3-6-3 beam forming network

1 Introduction

In order to establish a direct voice commu-nication link in the S-band between handheldterminals and a geostationary satellite [1]-[3],the satellite must feature an antenna with anaperture diameter of 10 m or more, which isdeployable in satellite orbit. Such antennaswill have very narrow antenna beams, and willentail the accurate pointing of multiple anten-na beams to cover desired service areas. Theonboard antenna system presents a number oftechnical challenges involving (among others)reflector structure, the architecture of the feedsystem, the beam-forming method, antennapointing method, and reduction in the weightand size of the antenna system.

In the ETS-VIII onboard antenna, the feedarray is “defocused” toward the reflector at acertain distance from the focal plane of thedeployable antenna reflector. Beam formingand pointing control are carried out mainly bychanging the phase of the feeding signal, inwhat is referred to as a phased-array-fed sys-tem [3]-[5]. This method offers the followingadvantages (1) the degree of freedom in beamformation is high, with redundancy ensured in

the event of failures of individual array ele-ments or of the feed system; (2) in the trans-mitting antenna system, problems associatedwith the electrical overload on the feed sectioncan be mitigated through the use of a spatialpower combine with multiple radiating ele-ments; and (3) allocation of onboard powerresource to antenna beams can be dynamicallyoptimized according to the regional trafficdemand without changing the driving levels ofthe power amplifiers, by sharing the excitationamplitude distributions of the feed arrayamong beams.

On the other hand, it is important to excitethe feed array in a highly precise manner inorder to secure the area gain required in amultibeam satellite communication system(i.e., the minimum gain required for a givenarea) and to ensure beam isolation (measuredas the ratio of the main-lobe gain of a givenbeam to the side-lobe level of another beamusing the same frequency). Moreover, since alarge number of feed elements are used inbeam formation, the scale of the BFN circuitincreases along with an increase in the numberof beams and feed elements. Therefore, theimportant technical challenges will consist of

MATSUMOTO Yasushi and IDE Toshiyuki 73

3-6-3 Beam Forming Network

MATSUMOTO Yasushi and IDE Toshiyuki

A newly developed BFN (Beam Forming Network) for a phased array feed on ETS-VIIIsatellite is described. The BFN controls the excitation weight of the feed arrays, which arelocated at a defocus position of the large deployable reflector antennas. The BFNdescribed here has a unique architecture by which excitation weight for beam steering canbe sheared by multiple beams. This enables simple and effective calibration of beam point-ing errors caused by mechanical, thermal, or electrical environment in satellite orbit. TheBFN also has a self-calibration function to reduce excitation weight errors caused by ther-mal environment.

Keywords Phased array antenna, Satellite antenna, Beam forming network, Multi beam antenna,Satellite communications

74

ensuring excitation accuracy and maintainingsimplicity and compact size in the BFN hard-ware.

The ETS-VIII will carry two kinds ofBFNs (BFN1, BFN2) designed to addressthese technical challenges, each using a differ-ent approach [3]. BFN1 embodies a uniformpointing control system [6], the result of sim-plification of the pointing control method.The BFN2, on the other hand, applies highlyintegrated large-scale MMIC technology toestablish independent pointing control for allbeams [7]. The BFN1 and BFN2 have beendeveloped by Communications Research Lab-oratory and NTT, respectively.

This paper describes the BFN1 the BFC(Beam Forming Controller). Based on the factthat pointing errors for multiple beams whichshare an antenna reflector are caused by manycommon factors, the BFN1 for the transmit-ting antenna has been designed to correctthese common pointing errors uniformly, withshared variable attenuators and shared phaseshifters. This method (uniform pointing con-trol method) can greatly reduce the number ofcontrol weights for pointing control, and cansimplify the control algorithm for beam point-ing. The BFC compensates for fluctuations inarray weight through onboard calculation, inresponse to temperature changes in theonboard equipment, in addition to its variousfunctions in beam-pointing experimentation.

