1 ΐlvXbNnVXe A E]K瑏 ݁An[_𗘗pݒn̉@̌ɍsĂB̃O[v܂łɁAfȑgD}bv(Complex-valued self-organizing map: CSOM)𗘗pVn[_𗘗pVXeĂĂB̃VXe̊Tv́Aԗ̈您юg̈̕feNXɒڂĂ̓ʂ𒊏oA𕡑fȑgD}bvɂK敪āAn݉ӏ̂łB܂łɁA܂܂ȍȏꍇł̉ɐĂBAԃeNXƎgeNXƂ2̈قȂ郂[h̏񓝍̂ɂẮAꂪKœKĂ킯ł͂ȂBNx́A擾ꂽf[^玩ȑgDIɒn̊TO\zĂ䂭ƓɁA2̈قȂ郂[h̏𓝍Ă䂭@ĂBTO\z@ƂāAݏʂɂCꂽȑgD}bvĂB 2 ~gE}CNgC[WO A EfA}h@_E쓇 ANeBuȃ~gC[WÓAL͂ȉp҂ĂBA~gH͈ʓIɑ傫A܂Av͍ɂȂ肪łāAɃACɍۂĖɂȂBᑹňȃ~gtgGh邽߂ɁAEPDȂǂĂĂB́AAeiŎMシɌgŝł~gAvAHŜƂčx]܂B{_ł́AtBoNe[pXbgAeiibulk TSAjp鍂xȃ~gC[WOptgGhĂBTSACs[_Xł邽߁AMp[傫łBɂAtB1mmTSApEPDł́A]^tBɂTSAɔׂ22dBxꂽBtgGh̓𐔒lIɉ͂B܂AC[WOVXepɃACۂ̍\V~[Vɂ茟AX^bN^ʌ^̕AC\[V邱ƂmFꂽB 3 J[_ A EďH@ lHqq@̂悤ȁAڂłAeĩTCYɐō\𓾂邽߂̋ZpƂāAJ[_(SAR) BɊJ[_(InSAR) ́AnkΎRɂnϓȂǂϑۂ̏dvȋZpłAߔNϋɓIɌȂĂBInSAR̊摜́An2ӏȂ͎Ԃu2ϑēꂽA master slave Ƃ΂2̕fUf[^pÄ݂ʑ邱ƂɂēB̊摜̎邱ƂŃfW^n}(Digital Elevation Model: DEM) BW͖{ۑł͂Aۂ̊摜ɂ͈ʑٓ_(Singular Point: SP)ƌĂ΂]n_݂Am DEM̍쐬ɂ[ȏQƂȂĂB{ŃA萳mȊ摜쐬@ƂĊ SP ̐wWɂVʒu킹ƁAUɊR̈ʒu𐄒肵ÅRČ̂ɕKvSPcĂB̓Iɂ́Aʑٓ_wWƂʒu킹@AShape-from-ShadingpU񂩂̊R̐ĂBĎ@p摜AbvDEMSNɂ́APF߂ꂽB 4 A_veBu Aei VXe A E͑Y ړ̒ʐMŃA_veBuAeinǂɗpꍇAnǂ͎MAbvNM烆[U[̕𐄒肵ăAei׏d肵AԕdłB̍ہAɎdʐMł́AAbvNMMɂČ肵׏d_ENɂpł\B̂߂ɂ̓[Üړ̗\KvƂȂ邪A_ENԂ߁A܂ŏ\Ȑ\𓾂邱ƂBNx܂łɂ́A[vzϊ(CZT)𗘗pAmȗ\\ŌvZʂȂVȎ@ĂĂB̌vZʂ́A[ߍݖ@⎩ȉA@A\@Ȃǂ菬DĂBNx͂ɁAg̈ŎԓIȈꎞsƂɂA\̐\コ邱ƂɐB̎@́Ag̈ł̉͂̍xƎԗ̈ł̌ʓIȗ\Ƃ̗𕹂łB{ɂāAÛڋ߂ĈʒuĂꍇłAxŃl\ł邱ƂB 5 @\Aeifq A EFi[@VetFEnMO ACAei\ꍇA]̖_Aei͑݌傫A܂܂ȃVXeœ̑傫ȗ򉻂NĂB̖͂邽߁AAeifqԂɓʂɐ݌vꂽUd̕ЂuƂĂBɂď]^20dB̑݌ጸ}邱ƂłB̌ʂ́Aш15%ŗLłƂLшB̓Aeifq̐P邱ƂƂ֌WĂB܂IȂƂƂāAACAeiƂĂ̕˃p^[e󂯂AđMɉeyڂɁA݌݂̂ጸʂł邱ƂmFA̗LpƂłB 6 tNVi uE^̐Ɨp A E_kfBW hj[ tNVi uE^(fractional Brownian motion: fBm)́ARÊ܂܂ȌۂLq郂fƂėLłA̐ɂĂ܂܂ȌsĂBNx2邢͂ȏ̎fBmVȐ@ƂāA񖄂ߍ݂ɂ鐶ASYĂB񖄂ߍ݂ɂẮA1fBm̏ꍇɂ͂łɕ񍐗Ⴊ邪A2Ɋg邱Ƃ͂܂ŊȒPł͂ȂƍlĂB͂AŏIڕWfBm2̍l邱Ƃɂ胉vXg̈ŉɋAA𓾂邱ƂɐB2ȏ̑ɗeՂɊg\ȂƂƂẢ@̗_łBlɂA_vʂ̐fBmႢvZRXgŐ邱ƂB 7 j[lbg[Ñ_Ci~NX̊b A EgcYE w} j[lbg[N̍łIŏdvȋ@\̈́AĉłB{éAfj[lbg[NƎj[lbg[N̔ĉrBɁAfj[lbg[NgIȃf[^ۂ̔ĉAM̔gɂĂǂ̂悤ɕω邩Bfє{̎j[lbg[NƂătB[htH[h^wj[lbg[NグBȌȗƂāA炪ԐMɑ΂Ċ֐ߎ邢̓tB^IȓƂ̔ĉ𒲂ׂBvZ@̌ʁAfj[lbg[N͔{j[lbg[NɔׂāA̔ĉǍDłAɐM̔gꍇɂ̍ł邱ƂmFꂽB
 1 Antipersonnel Plastic Landmine Visualization System A.Hirose and A.Ejiri The complex-valued self-organizing map (CSOM) realizes an adaptive distinction between plastic landmines and other objects in landmine visualization systems. Previously we developed our walled linearly tapered slot antenna (walled LTSA) to construct a high-density array. We conducted field experiments to confirm the high ability to visualize plastic landmines buried in laterite soil. This year we proposed a concept generation method using multiple self-organizing maps (SOMs) for a landmine visualization system where we integrate two modal information (space- and frequency- domain texture) by considering their mutual information. Our method makes clusters of landmine information in their space- and frequency-SOMs that interact with each other based on their mutual information. 