We propose a learning algorithm for self-organizing neural networks to form a topology preserving map from an input manifold whose topology may dynamically change. Experimental results show that the network using the proposed algorithm can rapidly adjust itself to represent the topology of nonstationary input distributions.

An intrinsic property of a tapped resonator is elucidated here, and a novel bandpass filter (BPF) with improved skirt characteristics based on a tapped half-wavelength resonator is proposed by this intrinsic property. "Tapping" for both I/O and interstage couplings of the resonator is the key concept here because a resulting open-ended resonator makes shunt open stubs which give anti-resonance near the center frequency. Multiple attenuation poles appear near the center frequency, namely, close to the passband. A BPF is designed on the basis of the general filter theory with a narrow band approximation. An experiment is carried out to confirm the concept by using a coplanar structure. The expected bandpass characteristics with multiple attenuation poles have been obtained by the novel BPF designed by the present concept.

The onboard antenna beam forming network (BFN) of the next-generation communication satellites must offer multiple beam forming and beam steering. The conventional BFN, which directly controls the array elements, is not suitable for a large-scale array antenna because of the difficulty of BFN control. This paper proposes a new BFN configuration that consists of three/four-way variable power dividers and a Butler matrix (FFT circuit). This BFN can offer continuous beam steering with fewer variable components. By introducing new techniques based upon excluding FFT periods and power evaluations by definite integration, the deviation in beamwidth is reduced by 75% or more and the maximum sidelobe level is improved by 10 dB or more.

One effective way to make a large-capacity mobile communication system is to use a microcellular system. One way to do this is to place the base station antennas lower than the surrounding buildings. This result in what is called a street microcellular system. We previously presented basic simulation results evaluating the performance of a different-sized cell combination algorithm (DCC) designed to avoid the problems due to an unbalanced distribution of traffic in street microcellular systems. In this paper, we present the results of simulations evaluating the performance of an improved active different-sized cell combination method (ADCC) that controls the transmission power of each base station.

This paper discusses microwave path-loss characteristics as a function of mobile antenna height in an urban line-of-sight environment. Measurements were made in metropolitan Tokyo with high-density buildings, using base station antenna heights of 4 and 8 m. We describe the path-loss characteristics of vehicle-mounted mode (mobile antenna height is 2.7 m) and portable mode (mobile antenna heights are 1.6 and 0.5 m). Dependence of path loss on the distance between base and mobile stations was analyzed. This reveals that the break points shift to the near side in the vehicle-mounted mode. This phenomenon can be interpreted by the existence of an effective height h of the road. The typical value of h was found approximately 1.4 m. In the portable mode, on the other hand, break points were not observed. The mobile antenna heights (1.6 and 0.5 m) in this mode are close to or less than the average height (1-2 m) of pedestrians on the sidewalk; and the received waves at the mobile station are often disturbed by pedestrians. This explains the nonexistence of break points in portable mode. The average attenuation coefficients is observed 3.2 in this mode. The attenuation coefficients tend to be larger at lower base station antenna heights and narrower road widths.

This paper proposes an adaptive transmission control scheme for code-division multiple-access (CDMA) cellular slotted-ALOHA systems. This scheme adaptively controls the target received power and the processing gain according to both channel load and location of the mobile station. The target received power of each mobile station is controlled so that the difference between the target received powers by distance becomes large under heavy load conditions. As the distance from the base station increases, the target received power becomes smaller. The processing gain of transmitted packets is concurrently controlled with their target received powers. The packets transmitted with low signal power are spread by a large processing gain in order to reduce the unfairness in packet reception. The radio channels with distance attenuation, shadowing, slow Rayleigh fading and imperfect power control are taken into consideration in order to evaluate the performance of this scheme in the case that mobile stations transmit short massages to the base station in cellular environments. Computer simulation validates the effectiveness of this scheme: the capture effect can be achieved under heavy channel loads, and therefore, throughput performance is improved. Detailed evaluation of throughput, packet reception probability and transmission complete probability is presented. The effect of movement of mobile stations is also discussed. Calculated results show that the proposed scheme has superior characteristics and thus can expand the allowable load area in the cellular environments with slow Rayleigh fading channels.

This paper describes a new system for extracting and classifying bibliography regions from the color image of a book cover. The same as all the color image processing, the segmentation of color space is an essential and important step in our system; and here HSI color space is adopted rather than RGB color space. The color space is segmented into achromatic and chromatic regions first; and the segmentation is completed after thresholding the intensity histogram of the achromatic region and the hue histogram of the chromatic region. Then text region extraction and classification follows. After detecting fundamental features (stroke width and local label width) text regions are determined by comparing smeared blocks to the original candidate image. Based on the general cover design model, text regions are classified into author region, title region, and publisher region furthermore, and a bibliography image is obtained as a result, without applying OCR. The appearance of the book is 3D reconstructed as well. In this paper, two examples are presented.