Below we will describe an outline of thesatellite’s onboard beam-forming section; thestructures, functions, and performance of theBFN1 and the BFC; and the beam scanningperformance of the antenna.

2 Outline of phased-array-fedsystem and beam forming sec-tion

The ETS-VIII satellite carries two phased-array-fed large deployable reflector antennas,dedicated to transmission and reception,respectively [8][9]. The antenna reflector is anoffset parabola consisting of fourteen hexago-nal deployment modules. A feed array, in

which thirty-one cup MSA (microstrip) ele-ments [10] are arranged in a triangle is dis-placed toward the reflector by a given defocusdistance from the focal plane of the reflector.The position of the feed array, the number ofelements, element arrangement, and excitationweight are designed to optimize [9] antennaperformance. In the transmission system, theexcitation amplitude distribution of the arrayis shared, so that the transmitting power canbe freely allocated to multiple beams.

The beam forming section features twokinds of beam-forming networks (BFN1 andBFN2), for transmission and reception respec-tively, each of which can form three beams.The uniform pointing control method withshared weight is applied to two beams of thetransmitting BFN1. This method is theoreti-cally applicable to both transmission andreception, but the receiving BFN1 is operatedby independent pointing control, for the fol-lowing reasons. In the transmission system,the excitation amplitude distribution of thearray is shared by multiple beams, which ismore compatible with the uniform pointingcontrol method because both the excitationamplitude and the phase are controllable incommon. On the other hand, the receivingantenna is required to have greater beam isola-tion than the transmitting antenna. This isbecause one must consider the situation thatthe uplink signals are attenuated by fading,whereas uplink interfering signals from fre-quency-shared areas do not suffer attenuation.Moreover, the link margin is less in the uplinkbetween the ETS-VIII and a handheld termi-nal than that in the down link [11]. Therefore,it is best to individually optimize the arrayweight of each beam.

The excitation weight of the transmi-tting/receiving BFN1 is controlled by thebeam-forming controller (BFC). In order toavoid a critical problem that a single failure inthe built-in processor may cause the loss ofthe beam forming function, two BFC compo-nents are equipped separately on the satelliteto configure a cold standby system.

Fig.1 shows an example of the beam allo-

Journal of the National Institute of Information and Communications Technology Vol.50 Nos.3/4 2003

cation for multi-beam mobile communication.Each circle indicates a beam position corre-sponding to an area gain of 42 dBi. The samefrequencies are re-used for beam positions #1and #4 and for beam positions #2 and #5,respectively. Of the five beam positions indi-cated in the figure, the uniformly pointing-controlled beams correspond to beam posi-tions #3 and #4 among the three beamsformed by the transmitting BFN1. On theother hand, the independently controlled beamis formed at an arbitrary beam position.Moreover, when carrying out broadcast exper-iments, the array weight of the independentlycontrolled beam is modified to form a singlededicated beam.

3 Functions of BFN and beamscanning method

The functional block diagram of the trans-mitting BFN1 and the receiving BFN1 areshown in Fig.2. In the transmitting BFN, theinput signals to the uniformly pointing-con-trolled beams (to beam ports 2 and 3 inFig.2a) are output to the element ports throughthe shared variable attenuators and the sharedphase shifters shown in the figure. The sharedattenuators and phase shifters are used for cor-rection of the excitation error produced by fac-tors such as temperature changes in the com-

ponents of the feed system which are sharedby the beams (e.g., power amplifiers, filters,and feed lines in the transmitting feed sec-tion). Moreover, beam pointing errors causedby common mechanical factors (satellite atti-tude error, alignment error of feed-elements,alignment error of the reflector after deploy-ment, or reflector distortion, for example) arecorrected by antenna beam scanning, using theshared phase shifters.

In Fig.3, the phase shift ξn of element n


Beam allocation for mobile communi-cations.

Fig.1

Architecture of the BFN1.ATT:Variable attenuator, VPS:Variablephase shifter, HYB:Hybrid circuita)Transmitting BFN1, b)Receiving BFN1

Fig.2

Coordinate systems of a phased-array-fed reflector antenna.

