Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
BEMI - BÄTTRE ELMILJÖ Sida 1 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT Törnevalla 2010-04-26 REPORT #BEMI1004261 Description of measurement methodology BEMI – BÄTTRE ELMILJÖ Clas Tegenfeldt Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 2 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT NATURE OF RADIATION FROM ANTENNAS A wave is a sort of oscillation, it could be sound (mechanical vibration in a medium) or electromagnetic (no medium necessary) such as radio or light. A wave has an amplitude which can be described as strength or intensity. The rate it oscillates, its frequency, is measured in Hertz (Hz). Because it also travels the waves peaks and zero crossings also moves. The distance it moves while oscillating one time is called the wavelength. The wavelength is measured in meters. The wavelength depends on the propagation speed and the frequency. The product of wavelength and frequency is equal to the propagations speed. For an electromagnetic wave in free space the speed is equal to the speed of light (which also is an electromagnetic wave!). Let c be the speed of light, f is the frequency in hertz and λ is the wavelength in meters, then c 299792458 c=⋅ f , eller = = f f The amount of power a wave carries is called Poyntings vector Taking the magnitude of that vector gives the power density S= E2 2 [W / m ] in watt per square meter. 377 In fact, it is almost as simple as Ohms law and the definition of power (P=UI). What it tells you is that the electric field and the magnetic field in combination carries the power in the travelling electromagnetic wave, or if you prefer - the electromagnetic radiation. The number 377 Ohm is the impedance of free space. Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 3 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT Since the wave/radiation can move in any direction one often refer it to some coordinate system, such as x, y and z. The electric field E can then be described either as the vector in space or as three components along the x, y ans z axis. When measuring the electric field at some point, one either has to make sure the sensor antenna is directed along the E vector to collect the full strength of the field/wave/radiation, or measure in the three orthogonal directions x, y, and z and then use the above formula to find the E vector. Antenna facts An antenna is just something electrically conductive that is used to emit electrical energy into the air as a electromagnetic wave. It could be in almost any shape and size, however, since the wavelength is dependent on the frequency used, the size of the antenna has to be related to the wavelength. For low frequency radios the antennas are large, for mobile phones using high frequency (microwave) the antennas can be small. The antennas used for basestations could be the same as for the mobile phones, however the mobile is driven by battery and has limited power output at the same time the base station has no such limitation. The reciver in the base station can be made sensitive (and expensive) at the same time the mobiles reciever should be simple and cheap. This in combination makes it necessary for base station antennas to be larger, typically a meter or so high and a decimeter or two wide. Also the power output through the base station antenna will be tens of times higher than from the handheld mobile. Antenna diagrams An ideal antenna could radiate equally in all directions, this is called an isotropic radiator. This antenna isn't physically implementable but can be approximated at larger distances. For mobile phone systems as well as broadcasting there is actually no need to send signals straight up into the air or down directly into the ground. Thus in actual use one want directional antennas., antennas that can throw the radiation in the needed direction along the earth surface. To characterize antennas one uses antenna diagrams that describe the distribution of radiated energy along different directions. Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 4 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT The most of the energy is radiated along some direction, this is called the main lobe. If the directivity of the antenna is high this main lobe is narrow, this is called a "high gain antenna" since the signal will be strong in that direction compared to other directions. Gain is a measure of how narrow an antenna radiates its energy. High gain means that the field levels and power density is increased in the main direction (in the main lobe) while at the same time decreases in other directions (back lobe, side lobe and nulls). To simplify the antenna diagrams are flattened in 2D either as a linear or polar diagram. Let us take a closer look at one of the antennas used in Zambia. It is identified as "AP901213 120 degree directional panel antenna 890". It is a GSM900 antenna and has a typical antennadiagram in the horisontal and vertical planes as shown below: The horisontal diagram (above left) shows that even though most energy is thrown in the main lobe along the xaxis (to the right), radiation also goes in other directions. Since a mobile phone user may be anywhere this is often desirable. In the vertical diagram (above right) it is even more pronounced that the energy is along a line parallel to the earth (ground). Some sidelobes will reach the ground near the antenna mast, but most of the energy is thrown far away. GSM900 can reach 35 km and even 70 km if placed at high masts. The antenna diagram implies that simulations in the vicinity of a base station must take both antenna diagrams (or a complete 3D data set) into consideration to be able to make a proper prediction of field strengths. Farter away, say 500 meter or more, the radiation more or less follows only the horisontal antenna diagram. Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 5 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT Radiation around base stations The image is from the ITU-K.61 document. When an antenna transmitts electromagnetic energy this radiation can be reflected on the ground or other surfaces such as buildings. Thus there exists multiple pathways for the electromagnetic radiation to reach the body (or measurement point). To be able to understand the various factors influencing the exposure one needs to begin with the antenna. In the diagram below there is a 50% reflective flat ground but no buildings (no other reflections than from the ground). 140 Red line is spherical propagation in free space (no ground plane, a perfect isotropic antenna) Blue line is with antenna diagrams, measured along the main horisontal lobe direction, and using the vertical diagram for every distance point, above a groundplane. 