INFORMATION AND COMMUNICATION ENGINEERS SPS6-1 (6) Comparison of Different Kinds of Edge Tapering System in Microwave Power Transmission A.K.M.Baki a), K.Hashimoto b), N. Shinohara c), H. Matsumoto d), and T. Mitani e), The authors (a-c, e) are with the Research Institute for Sustainable Humanosphere (RISH), Kyoto University, Uji, Kyoto 611-11, Japan. The author (d) is the Executive Vice-President of the Kyoto University, Japan. Tel: +81-774-38-3864, Fax: +81-774-31-8463, a) E-mail: baki@ rish.kyoto-u.ac.jp Abstract Side Lobe Level (SLL) minimization is equally important as the achievement of highest Beam Collection Efficiency (BCE) for Solar Power Station/Satellite (SPS). When the amplitude distribution of all antennas is uniform then the main beam carries only a part of the total energy due to the higher SLL. If edge tapering is adopted for SPS then SLL will decrease and BCE will increase though edge tapering is a complex technical problem for SPS. So an optimization is needed between uniform amplitude distribution and edge tapering system. We have derived a new method of edge tapering called Isosceles Trapezoidal Distribution (ITD) edge tapering [1]. Only a small number of antennas from each side of the phased array antennas are tapered in this method. ITD edge tapering is almost uniform so it is technically better. We have compared different amplitude distribution systems; uniform, Gaussian, Dolph-Chebyshev and the newly derived ITD method. We have found that the BCE performance is almost equal in Gaussian and Dolph-Chebyshev edge tapering. Therefore in this paper we have described ITD, Gaussian and uniform distribution. The SLL reduction in ITD (except Maximum SLL) is even lower than those of other kinds of amplitude distributions. As a result the interference level becomes lower and BCE becomes higher in ITD. Here, 1. Introduction Solar panels of the SPS would be placed in geostationary orbit (GEO) at a distance of 36 km from the Earth s surface as shown in Figure-.1. Microwave Power Transmission (MPT) depends on the frequency, element spacing, and the size of antenna, output power and maximum power density. The SPS designers must be concerned with Beam Collection Efficiency (BCE), Side Lobe Level (SLL), size, weight, and cost among many other factors. The improvement of the MPT depends on efficiency in order to reduce the SPS costs. BCE and Maximum Side Lobe Level (MSLL) are used for an evaluation of the microwave beam. BCE of the MPT systems is the ratio of energy flow that is intercepted by the rectenna to the whole transmitted power. BCE for two dimensional array and rectangular/square rectenna can be expressed as: ( θ, θ ) dθ dθ P( θ θ ) BCE D = P x y x y x, y θryθrx θtyθtx dθxdθ y (1.1) and θ tx ; θ ty ; are ±9 degree angle sector. θ rx ; angle sector due to x dimension of rectenna. θ ry ; angle sector due to y dimension of rectenna. ( θ ) P θ x, y is the energy of the radiated beam pattern. BCE for one dimensional case can be expressed as: BCE = P θ r ( θ ) dθ P( θ ) θ w dθ (1.) θ r is the angle sector due to one dimensional rectenna θ w o is the angle sector ± 9. P ( θ ) is the energy of the one dimensional beam pattern. BCE should be more than 9 % to reduce the SPS cost. The BCE of uniform, Gaussian and ITD were compared and that in newly derived ITD is found highest. ITD also showed better performance in two dimensional arrays. - 13 -
INFORMATION AND COMMUNICATION ENGINEERS SPS6-1 (6) and the values of tapering level in ITD. Amplitude distribution A 1 =1 A =1 A 3 =1 A N =1 A tn A tn A t(n-1) At(n-1) A t A t U t U t(n-1) U tn U 1 U U 3 U N-1 U tn U t(n-1) U t h D U L U N- Fig-.1: ITD tapered phased array antenna. U N 3 Different kinds of Edge tapering for large array antenna Fig-1.1: Concept of Solar Power Station/Satellite. A laboratory experiment on uniform and ITD tapering system was conducted. A good agreement was found between the simulations and the practical results. The reduction of SLL and increase in BCE were achieved with the ITD tapering. Decreased SLL will decrease the ionospheric impact and increase the BCE.. The Concept of ITD ITD concept for 1D array is shown in Fig -1. Amplitude of only a few edge transmitting