Fig.3

76

necessary to scan in a direction (in units of[rad]) from (AZ, EL) to (AZ +δAZ, EL +δEL) is expressed by the following formula, inwhich virtual feed positions on the focal planebefore and after scanning are designated as rf

and rref [6].

ξn＝k(｜rn－rf｜－｜rf－rref｜) (1)

≒k(Foff /Df)(Xn･δAZ－Yn･δEL), (2)

where

k=2π/λ

rn=(Xn,Yn,Df): element position (n=1 to N),

rref =(Xref,Yref,Zref)=(-Foff･tan(AZ),Foff･tan(EL),0),

rf=(Xf,Yf,Zf)=(-Foff･tan(AZ+δAZ),Foff･tan(EL+δEL),0),

(δAZ,δEL): beam scan angle [rad]

Foff =F/cos2(θoff/2): distance between the aper-ture center and the focal point,

F: focal length of the reflector,

Df: defocus distance, and

θoff : reflector offset angle.

Equation (2) indicates that the phase shiftdoes not depend on the initial direction andthat beams with different directions can bescanned jointly. The phase shift with respectto a unit scan angle becomes Foff /Df times thephase shift of the direct radiation array.

In a phased-array-fed reflector antenna, asthe scan angle from a boresight increases,directivity is generally degraded due to aberra-tion, a phenomenon not seen in a direct radia-tion array. In some cases the scan angle usedin Eq.(2) (referred to as the “steering angle inEq.(2)”) may not necessarily agree with theactually obtained scan angle. Furthermore, fora shaped antenna beam that has been formedunder gain constraint (in order to secure suffi-

cient area gain and beam isolation), it is diffi-cult to define the beam scan angle using thedirection in which the antenna gain is maxi-mum. In this paper, desired scan angles (δAZ’,δEL’) are defined using an area in whichthe gain constraints are imposed, and thesteering angles (δAZ’,δEL’) to be substitutedinto Eq.(2) are determined as follows.

1) Areas of constraint are defined withinwhich area gain or beam isolation must beensured, with array weight optimizedaccordingly (referred to as the “initial arrayweight”). The obtained excitation phase isdesignated as ηn (n = 1 to N).

2) All constraint areas are translated bydesired scan angles (δAZ,δEL) [rad], andthe excitation phase η’n is obtained similar-ly as in step 1). The difference in excitationphase before and after scanning is designat-ed as δηn = η’n-ηn.

3) Determine the phase gradients Nx, Ny to bethe best fit to the distribution of the phasedifference of δηn on the feed plane by min-imizing the square phase deviation ε2.Here, in order to avoid the phase ambiguityof 2π,ε2 is defined by the following for-mula (the range of the function arg (*) is setbetween -πandπ).

ε2=Σarg2 [exp(jk(Xn･Nx－Yn･Ny+C)－jδηn)] (3)

4) The steering angle to be substituted intoEq.(2) is determined as follows.

δAZ=Df／Foff･sin-1(Nx),δEL=Df／Foff･sin-1(Ny) [rad] (4)

Next, numerical calculation was per-formed based on the specifications of theantennas equipped on ETS-VIII. Since theexcitation amplitude distribution for transmis-sion is a specified constant regardless of thedirection of beam pointing, only the phaseoptimization was carried out. The conditionsof constraint consist of an area gain of 42 dBior more and a beam isolation of 27 dB ormore (equivalent to gain of 15 dBi or less).Although the specification value for beam iso-


lation of the transmitting antenna is 20 dB ormore, the constraint condition was set morestrictly in order to allow for degradation dueto excitation errors (weight optimization isperformed based on the assumption of noexcitation errors).