An efficient algorithm for reconstructing all polyhedral 3D objects from two orthographic views is presented. Since the two-view orthographic representation of a 3D object is ambiguous, it requires a numerous amount of combinatorial searches in the process of reconstruction. Also, multiple number of solutions in contrast to the designers intention can be existed in the problem. This paper proposes a two phase algorithm to reduce the search space and to select the most plausible solution described by the given projections. First, the partially constructed objects are reconstructed from the restricted candidate faces corresponding to each area on the two-view drawings in its first phase. Then the complete objects are obtained from the partially constructed objects by adding other candidates with geometrical validity in the second phase. The algorithm performs a combinatorial search based on the face decision rules along with two heuristics. Decision rules check geometrical validity and heuristic rules enhance the search speed. In addition, the reconstruction finds the most plausible 3D object that human observers are most likely to select first among the given multiple solutions. Several examples from a working implementation are given to show the completeness of the algorithm.

The architecture design and test results of simulation facility named millimeter-wave Test Bed has been described. Contrast with a millimeter-wave sounder, the Test Bed proposed in this paper can characterize radio channels, received signals, target reflections and radio link performance at the millimeter-wave band of 60 GHz. For fixible simulation and analysis of the performances of newly designed millimeter-wave systems, major digital signal processing parts like a sophisticate waveform generator and an analyzer, a modulator, a demodulator, an encoder, a decoder, an equalizer in the Test Bed are implemented by a software using SPW. This software based Test Bed can be used as a "deign tool" for the simulation of the millimeter-wave communication systems very flexibly without hardware modification in different specifications. The Test Bed consists of a millimeter-wave transmitter, a receiver of 60 GHz, 1.95 GHz up/down converter as IF module and a digital signal processing module. The I/Q vector modulator and demodulator with a video bandwidth of 37.5 MHz in the Test Bed can simulate or test the application of high data rate communication systems of short distance.

In a personal communication system (PCS), a scheme for reforwarding call-terminating setup messages (SETUP messages) from a network or a cell station is used to guard against their loss. We have developed a method for evaluating the loss probability of a reforwarding scheme in which the network monitors the response messages from a personal station after forwarding a SETUP message to that personal station and reforwards the SETUP message only if a response message is not received. We started with a stochastic model in which messages registered in the paging-channel queue in a cell station are cyclically forwarded to the wireless area. This model corresponds to the finite-capacity M/D/1/N model with vacation time. We then added a method for calculating the "timeout" probability. Next we expanded the model into one in which the SETUP messages are reforwarded when a response message is not received by the network. This model corresponds to the M/D/1/N model with vacation time and retrials. We then added an approximate method for calculating the loss probability. Finally, using the proposed methods, we clarified the traffic characteristics of PCS call-terminating control.

An ad hoc wireless network is a collection of mobile hosts that self-forming a temporary network without any required intervention of centralized system. In such environment, mobile hosts, which are not within transmission range from each other, require some other intermediate hosts to forward their packets to form a multi-hop communication. In this paper, an ad hoc network is modeled as a graph. Two nodes within the transmission range of each other are connected by an edge. Given a finite set of mobile nodes, a finite set of edges and a new multicast request, the wireless multicast tree problem (WMTP) is to find a multicast tree for the request so that the multicast loss probability is minimized. We prove the WMTP is NP-complete and a heuristic algorithm, called Degree-Based Multicast Routing Algorithm (DBMRA), is proposed. Based on the DBMRA, one algorithm was proposed to establish backup nodes for the multicast tree to improve the reliability. A node is needed to be backup only when it has a high probability to disconnect the multicast tree seriously. The qualification of a node to be backup is subject to a computed threshold, which is determined by a statistic analysis. The theoretical and experimental analyses are presented to characterize the performance of our algorithms.

We propose a novel planar filter design for narrow-band applications. The filter consists of half-wavelength ring resonators with open gaps. This design has three advantages over conventional planar designs: a smaller size despite narrow bandwidth, a sharper skirt response at the passband edge without notch, an excellent out-band attenuation. We demonstrated these advantages by fabricating an 8-poles filter centered at 1.95 GHz with a 5 MHz bandwidth using YBCO films on a 2 inch diameter MgO substrate.