2 Millimeter wave and microwave imaging for antiterrorism screening A.Hirose, D.Radenamad and S.Onojima Millimeter-wave (MMW) active imaging has potentially a variety of applications. However, MMW circuits are often lossy and amplifiers are costly, which is a serious problem in particular in arrays using many front-end circuits in parallel. To overcome this point, there have been some proposals such as the envelope phase detection (EPD) method where MMW is directly detected without amplification just after received at the antenna. To realize a high sensitivity, this paper proposes a MMW imaging front-end using a bulk tapered slot antenna (bulk TSA). Since bulk TSAs possess an intrinsically low impedance characteristic, we can obtain a high detection power. Experiments demonstrate that a front-end using 1mm-thick TSA achieves 22dB-higher sensitivity than a conventional circuit using 0.06$\mu$m TSA. We analyze the front-end numerically to elucidate its characteristics. We also investigate array configurations to find that a coplanar arrangement exhibits higher isolation than a stacked configuration. 3 Synthetic Aperture Radars A.Hirose and R.Natsuaki Theoretically, synthetic aperture radar (SAR) interferogram should be a conservative field which mirrors the geologic distribution. However, actual interferograms contain a mass of rotational points named singular points (SPs). They prevent creating accurate digital elevation model (DEM). We think that many singular points indicate the local distortion in the interferogram. To solve this problem, we proposed local coregistration and phase correction by using amplitude information. The signal-to-noise ratio has increased by this method. 4 Adaptive Antenna System A.Hirose and S.Ozawa Channel prediction is one of the most important issues for dealing with fading environment. If a future channel response is predicted, adaptive techniques such as pre-equalization and beam forming can be employed before transmission in order to avoid the degradation of communication quality. This year we proposed a highly accurate method to predict time-varying channels by using a linear prediction of Fourier-domain channel parameters. The method combines frequency-domain and time-domain techniques. Simulation assuming a time division duplex (TDD) system demonstrates that the proposed method predicts varying channels very accurately. The paper also presents the dependence of the prediction error on the length of observation window and on the distance between scatterers. It is found that our method reduces the prediction error even when the scatterers are located very closely. 5 Highly functional antenna elements A.Hirose, S.Werner and T.Baba Conventional rod antenna arrays exhibit considerable mutual coupling, which limits or degrades the performance of many systems. In the scope of this work we propose a dielectric particle that is placed between the antenna elements and lower the mutual coupling by more than 20dB compared to conventional arrays. The designed particle shows a broadband isolation effect that has a fractional bandwidth of more than 15%. Moreover, the particle can provide an enhanced matched environment and a largely maintained radiation pattern as well as gain compared to a conventional array. 