Fig.4 a) shows an example of the obtainedantenna pattern. In the figure, the five pointsin the Tokai area indicate constraint points for

area gain, and the five points in the Kyushuarea indicate the constraint points for beamisolation. The gain contour line of 42 dBi cor-responds to the constraint value for area gain,and the contour line of 15 dBi corresponds tothe constraint value for beam isolation. Fig.4b) shows a pattern optimized by translating allconstraint points, with the desired scan angleset to 0.5 degrees in the EL direction. Fig.4 c)shows the pattern obtained by scanning thebeam using Eqs. (2) and (4) on the initialarray weight. Comparing Fig.4 b) with Fig.4c), the contours corresponding to the area gainand beam isolation in Fig.4c) nearly agreeswith those in Fig.4b). In both Fig.4 b) and c),the areas in which the required beam isolationis obtained become narrower than the corre-sponding area in Fig.4 a). This is attributed tothe increase in the scan angle from the bore-sight of the main lobe. In actual geostationaryorbit, there is little probability that antennapointing error will exceed 0.5 degrees due tosatellite attitude error and/or thermal deforma-tion of the reflecting surface.

Fig.5 shows the relationship between thedesired scan angle (to the gain constraintpoint) and the steering angle obtained with Eq.(4). The steering angle from Eq. (4) and thedesired scan angle are proportional with aratio of 1:1.1. Nearly the same results wereobtained for beam scanning from other initialbeam positions.


Calculated transmitting antenna pat-terns. Marks in Figs. (a) and (b) indi-cate the gain constraint.

Fig.4Steering angle defined by Eq.(4) vs.desired steering angle.

Fig.5

78

Fig.6 shows the relationship between rmsphase deviation ε(defined by Eq. (3)) and thedesired scan angle. Phase deviationεincreas-es with the scan angle, and becomes 12 to 22degrees rms for a scan angle of 0.5 degrees.When a random phase error of this amountarises in the feed array, large degradation isexpected, especially in terms of beam isolation[12]. The results shown in Fig.4c) correspondto the case of maximum phase deviation (inthe +EL direction, with a scan angle of 0.5degrees) shown in Fig.6. Significant degrada-tion is not found in the scanned beam. There-fore, we conclude that phase deviation causedby the beam scanning method presented in thispaper does not affect degradation of directivi-ty to the same extent as random phase errorhaving the same rms value.

4 Development of onboard equi-pment

4.1 Beam-forming network 1The structure of the transmitting BFN1 is

shown in Fig.2. Beam port 1 corresponds tothe independently controlled beam, and beamports 2 and 3 correspond to the uniformly con-trolled beams. An input signal to the beamport is divided into thirty-one portions, whichare subject to the required array weight (atten-uation and phase sift) and amplification for

subsequent output to the respective elementports. The attenuation and phase shift are per-formed by an MMIC active module in which a5-bit variable attenuator, a 6-bit variable phaseshifter, and an amplifier (to compensate forinsertion loss) are integrated in a single pack-age [13][14]. Table 1 shows the specificationsof the module. Each MMIC module is mount-ed on a space-grade multilayer polyimideboard. The MMIC modules are mounted onthe front surface of the board with micro-stripe RF lines. On the rear surface, interfacecircuits between the MMIC modules and BFCare formed. Note that beam 2 uses MMICmodules that can be pre-set upon equipmentstartup so that the initial array weight can bemodified. Table 2 shows the major specifica-tions of the BFN1 (common to the transmit-ting BFN1 and to the receiving BFN1) and thespecifications of the BFC.

4.2 Beam-forming controllerThe beam-forming controller (BFC) calcu-

lates the array weight, controls the MMICmodules in the BFN, and interfaces with thetelemetry and tracking control subsystem(TTC) of the satellite. Table 3 shows the oper-ating modes and main functions of the BFC.

The test mode is used mainly to confirmthe functions of the MMIC modules. In oper-ating modes other than the test mode, the BFCprovides amplitude and phase values for tem-perature compensation, based on the tempera-tures of several components of the feed sec-


Phase deviationεvs. the desired steer-ing angle.

Fig.6

Specifications of the active MMICmodule for BFN1.

