Semi-Orthogonally Associative Memory neural network model (SAM) uses the orthogonal vectors in Un = {-1, 1}n as its characteristic patterns. It is necessary to select the optimum characteristic parameter n so as to increase the efficiency of this model used. This paper investigates the dynamic behavior and error correcting capability of SAM by statistical neurodynamics, and demonstrates that there exists a convergence criterion in tis recalling processes. And then, making use of these results, its optimum characteristic parameter is deduced. It is proved that, in the statistical sense, its recalling outputs converge to the desired pattern when the initial similar probability is larger than the convergence criterion and not true otherwise. For a SAM with N neurons, when its characteristic parameter is optimum, its memory capacity is N/2 ln ln N, the information storage capacity per connection weight is larger than 9/23 (bits/weight) and the radius of attractive basin of non-spurious stable state is about 0.25N. Computer simulations are done on this model and the simulation results are consistent with the results of theoretical analyses.

This paper proposes a new RSA-type scheme over non-singular parts of singular cubic curves En(α,β):(y-αx)(y-βx)x3(mod n). In usual one-to-one communication, we prove that breaking the proposed scheme is not easier than breaking the RSA scheme for the whole ciphertexts. If encryption key e is larger than 19 for 512 bits modulus n, then the proposed scheme is secure against the Hastad attack in broadcast applications. A plaintext of two blocks, i.e., x and y coordinates in En(α,β), is encrypted to a ciphertext of three blocks, where the size of one block is log2n bits. The decryption speed ofthe proposed scheme is as fast as that of the RSA scheme for the even block plaintext.

In this paper we discuss the relation between a one-way group homomorphism and a one-way ring homomorphism. Let U,V be finite abelian groups with #U=n. We show that if there exists a one-way group homomorphism f:UV, then there exists a one-way ring homomorphism F:ZnUZnImf. We also give examples of such ring homomorphisms which are one-way under a standard cryptographic assumption. This implies that there is an affirmative solution to an extended version of the open question raised by Feigenbaum and Merrit: Is there an encryption function f such that both f(x+y) and f(xy) can be efficiently computed from f(x) and f(y)? A multiple signature scheme is also given as an application of one-way ring homomorphisms.

It is important to find the best linear expression to estimate the vulnerability of cryptosystems to Linear Cryptanalysis. This paper shows the results of the best linear expressions search of FEAL-N (N32) and discusses the security of FEAL against Linear Cryptanalysis. We improve Matsui's search algorithm which determines the best linear expressions, and apply it to FEAL. The improved search algorithm finds all the best linear expression of FEAL-N (N32) much faster than the original; the required time is decreased from over three months to about two and a half days. We find the best linear expressions of FEAL-7, FEAL-15, and FEAL-31 with deviations of 1.152-8, 1.482-20, and 1.992-41, respectively. These linear expressions have higher deviations than those derived from Bi-ham's 4-round iterative linear approximations. Using these data we calculated the number of known plaintexts required to attack FEAL-8, FEAL-16, and FEAL-32. It is proved that FEAL-32 is secure against Linear Cryptanalysis.

RDES cryptosystem is an n-round DES in which an probabilistic swapping is added onto the right half of the input in each round. It is more effective than a simple increase of DES rounds for a countermeasure against differential attack. In this paper, we show that the RDES is also effective against linear cryptanalysis. We applied Matsui's search algorithm to find the best expression for RDES-1 and RDES-2. The results are as follows: (a) The 16-round RDES-1 is approximately as strong as a 22-round DES, and the 16-round RDES-2 is approximately as strong as a 29-round DES. (b) Linear cryptanalysis for a 16-round RDES-1 and a 16-round RDES-2 requires more than 264 known-plaintexts.

We create a new version of the FEAL-N(X) cryptographic function, called FEAL-N(X)S, by introducing a dynamic swapping function. FEAL-N(X)S is stronger against Differential Cryptanalysis in the sense that any characteristic for FEAL-N(X) is less effective when applied to FEAL-N(X)S. Furthermore, the only iterative characteristics. that may attack the same number of rounds for the two versions are the symmetric ones, which have an average probability bounded above by 2-4 per round, i.e., the FEAL-N(X)S is at least as strong as DES with respect to this type of characteristic. We also show that in general the probability of an iterative characteristic for the FEAL-N(X) that is still valid for FEAL-N(X)S is decreased by 1/2 per round. Some of the best characteristics are shown. Experimental results show that the running time required by FEAL-N(X)S is around 10% greater compared to FEAL-N(X), in software; but this price is small compared to the gained strength against Differential Cryptanalysis.