130 120 110 100 90 80 0 100 200 300 400 500 600 700 800 900 1000 The red line is the simple square law propagation from an ideal isotropic radiator in free space. This is just a simple reference line. The blue line is a GSM900 transmitter with 35 Watt output, neglecting cable losses, into the antenna "AP901213 120 degree directional panel antenna 890" which had its diagrams shown earlier. The antenna is at 20 meters height and downtilted 3 degrees. What is interesting to note is the in real life the field strength do not increase when the antenna mast is approached. This is due to the fact that the antenna height makes an increasingly steep angle upwards from measureing point and the antenna. Reversely, the antenna "looks" more steeply down on the observer and thus the vertical antenna diagram shows that less energy is radiated downwards than forward (horisontally). The maximum field strength will occur at 50-500 meters depending on the antenna height and the down-tilt angle. However, it is safe to say that the levels will be about 120 dBuV/m at maximum in most cases. Even with a low (6 meters) wall mounted antenna, downtilted 10 degrees, and full 35 W transmitter power, the maximum will be below 140 dBµV/m. Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 6 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT Antenna height relative to the person matters The illustration is from ITU-K.52 and shows a circular exclusion area around an antenna. The need to actually use an exclusion zone depends on the height h and type of antenna and transmitted power. For mobile phone basestations the exclusion area is mostly a concern for roof mounted antennas. If the height is more than 10 meters and the system used is GSM/UMTS then there is no need for an exlusion zone at ground level. However, if the antenna is placed on a mast or another house at the same height h and the distance d is short (less than 10 meters) then the exposure will be higher and the exclusion zone need to be evaluated. Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 7 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT Base station power variations in time When a mesurement is done, is the value representing the "normal" field strength? In other words, are there significant variations in the power output from the transmitter(s)? For FM radio the answer is simply, no the radiation is almost perfectly constant. For TV the answer is a little more complex, analog TV varies somewhat, but at a fast rate (image to image), while digital TV is more like the FM case. The analog TV, if measured a few minutes will statistically collect various image contents and thus the level is representative. However, if the transmitter is switched off during mesurements then it is of course missed completely! FM is usually around the clock while TV may be switched off certain during some hours each day. For the GSM case, the transmitters are always on. The total power variation for the GSM case is small. This is due to the fact that the first base station channel is always active with all eight time slots, filled with dummy data if necessary. This means that the transmitted power will remain the same until all timeslots are used by phone traffic. If there is more traffic channels available at the site the power will be increased by the transmitters output power in steps of 1/8 (eight time slots), if these are also filled with phone usage the output power has doubled (equals +3 dB). This variation is insignificant compared to other variations when measuring the exposure. In real life long time measurements, such as reported by Joe Wiart et al, or as BEMI has performed in Sweden, shows variations from 80% to about 130% as compared to the mean level. This is smaller variations than the theoretical 200% above base level. [See "Analysis of the influence of the power control and discontinous transmission on RF exposure. GSM mobile phones, Joe Wiart et al. IEEE Transactions on electromagnetic compatibility, vol 42, no 4, pp 376385, nov 2000.] For the 3G/UMTS case, the transmitters are also always on. The power variation is more complex than the GSM case, however the levels do not change significantly more, just faster. It is safe to assume that most power variations in time are accounted for during a 6 minute time, but bearing in mind that high traffic times may increase the exposure by a few dB, and bearing in mind that other transmitters such as TV may be switched off certain hours. Uncertainty of selected measuring point Waves may add or subtract onto eachother, creating interference, standing waves and something called fading. The so called superposition principle is the basis for all these phenomena. The radiation from a transmitter that may travel by many routes - by multipath propagation, will at the measurement point be blended. By measuring in one point only the question arises if this point is on a maximum peak, lowest valley or somewhere in between? How great difference is there to a point a few centimeters away? There are two ways to answer this question, theory or simply by measuring many points! Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/ BEMI - BÄTTRE ELMILJÖ Sida 8 av 8 Copyright © 1994-2010 Clas Tegenfeldt BEMI1004261.ODT For GSM1800 the wavelength is 16 cm, by moving the antenna and measuring the field strength variation, it easy to see quite large deviations due to fading. In some cases with nearby reflections the differencies may be even larger. Deviations of plus or minus 10 dB is very common. No matter how you coose one point there is no way of knowing if that is a "good choice". If you know the wavelength you can construct a measuring grid at certain distances to make statistically sure to catch a peak with a high degree of confidence. However, if you do not know the frequency or if there are multiple frequencies from different sources there are no common grid to use since it is frequency dependent. There have been many studies on this problem, and by increasing the measurement points at a location the error can be estimated. The graph on the right is one such attempt. For an average value (NOT peak field strengths) only three points are needed to get the error down to about 3 dB. However, for peak field strengths 20 points or more are needed. BEMI has over many years developed the method of a moving antenna. In fact, the ONLY way to make sure that you find the peak along one cycle is to actually MOVE the antenna and measure during the movement. This means that you sample the waves spatially. By keeping the instruments "peak-hold" value you will make sure to find the peak field strength at the location. By moving an antenna a few wavelengths in each direction the statistics gets very strong. Furthermore, polarisation also differs point to point making it either necessary to do a measurement along x, y and z-axis or tilting the antenna until maximum peak is found. The graphs can be found in ["Analysis of electric field averaging for in situ radiofrequency exposure assessment", Emmanuel Larcheveque, Christian Dole, Man-Fai Wong and Joe Wiart, IEEE VT 2005.] Measurement uncertainties for measurements of fields are the results of errors due to system instrumentation, field probe response and calibration, and the extrapolation, interpolation and integration algorithms used to determine the averaged field. For evaluation and expression of uncertainties, see [ISO/IEC GUM], [IEC 62311], [EN 50383], [EN 50400] and [b-IEC 62232], [b-FprEN 50492], [b-FprEN 50413] and [b-IEEE P.1597.1]. Föreläsning/utbildning Adress BEMI - Bättre Elmiljö Mejl [email protected] • Mätteknik DC-GHz • Analys/åtgärder/policy • Forskning/utveckling Törnevalla Gamla Skola 585 61 LINGHEM Internet http://www.bemi.se/