6 Generation and application of fractional Brownian motion A.Hirose and D.Danudirdjo Fractional Brownian motion (fBm) is a useful model to represent various natural phenomena and its synthesis has been a topic of interest in literature. We presented an algorithm to generate two-dimensional fBm based on circulant embedding method. Although this method has been proven in simulating one-dimensional fBm, its extension into two-dimensional is not straightforward. To solve the problem, we use circulant embedding as an exact method to synthesize the second-order increments of fBm, and recover the desired fBm via integration in frequency domain. The advantage of using second-order increments is its direct extensibility to any dimension. Experimental results show that the proposed method works efficiently fast and offers acceptable results when compared to the theoretical statistics of fBm. 7 Fundamental analysis of neural network dynamics A.Hirose, S.Yoshida and Gema M.F.Merchan Generalization is one of the most important functions of neural networks. We compared the generalization characteristics in complex-valued and real-valued neural networks. Assuming that a network deals with wave-related data, we paid attention to the generalization variation dependent on the degree of wave nature. We considered feedforward layered neural networks in complex and real domains. To elucidate the operation difference in the domains, we investigated a simple case where the networks function for function approximation or filtering of temporal signals. Computer experiments demonstrated that the generalization ability of the complex-valued neural network is better than that of real-valued one in particular when the signal possesses higher degree of wave nature. _Eꗗi2010Nxj Publications List _ [1] A.Hirose, "Nature of complex number and comlex-valued neural networks," Frontiers of Electrical and Electronic Engineering in China, 6, 1 (2011) 171-180 [2] R.Natsuaki A.Hirose, "SPEC method - A fine co-registration method for SAR interferogram," IEEE Trans. on Geoscience and Remote Sensing," 49, 1 (2011) 28-37 [3] Y.Nakano A.Hirose, "Adaptive identification of landmine class by evaluating the total degree of conformity of ring-SOM," Australian Journal of Intelligent Information Processing Systems, 12, 1 (2010) 23-28 [4] D.Radenamad T.Aoyagi A.Hirose, "Low impedance bulk LTSA," Electron. Lett., 46, 13 (2010) 882-883 ۉc_ [5] S.Werner L.Vietzorreck A.Hirose, "Two-element particle isolator to reduce mutual coupling in rod-antenna arrays," Asia-Pacific Microwave Conference (APMC) 2010 Yokohama, Proc. (Dec. 7-10, 2010, Yokohama) 1581-1584 [6] T.Baba A.Hirose, "Planar wideband