Table 1

Specifications of Tx-BFN1 and BFCTable 2

tion, and converts these values into control bitpatterns for the MMIC modules. In this con-version, control bit pattern for a variable-attenuator is determined, to compensate theinsertion loss of a variable phase shifter that isdependent on the phase shift. The determinedbit patterns are buffered sequentially in theinterface circuits of the MMIC modules in theBFN, and finally the settings of all the mod-ules are updated simultaneously by an updatecommand. The transition time during updateis 10 ns or less, significantly shorter than thesymbol duration of a signal. Therefore, beampointing control is performed without inter-ruption in communication link and without theinter-beam interference due to the deformationof the antenna beam during the above transi-tion time.

The normal mode is used communicationand broadcasting experiments. In this mode,phase control Eq.(2) enables independentscanning for beam 1 and uniform scanning forbeam 2 and beam 3. Program scan mode runsthese scanning operations repeatedly at inter-vals of 0.2 sec. This mode is used to evaluateantenna patterns in orbit and to simulate hand-over between beams in mobile communicationexperiments. Although the steering angle inEq.(2) is based on the antenna coordinate sys-tem in Fig.3, it is more convenient in actualoperation to denote east and north as AZ andEL, respectively, with reference to the satel-lite’s nadir. Since the onboard antenna isinstalled on the satellite so that the boresight(+Z of the antenna coordinates) may point tothe central part of the Japanese islands (135E,34.67N), the BFC calculates phase shift byconverting the coordinate. Moreover, the BFCalso calculates the scan angle required to com-pensate for the attitude error extracted fromthe satellite’s telemetry data.

The REV (rotating element electric fieldvector method [15]) measurement mode isused to evaluate the feed section in orbit.Since the REV method does not require phasemeasurement of a received signal, it is suitablefor evaluation of satellite phased array anten-nas in orbit [16]. However, after the satellite islaunched, time averaging of received signalamplitude is needed to improve the measure-ment accuracy; therefore, it is necessary toperform phase rotation at sufficient time inter-vals. The REV mode is convenient especiallyfor such in-orbit evaluation, as the excitationphase of an arbitrary element can be rotatedby a specified phase step at specified timeintervals. The in-orbit programming mode isused in the event of a fault after launch orwhen modifying experimental functions of theBFC.

4.3 Evaluation of array weight control In order to verify the onboard temperature

compensation function, a thermal test wasconducted for the beam-forming network andthe beam-forming controller. During the test,weight error was measured by changing thecomponent temperature. According to theresults of the thermal test of a BFN withouttemperature compensation, when the tempera-ture (in this case, the temperature of the baseplate the component) rose from －10℃to＋40℃, the insertion loss increased by about 4dB. On the contrary, the results with the com-pensation shown in Fig.7 show that variationin insertion loss with a temperature changefrom －10℃to＋40℃ was within 1 dB.Moreover, among the thirty-one elements,amplitude and phase deviations were 0.4dBrms or less and 4 degrees or less, respec-tively, which clearly demonstrated the effec-tiveness of the temperature compensationfunction of the BFC.

5 Evaluation of beam-scanningperformance

To verify the performance of a reflectorantenna, measurement of secondary radiation


Operating modes and functions ofBFC

Table 3

80

patterns is generally the most direct way.However, it is difficult to measure the second-ary pattern on the ground due to the influenceof terrestrial gravity on the surface accuracy ofthe reflector, particularly given the reflectorsize of the ETS-VIII antenna (19 m in outerdiameter). We therefore divided the verifica-tion process into two steps, as follows.

The first step involved evaluation using aBBM (breadboard model) for the BFN and thefeed array. In this evaluation, the antennareflector was not used. First, the primary elec-tric field radiated by the feed array was meas-ured by the near-field measurement method.Next, by assuming the shape of the antennareflector (in this case, an ideally parabolic sur-face was assumed), electric-field distributionon the reflector surface was calculated, andthen the secondary radiation pattern was eval-uated [9]. Although in the BBM some of thestructural parameters of the feed element, theexcitation amplitude distribution, and thereflector structure were different from those ofthe flight model [9][17], the BBM is sufficientto verify the formation and pointing control ofthe antenna beams. Here it should be notedthat the phase center of a cup MSA element is

not on the feed patch but rather is located onthe aperture plane of the metal cup, based onour measurement results. Therefore, the defo-cus distance is defined in terms of the distancefrom the reflector focal plane to the cup aper-ture plane.

