the equal amplitude surfaces, which are, in te case of refraction, parallel tothe separation surface. The wave is then said to be disassociated or heterogeneous. In general, the two vectors p and q are not identical, and 6 then represents the angle formed with the normal by the complex vector: c=atip (4a) ‘The real part of this vector gives the phase velocity u: (2.45) 2. FUNDAMENTAL PROPAGATION PHENOMENA 4s ‘and the imaginary part gives the attenuation a perpendicular to the surface: ane 2.46) Ifthe incident wave s perpendicular to the separation surface, the two vectors pand q are collinear, and the refractive wave has the same structure as a plane wave, except that fields E and H are not in phase and there is a gradual attenuation. If v and x denote the real and imaginary parts of the refractive index, e.: Viav tin, ean) we obtai x eae 2.48) a ow) where 2 is the wavelength in the vacuum. tn ee va ee eta ease Oy ce tty eas Ser arate aces occas a: refracted wave enters the conducting medium at almost right angles to the surface. ‘This is the case in particular for refraction of a radio wave by water, for which ¢ is close to 80 (it is approximately 2 for light waves). 2.4.3, Depth of penetration and wavelength in the conducting medium Consider Equations (2.48) and (2.49) once again, which are exactly true for normal incidence and approximately true for any incidence when ¢ is relatively large, two other quantities can be defined. On the one hand, the wavelength 2 in the conducting medium: Jn ay and on the other hand the depth of penetration, ie the distance atthe end of which the amplitude is reduced by a factor e: eae a Ta In Equation (2.43) suppose that: (2.50) Qs) od ae in 5, 252)