adaptive antenna consisting of radially arrayed multiple taper-slot antenna elements having wide fins," Asia-Pacific Microwave Conference (APMC) 2010 Yokohama, Proc. (Dec. 7-10, 2010, Yokohama) 738-741 [7] S.Ozawa S.Tan A.Hirose, "Channel Prediction Experiment Based on Linear Prediction in Frequency Domain," Asia-Pacific Microwave Conference (APMC) 2010 Yokohama, Proc. (Dec. 7-10, 2010, Yokohama) 1280-1283 [8] D.Radenamad A.Hirose, "Optimization of Microstrip-to-slot Transition for Ultra-wideband Bulk LTSA," Int'l Conf. on Antennas and Propagation (ISAP) 2010 Macao, Proc. (Nov., 23-26, 2010, Macao) Paper 122 [9] Y.Nakano A.Hirose, "Adaptive identification of landmine class by evaluating the total degree of conformity of ring-CSOM weights in a ground penetrating radar system," Int'l Conf. on Neural Inform. Processing (ICONIP) 2010 Sydney, Proc. (Nov., 22-25, 2010, Sydney) Th10.30-Rm38 [10] A.Hirose A.Ejiri K.Kitahara, "Ground penetrating radar system with integration of mutimodal information based on mutual information among multiple self-organizing maps," Int'l Conf. on Neural Inform. Processing (ICONIP) 2010 Sydney, Proc. (Nov., 22-25, 2010, Sydney) 415-422 [11] S.Ajisaka S.Nakamura H.Hashimoto K.Takiguchi A.Hirose, "Human Body Position Estimation System using Electric Field," Int'l Conf. on Control, Automation, and Systems (ICCAS) 2010 Goyang, Proc. (Oct. 27-30, 2010, Goyang) TA0-5 [12] S.Nakamura S.Ajisaka H.Hashimoto K.Takiguchi A.Hirose, "Electric-field resonance coupling between human and transmitter for human position estimation system," Int'l Conf. on Control, Automation, and Systems (ICCAS) 2010 Goyang, Proc. (Oct. 27-30, 2010, Goyang) TA0-2 [13] A.B.Suksmono A.Hirose, "Numerical reconstruction of holographic microscopy images based on matching pursuits on a pair of domains," Int'l Conf. on Image Processing (ICIP) 2010 Hong Kong, Proc. (Sept. 26-29, 2010, Hong Kong) WP.L5.8 [14] R.Natsuaki A.Hirose, "Enhancement of filtering performance by local and nonlinear co-registration in InSAR processing," Asia-Pacific Radio Science Conference (AP-RASC) 2010 Toyama, Proc. (Sept. 22-26, 2010, Toyama) FP-20 [15] S.Ozawa S.Tan A.Hirose, "Errors in Channel Prediction Based on Linear Prediction in Frequency Domain (Young Scientist Award Finalist)," Asia-Pacific Radio Science Conference (AP-RASC) 2010 Toyama, Proc. (Sept. 22-26, 2010, Toyama) FBC-4 [16] T.Aoyagi D.Radenamad Y.Nakano A.Hirose, "Complex-valued self-organizing map clustering using complex inner product in active mmillimeter-wave imaging," Int'l Joint Conference on Neural Networks (IJCNN) 2010 Barcelona, Proc. (July 18-23, 2010, Barcelona) 1346-1351 (Runner-up Best Student Paper Award) [17] Y.Nakano A.Hirose, "Taper-walled LTSA: A low direct-coupling antenna for subsurface imaging," Int'l Conf. on Ground Penetrating Radar (GPR) 2010 Lecce, Proc. (June 21-25, 