antennas/units are tapered in this method. Amplitudes of remaining most units are uniform. In recent SPS design, a concept of unit is adopted. Each unit consists of phased array antenna elements. A large number of units is used in MPT. In this paper 5 phased array antenna elements are considered as one unit for one dimensional case and the operating frequency is 5.8 GHz which falls in the ISM (industrial, scientific and medical) band. The operating frequency of each unit is same. It is possible to change the number of antennas/units to be tapered from each side Beam patterns, SLL and BCE for total 155 SPS 1D units (53875 antenna elements) were studied for total array length of 11.13m. Location of SPS is at a distance 36 km from the rectenna site. We studied beam patterns, SLL and BCE from db to db ITD and Gaussian distributions with a step of db variations. The number of units tapered from each side was 4 in ITD. The rectenna length was 3.9 km in this case. BCE and MSLL for the study are shown respectively in Fig-3.1 and Fig-3.. BCE is the highest in ITD for db and db tapering levels and slightly lower than BCE (%) 1 99.5 99 98.5 98 97.5 97 Level of tapering in ITD and Gaussian ( db ) Gaussian ITD Uniform Fig-3.1: BCE for ITD (4 units), Gaussian and uniform amplitude distribution system as a function of different tapering levels. those of Gaussian distribution for db and db tapering levels. But the same values in ITD can be made 14
INFORMATION AND COMMUNICATION ENGINEERS SPS6-1 (6) higher than those of Gaussian distribution by increasing the number of units to be tapered in ITD from 4 to 5 from each side. BCE in uniform distribution are the lowest. MSLL for both kinds of tapering are lower than those of uniform amplitude distribution. MSLL ( db ) Level of tapering in ITD and Gaussian ( db ) Gaussian ITD Uniform Fig-3.: MSLL for ITD (4 units), Gaussian and uniform amplitude distribution system as a function of different tapering levels Fig-3.3 shows the beam pattern for uniform, db Gaussian and db ITD (8 units). The MSLL is same for both ITD and Gaussian distribution in this case. BCE for uniform, db Gaussian and db ITD (as a function of number of units to be tapered) is shown in Fig-3.3. From the study if we compare up to db ITD and db Gaussian distribution we can find that it is possible to achieve the same BCE in both Gaussian and ITD cases. But in ITD it is needed to taper only 9.8% of the whole length from each side. The length of units is 9.8 % (total 18.54% from both sides) of the whole length. In ITD total 1755 units are of uniform amplitudes Normalized Radiation Pattern ( db ) -4-6.5.1.15. Scan Angle (deg) Uniform Gaussian ITD Fig-3.3: Beam patterns for uniform amplitude, db Gaussian and db ITD (8 units) cases. BCE ( % ) 1 99 98 97 96 95 4 6 8 No. of units tapered in ITD ITD Gausian Uniform Fig-3.4: BCE for uniform amplitude, db Gaussian and db ITD (as a function of number of units to be tapered) with large transmitting antenna at a distance 36 km. but in Gaussian case it is fully tapered. It is also possible to achieve higher BCE in ITD than that of Gaussian distribution by selecting tapering level between db and db and optimizing the number of units tapered from each side in ITD. ITD reduces not only the costs but also decreases the different levels of power distribution. ITD is technically better than Gaussian distribution. Other SLL reduction in ITD is even higher than those of Gaussian type power distribution. Therefore by incorporating ITD in SPS, it is possible to increase the BCE and reduce the costs, SLL and technical complexity those can arise in Gaussian or other kinds of distribution. 4. Two dimensional arrays with ITD technique. MPT technique was studied for two dimensional array and by taking uniform, -18 db Gaussian and -18 db ITD amplitude distribution of 5 5 elements. 6 elements were tapered in ITD from both side of the X and Y dimensions of the array. Element spacing was.75λ for both the two dimensions. We studied the beam pattern, MSLL and BCE. MSLL and BCE were calculated at 5.8 GHz MW frequency and by taking a rectenna (area 94. 94. sq.m) at a distance 1km from the transmitter. The area of the transmitting antenna array was.97.97 sq.m. The results of the simulation are summarized in table 4-1. The beam pattern of -18 db ITD in the XY, XZ and YZ projected planes is shown in Fig-4.1. Pattern in the XY plane is the contour plot of the beam pattern of the two dimensional array antennas. Highest BCE is achieved in ITD. ITD performs even better in two dimensional arrays. The main reason of this better performance is the more concentration of power in the 15