Figs.8 a), b) show the estimated secondaryradiation patterns with weights optimized forbeam positions #3 and #4 shown in Fig.1,respectively. Figs.8 c), d) show patternsobtained with uniform scanning of thesebeams before and after the steering of -0.5degrees in the AZ direction, respectively. Thecircles in the figure indicate desired beampositions before and after scanning, and dottedlines show measurement results as contourlines of 43, 42, and 20 dBi. The results of uni-form scanning show that the desired scanangle was obtained, and that the required areagain (42 dBi) was obtained. Note that all ofthe 42-dBi contours are shifted from thedesired positions by about 0.2 degrees towardthe -AZ side. It is possible that this is attribut-able to alignment error in the feed array at thetime of primary pattern measurement. On theother hand, although the specification forbeam isolation (>20 dB) imposes a gain con-dition of 22 dBi or less in the areas of sharedfrequencies, the gain contour line of 20 dBiafter the scan was slightly extended, indicatingthe presence of areas of deteriorated beam iso-lation. We conclude that this was caused byexcitation error in the feed system and by themeasurement errors in the primary radiationpattern, in addition to the phase deviationoccurring through the beam scanning men-tioned as shown in Fig.6.

In the next stage, secondary-pattern meas-urement was performed by combining an EM(engineering model) of the reflector and themajor components of the feed section [18].This measurement is intended (1) to confirmthe total performance of the feed componentsin a configuration identical to that of the flightmodel, and (2) to verify the validity of antennapattern evaluation based on the measured val-ues of the geometrical shape of the reflectorsurface. The specifications and arrangement


Measured weight error of TX-BFN1.(a) Amplitude error, (b) Relative phaseerror●T=-10deg., ■T=22.5deg., ▲T=40deg.

Fig.7

of the EM components used are basically thesame as those of the flight components exceptthat the large deployable reflector was con-structed with only seven modules of the totalfourteen modules to be used in the flight con-figuration due to the size constraints of theradio anechoic chamber.

The reflector shape was measured optical-ly, then by combining the reflector with thefeed section, the secondary radiation patternswere measured with near-field measurementmethod. The results of this measurementagree well with the predicted results based onthe geometrical shape of the reflector surface[18]. Consequently, we conclude that the

validity of the presented evaluation method oflarge reflector antennas has been demonstratedand that the performances of the feed systemhave also been verified.

6 Conclusion

This paper described the BFN1 and theBFC, both of which will be used in thephased-array-fed reflector antennas equippedon ETS-VIII satellite. We have shown thatthese devices have sufficient functions andperformances to set the array weights for theformation and pointing control of the antennabeams.


Measured patterns before/after the steering of –0.5 degrees in AZ direction.Circles indicate the desired beam position.

Fig.8

82 Journal of the National Institute of Information and Communications Technology Vol.50 Nos.3/4 2003

References1 E. H. Kopp, "Handheld telephony from geosynchronous orbit ", 17th ICSSC, AIAA-98-1396, 1998.

2 S. C. Taylor and A. R. Adiwoso, "The Asia cellular satellite system", 16th ICSSC, AIAA -96-1134, 1996.

3 Y. Kawakami, S. Yoshimoto, Y. Matsumoto, T. Ohira, and N. Hamamoto, "S-band Mobile satellite communi-

cations and multimedia broadcasting onboard equipments for ETS-VIII ", Trans. IEICE Vol. EB82-B, No. 7,

pp.1659-1666, 1999.

4 K. Tokunaga, K. Ueno, Y. Matsumoto, Y. Suzuki, N. Hamamoto, and M. Okumura, "Design of phased array-

fed reflector antennas for the Japanese Engineering Test Satellite 8 ", 50th International Astronautical Con-

gress, IAF-99-M.3.06, 1999.