2010, Lecce) PR_248 [18] A.Hirose, "Recent progress in applications of complex-valued neural networks (invited)," Int'l Conf. on Artificial Intelligence and Sof Computing. (ICAISC) 2010 Zakopane, Proc. (June 13-17, 2010, Zakopane) Invited Talk [19] A.Hirose, "Merits of complex-valued neural network processing (invited)," Int'l Workshop on Intelligence Science and Intelligent Data Engineering (ISciDE) 2010 Harbin, Proc. (June 3-5, 2010, Harbin) Invited Talk Session 2 [20] A:ufj[lbg[N(W)vAdCw_CA131, 1 (2011) 2-8. V|WEEE\ [21] ͑Y@^E\t@A: tF[WOɂlp[^̐\AMwdE_AEMT10-103 (Nov. 11, 2010, Inawashiro) [22] ďH@AFʑٓ_ƐU𗘗pSAR摜̏CAMwdE_AEMT10-121 (Nov. 11, 2010, Inawashiro) [23] nMO@AFLtBɂ镽ʌ^e[pXbgLшwAei̒āAMwdE_AEMT10-124 (Nov. 11, 2010, Inawashiro) [24] fA}hE_@AF}CNXgbvƃXbgȂ2̃rAɂLшoNLTSÂ߂̃o̒āAMwdE_AEMT10-125 (Nov. 11, 2010, Inawashiro) [25] A@gcYFgIȃf[^ۂ̕fj[lbg[N̔ĉAMwj[Rs[eBOANC2010-179 (March 8, 2011, Tokyo) [26] ]K瑏@kMI@AFݏʂ𗘗pSOMɂTOAFANV|WEAOS16-17 (Sept. 26, 2010, Tokyo) [27] ͑Y@^E\t@AFg̈ł̃lp[^\ɊÂl\̎IA2011MwAB-1-31 (Mar. 14, 2011, Tokyo) [28] nMO@AFˏɔzuꂽǌ^tBɂ镽ʍLшwAeiA2011MwAB-1-216 (Mar. 17, 2011, Tokyo) [29] ďH@AFUpSAR摜̋ǏIʒu킹ɂc݉A2011MwAB-2-41 (Mar. 16, 2011, Tokyo) [30] A.Hirose Y.Nakano, "Applications of complex-valued self-organizing maps to ground penetrating radarimaging systems," Self-organizing maps: New achievements, In-Tech (January 2011) 323-338, ISBN 978-953-307-546-4 ̑ [31] A.Hirose, "Neural network systems for intelligent imaging," Japan-Taiwan Cooperation Meeting at National Chiao Tung University (Oct. 27, 2010, Hsinchu) [32] D.Mandic I.Aizenberg A.Hirose, "Complex-valued neural networks: Theory and applications (Tutorial)," Int'l Joint Conference on Neural Networks (IJCNN) 2010 Barcelona, Proc. (July 18-23, 2010, Barcelona) Tutorial T-21 [33] A@Fu܊v񏈗Ej[lbg[Nƃ}CNGNgjNXA37STARCAhoXgu@VXeA[LeNZ~i[@ ٕɊw VXeA[LeNZpJ̏ ` (July 15, 2010, Yokohama) [34] A.Hirose, "Neural networks and microelectronics" Japan-Taiwan Cooperation Meeting at NCKU(April 21, 2010, Tainan) [35] A.Hirose, "Neural networks and microelectronics," Japan-Taiwan Cooperation Meeting at NCTU (April 20, 2010, Hsinchu) [36] A.Hirose, "Neural network systems for intelligent imaging," Japan-Taiwan Cooperation Meeting at National Chiao Tung University (Oct. 27, 2010, Hsinchu) [37] A@TF摜uyщ摜vOAoA2010-157514 [38] T.Aoyagi D.Radenamad Y.Nakano A.Hirose, "Runner-up Best Student Paper Award, for "Complex-valued self-organizing map clustering using complex inner product in active mmillimeter-wave imaging," Int'l Joint Conference on Neural Networks (IJCNN) 2010 Barcelona (21 July 2010), IEEE Computational Intelligence Society (CIS) [39] S.Ozawa S.Tan A.Hirose, "Student Paper Competition Finalist for "Errors in Channel Prediction Based on Linear Prediction in Frequency Domain," URSI Asia-Pacific Radio Science Conference (AP-RASC) 2010 Toyama, (10 September 2010) Union