INFORMATION AND COMMUNICATION ENGINEERS SPS6-1 (6) main beam in ITD. Where as more power is absorbed in side lobes in two dimensions of Gaussian and uniform distributions. There fore it can be concluded that the validity of ITD technique does not change even if it performs extension to a two dimensional array. Fig -4. and Fig-4.3 show the beam patterns of two dimensional arrays of uniform amplitude and -18 db Gaussian distribution in the XY, XZ and YZ projected planes respectively. More powers are distributed in the SLLs in these two beam patterns. As a result beam collection efficiencies are less in these two cases as can be seen from Table-4.1. Table 4-1: Comparative study of different kinds of amplitude distribution system for two dimensional arrays. Type of Tapering BCE MSLL(dB) Tapering Level (db) (%) ITD -18-17.5 9.8 Gaussian -18-34.5 89.34 Uniform -14 81.15 5. Experiment We conducted a laboratory experiment on uniform and ITD edge tapering system by using the SPORTS (Solar POwer Radio Transmission System) 5.8 GHz beam forming subsystem in the Microwave Energy Transmission LABoratory (METLAB) of the Kyoto University. METLAB is an anechoic chamber for MPT experiment. Fig-4.1: Three dimensional projected beam pattern for -18 db ITD of two dimensional arrays in the XY, XZ and YZ planes. Projected pattern in the XY plane is the contour plot of the radiation pattern. Fig-4.3: Three dimensional projected beam pattern for two dimensional arrays of -18 db Gaussian amplitude distribution in the XY, XZ and YZ planes. Projected pattern in the XY plane is the contour plot of the radiation pattern. Fig-4.: Three dimensional projected beam pattern for two dimensional arrays of uniform amplitudes in the XY, XZ and YZ planes. Projected pattern in the XY plane is the contour plot of the radiation pattern. The experimental set up is shown in Fig.5-1. SPORTS 5.8 GHz beam forming subsystem consists of 1 1 micro strip antennas. Front view of SPORTS 5.8 beam forming subsystem is shown in Fig.. We used 1 elements phased array antenna elements. We kept this set-up for this experiment because this set up was used for some other experimental purposes. Last 3 elements from each side of the 1 elements were tapered. Other 6 antenna elements were of uniform amplitude. Remaining 1 1 antenna elements were stopped by connecting matched load. Element spacing of the phased array antenna can be set with in the wavelength of.6 to 1. 16
INFORMATION AND COMMUNICATION ENGINEERS SPS6-1 (6) these values is shown in Fig.5.5. Normalized Radiation Pattern ( db ) -4-3 -4-16 -8 8 16 4 3 Scan Angle ( deg ) Uniform distribution ITD Fig. 5-1: The experimental set up of SPORTS 5.8. Fig..3: Simulated horizontal beam patterns of 1 elements with uniform amplitudes and ITD. Normalized Radiation Pattern ( db ) -4-3 -4-16 -8 8 16 4 3 Scan Angle ( db ) Simulated ITD Measured ITD Fig.4: Simulated and measured horizontal beam patterns of 1 elements with ITD edge tapering. Adjustable Phased array antenna elements MW generator Fig- 5.: Front view of SPORTS 5.8 (during the experiment only upper 1 elements were used). We used.6λ element spacing during the experiment. Horizontal beam patterns were measured during the experiment. Reduction of SLL and increase in BCE were achieved with ITD edge tapering. Fig.5.3 shows simulated horizontal beam patterns of 1 element array antenna with uniform amplitudes and ITD tapering. Last 3 elements from each side were tapered in ITD. Simulated and measured horizontal beam pattern of ITD of 1 element arrays are shown in Fig.5.4.There was a difference between the simulated and measured beam patterns. The reason of this difference is explained here: Simulation result shows beam pattern for ITD of -.5 db,.9 db and db