5 R. J. Mailloux, "Phased array theory and technology ", Proc. of the IEEE, Vol. 70, No.3, pp. 246-291, 1982.

6 Y. Matsumoto, Y. Hashimoto, T. Ide, M. Sakasai, N. Hamamoto, and M. Tanaka, "Beamformer with a single

set of variable phase shifters for the pointing control of multibeam satellite antenna", Trans. IEICE, Vol. J80-

B-II, No. 7, pp. 617-621, 1997.

7 T. Ohira, Y. Suzuki, H. Ogawa, and H. Kamitsuna, "Megalithic microwave signal processing for phased-array

beamforming and steering ", IEEE Trans. Microwave Theory and Tech., Vol. 45, No. 12, pp. 2324-2332,

1997.

8 K. Nakamura, Y. Tsutsumi, K. Uchimaru, T. Tsujihata, and A. Meguro, "Large Deployable reflector on ETS-

VIII ", 18th ICSSC, AIAA-98-1229, 1998.

9 K. Tokunaga, H. Tsunoda, Y. Matsumoto, and T. Ohira, "Electric design and prototype development of

phased array feed for onboard reflector antennas", Trans. IEICE, Vol. J82 B, No. 7, pp. 1357-1365, 1999.

10 Y. Matsumoto, M. Tanaka, and M. Orikasa, "Development of cup-MSA feed array element for phased array

satellite antennas", Trans. IEICE, Vol. J82 B, No. 7, pp. 1420-1424, 1999.

11 S. Hama, Y. Matsumoto, S. Taira "GEO based mobile communication of the near future -BFN and packet

switch on the ETS-VIII-", Space Communications 15, PP. 55-63, 1998.

12 K. Tokunaga, H. Tunoda, Y. Matsumoto, and T. Ohira, "Feed development and secondary radiation charac-

teristics of phased array fed reflector antennas for satellites", IEICE Technical Report, SANE97-65, 1997.

13 C. Sakakibara, H. Takasu, M. Okumura, S. Kamihashi, S. Hama, and Y. Matsumoto, "S-band MMIC phase

shifter module for beam forming network ", IEICE 1998 General conference, C-2-37, 1998.

14 H. Takasu, C. Sakakibara, M. Okumura, S. Kamihashi, Y. Matsumoto, and S. Hama, "S-band MMIC active

module with small phase variation and low insertion loss for beam forming network ", Trans. IEICE vol. E83-

C, No. 1, pp. 122-125, 2000.

15 S. Mano and K. Katagi, "A method for measureing amplitude and phase of each radiating element of a

phased array antenna", Trans. IEICE, Vol. J65-B, No. 5, pp. 555-560, 1982.

16 T. Tanaka, Y. Matsumoto, S. Yamamoto, K. Suzuki, and Y. Arimoto, "Antenna pattern measurement of S-

band active phased array on ETS-VI ", Vol. J80 B-II, No.1, pp. 63-72, 1997.

17 Y. Matsumoto, T. Ide, M. Tanaka, T. Orikasa, and Y. Yamasa, "Array element and BFN design for phased-

array-fed satellite antennas ", 17th AIAA International Communication Satellite Systems Conference, 98-

1224, 1998.

18 K. Ueno, M. Atobe, A. Miyasaka, M. Orikasa, and M. Okumura, "Ground test of array-fed reflector antennas

for ETS-VIII ", IEICE Technical Report, A・P2000-155 /SANE2000-136, 2001.


MATSUMOTO Yasushi, Dr. Eng.

Associate Professor, Research Instituteof Electrical Communication, TohokuUniversity (present Leader, Communi-cation System EMC Group, WirelessCommunications Division)

Wireless Communications

IDE Toshiyuki

Senior Researcher, Mobile SatelliteCommunications Group, WirelessCommunications Division

Wireless Communications

84 Journal of the National Institute of Information and Communications Technology Vol.50 Nos.3/4 2003

3-6-3 beam forming network

Documents