We propose using punctured convolutional code for a Discrete Multitone (DMT) Asymmetric Digital Subscriber Line (ADSL) system to improve the data rate performance of ADSL. The theoretical performance of the systems with single coding rate for all subchannels, and with adaptive coding rate for each subchannel are presented respectively. It is found that the data rate performance of the proposed system is higher than the uncoded system.

In this paper, the performance of some communication systems using chaos synchronization is evaluated and compared. A new channel model taking the attenuation, impedance mismatch and noise into account, is proposed for the performance evaluation. The evaluation of bit error rate is done for both ideal and non-ideal conditions using the channel model. It is confirmed that some chaos-based communication systems have a good performance compared with conventional analog communication schemes.

In this paper, a novel design method for bandpass filter with attenuation poles (BAP) is presented. The changed inverter element values due to inserting either capacitors or inductors can be optimized using the linear relationship between inverter element values of a conventional bandpass filter (BPF) and those of the BAP using the Touchstone program. A 1800-1825 MHz bandpass filter with attenuation poles for duplexers is designed and fabricated using coaxial dielectric resonators. The validity of this design approach is demonstrated by a computer simulation. The resonators are simulated equivalently as shorted lossy transmission lines. The measured results of center frequency, bandwidth, and attenuation pole frequencies closely agree with the design values.

Coupling effects between resonators on a laminated Band Elimination Filter (BEF) is studied. The coupling degrades the filter attenuation performance. A new equivalent circuit of coupled-line BEF with loaded capacitors is investigated. The performance is simulated and the improvement by staggered resonators is confirmed. An experimental filter made of Low Temperature Co-fired Ceramics (LTCC) is constructed. It shows good performance.

In this paper CMOS VLSI circuit solutions are suggested for on-chip learning and weight storage, which are simple and silicon area efficient. In particular a stochastic learning scheme, named Random Weight Change, and a multistable weight storage approach have been implemented. Additionally, the problems of the influence of technological variations on learning accuracy is discussed. Even though both the learning scheme and the weight storage are quite general, in the paper we will refer to a class of networks, named Approximate Identity Neural Networks, which are particularly suitable to be implemented with analog CMOS circuits. As a test vehicle a small network with four neurons, 16 weights, on chip learning and weight storage has been fabricated in a 1.2 µm double-metal CMOS process.

This paper describes a multibeam antenna which uses switched parasitic and switched active elements to obtain multiple simultaneous directional beams that can be steered in azimuth. A 13 element monopole multibeam array has been optimised for gain and front to back ratio. Results from numerical simulation and measurements in an anechoic chamber are presented. The 13 element array can achieve up to three beams simultaneously with a minimum gain over 360 azimuthal coverage of 1.2 dB less than the maximum gain. Located on a ground plane with diameter of 2λ at 1.5 GHz, the maximum elevation angle was 20.2 with -3 dB vertical beamwidth of 88.

Anomalous leakage current which flows between source and drain in thin film SOI MOSFET's is investigated. It is confirmed that the leakage current is caused by enhanced diffusion of the source/drain dopants along the LOCOS-induced crystal defects. Stress analysis by 2D simulation reveals that thinning a buried-oxide effectively suppresses deformation of an SOI film associated with over-oxidation during LOCOS. It is experimentally confirmed that using a SIMOX substrate which has a thinner buried-oxide causes no noticeable deformation of the SOI film nor anomalous leakage current.

In microwave hyperthermia for cancer therapy, two power feeding techniques can be utilized: incoherent and coherent operations. In the incoherent operation, not-synchronized microwave power is fed into each array element, whereas the coherent operation is achieved by feeding synchronized microwave to the array elements. The authors have been studying the coaxial-slot antenna for interstitial microwave hyperthermia. The antenna is usually employed as an array applicator inserting several antennas into the tissue to generate large heating area. So far we have examined the control of the heating pattern by feeding techniques in order to obtain more uniform and enlarged heating region. Particularly, `tip-heating,' which means sufficient heating at the area near the tip of the applicator, is significant not to damage surrounding normal tissue in interstitial hyperthermia. In this paper, two feeding techniques are combined and calculated temperature distributions in a hexagonal array applicator are examined by solving Pennes bioheat transfer equation by finite difference method. As a result, in the coherent feeding, large heating area was obtained, while better tip-heating was achieved in the incoherent feeding. Moreover, an instance of sequential combination of two feeding techniques is depicted. In this case, temperature distribution had both characteristics of large heating area and tip-heating, therefore the ability of the control of heating characteristics by sequential combination of the coherent and the incoherent feedings was presented.