tapering levels from each side. But during the experiment -3 db, -6 db and db attenuators were used. The amplitude distribution using Moreover the attenuators used for the experiment had some attenuation as well phase errors. Maximum attenuation errors for -3 db, -6 db and db attenuators were -.56 db, -.8 db and -.5 db respectively. So the amplitude distribution of 1 elements array by using these attenuators had some asymmetry during the experiment and also the measured amplitude distribution was different from the simulated amplitude distribution. Phase difference between adjacent elements of the original 1 1 array was zero. But phase differences occurred when attenuators were inserted. Still there were some phase errors those could not be minimized though we tried to adjust the phases through computer adjustment (software adjustment) and by putting electrical stretchers (phase adjusters) with each attenuator. Moreover there were a chance of reflection. We recalculated the beam pattern by considering the amplitude and phase errors. 17
INFORMATION AND COMMUNICATION ENGINEERS SPS6-1 (6) Fig.5: Amplitude distribution of 1 elements with ITD edge tapering (practically there were some errors in attenuators). 1 elements were used instead of 1 1 elements, because this set-up was used for other experiments. The measured and simulated beam patterns are shown in Fig.6. The simulated and measured beam patterns show a close relationship with each other. Measured beam patterns could have been improved if it were possible to remove the errors mentioned above and by using 1D array. Normalized Radiation Pattern (db) -4-3 -4-16 -8 8 16 4 3 Scan Angle (deg) MEASURED ITD SIMULATED ITD WITH PHASE AND AMPLITUDE ERRORS Fig.6: Measured and simulated beam patterns (with phase and amplitude errors) and with ITD edge tapering. We calculated BCE by assuming a rectenna (length m) at a distance m from the transmitter. BCE was simulated from the measured beam patterns and the calculated BCE for uniform amplitude was 1% lower than that of ITD. Therefore it can be inferred that by implementing ITD of phased array antenna it is possible to decrease SLL and increase BCE. inspired by the pioneering work of W.C.Brown on MPT. The advancement of MW power from SPS is still in the early stage. BCE needs to be made as high as possible for MPT, because MPT is power transmission. It is difficult to increase the BCE with uniform amplitude distribution of the transmitting antenna because of higher SLL.Though it is possible to reduce the SLL and increase the BCE by adopting edge tapering in SPS units but it would be difficult to taper all the units and to maintain different power levels at different SPS units. It is possible to minimize the complexity of full edge tapering by introducing ITD for SPS system. Amplitude distribution in ITD is almost uniform. So ITD is technically better. ITD edge tapering is a new approach for the SPS system. Different kinds of amplitude distribution for 1D phased array antenna was investigated and it was found that it is possible to maintain higher BCE and reduced SLL by incorporating ITD edge tapering in SPS system. ITD technique is also valid for D array. Presently the authors group are also studying another approach of phased array antenna system for MPT with random element spacing []. Further study and research on ITD in SPS contexts are needed. Reference [1] A. K. M. Baki, N. Shinohara, H. Matsumoto, K. Hashimoto, and T. Mitani, Study of Isosceles Trapezoidal edge tapered phased array antenna for Solar Power Station/Satellite, IEICE Transaction on Communication ( under final review ), 6. [] Blagovest Shishkov, Naoki Shinohara, Hiroshi Matsumoto and Kozo Hashimoto, On the Minimization of Side lobes in Large Antenna Arrays for Microwave Power Transmission, proceedings of the 3rd International Symposium on Sustainable Energy System, Kyoto, Japan, Aug.3-Sept.1, 6, pp. 18-11. 6. Conclusions SPS can provide environmentally clean power and can meet the future energy demand. Microwave power from the SPS needs to be steered and controlled precisely to avoid any serious accident. Researches on SPS have been conducted by the engineers and scientists worldwide 18