GDL is the language whose membership problerm is polynomial-time Turing equivalent to the discrete logarithm problem for a general finite group G. This paper gives a characterization of GDL from the viewpoint of computational complexity theory. It is shown that GDL NP co-AM, assuming that G is in NP co-NP, and that the group law operation of G can be executed in polynomial time of the element size. Furthermore, as a natural probabilistic extension, the complexity of GDL is investigated under the assumption that the group law operation is executed in an expected polynomial time of the element size. In this case, it is shown that GDL MA co-AM if G MA co-MA. As a consequence, we show that GDL is not NP-complete unless the polynomial time hierarchy collapses to the second level.

This paper describes a novel shared file system, whose main features are enhanced security and its concurrency control mechanism. The system is especially suitable for access from mobile hosts. Users can edit their shared files concurrently. Shared files are encrypted and decrypted only by clients, and the file server cannot know the contents. The server asynchronously receives the edited parts, which are already encrypted, and merges them into the current version, deciphering neither the stored file nor the encrypted editing data. We call the mechanism 'privacy enhanced merging'. The mechanism and the underlying encryption algorithm, shared file data structure and procedures followed by clients and the server are shown.

The authors propose a waveguide π-junction with an inductive wall. Galerkin's method of moments is applied to analyze it and small reflection and desired power division ratio is realized. Good agreement between the calculated result and the measured one verifies the design of a unit π-junction. The characteristics of aπ-junction with a wall are almost the same as those of a conventional π-junction with a post. Important advantage of the new π-junction with a wall is that it can be manufactured in the die-cast process of the waveguide while a post in the conventional one must be attached in an additional process. A 16-way power divider consisting of 8 π-junctions is designed at 11.85 GHz and the characteristics are predicted.

To realize better bit error rate performance in fast fading environments, this paper proposes the open loop reverse modulation carrier recovery scheme which employs a new open loop carrier extractor and regenerator instead of using a feed back loop. The proposed scheme realizes stable regenerated carrier signals to achieve low bit error rate not only under additive white Gaussian noise environments but also under fast fading environments. Computer simulations clarify that the proposed scheme always achieves better bit error rates than conventional differential detection or coherent detection with feed back loops under the various fading environments examined.

A radial line slot antenna (RLSA) is a high efficiency and high gain planar antenna. The efficiency of RLSA becomes lower as the aperture size reduces due to rotational asymmetry of the illumination. A concentric array RLSA (CA-RLSA) was proposed to overcome this difficulty. It adopted three new techniques; (1) not spirally but concentrically arrayed slots, (2) a rotating mode feed circuit, (3) matching slot pairs eliminating termination loss. This paper proposes the basic slot design for (1) and (2) of CA-RLSA. Excellent characteristics of very small CA-RLSAs based upon this design are confirmed by measurements.

This paper presents a 15-GHz MMIC direct optical injection-locked oscillator (MMIC OILO) with very-wide locking range that uses photosensitive HBTs. The MMIC OILO consists of an HBT and a positive feedback circuit including a Q-damping variable resistor. By utilizing the high-fT/fmax photosensitive HBT, we realize both high-frequency oscillation of 15 GHz and increased equivalent electrical injection power. In addition to increasing the RF injection power, the Q-damping variable resistor effectively reduces the quality-factor of the oscillator, thus realizing the very wide locking range (f) of 567 MHz (f/fosc3.8%). The locking bandwidth of 3.8% is over 10 times wider than that of any yet reported microwave direct OILO. Furthermore, it is shown that the MMIC OILO can also work as a high-gain Q-variable filter photoreceiver by increasing a Q-damping variable resistance over the self-oscillation suppression range.

This paper presents the performance of optically controlled MESFETs as photodetectors. The optical performance characteristics such as optic-to-electric responsivity, and BER for a π/4-QPSK signal are experimentally investigated. Measurements are performed by using MMIC compatible MESFETs. Experimental results are also evaluated in comparison with calculated PIN-PD limit. Optic-to-electric responsivity has high gain at lower received optical powers. It is shown experimentally that MESFET photodetectors improve the permissible optical power by 6 dB compared to the PIN-PD limit. Optically controlled MESFETs will provide a novel receivers for fiber-optic systems.

As an optically controlled array antenna, a "two-laser type" using two laser diodes whose frequency difference is set to the desired microwave frequency has been proposed. In this paper, we confirm experimentally that the beamforming network of the two-laser type array antenna has very broadband characteristics by measuring the amplitude and phase of microwave signals at 1.5 GHz, 10 GHz and 20 GHz. Using the optically controlled beam forming network, the radiation pattern of a 4-element linear array antenna was measured at 1.5 GHz.

Exploitation of air interfaces for mobile communications is rapidly increasing because of diversified service demands, technology trends and radio propagation conditions. This paper summarizes the radio and optic interaction devices and systems that can solve the future problems resulting from spreading demands in mobile multimedia communications. The concept of the Virtual Free Space Network (Radio Highway Network) is proposed for universal mobile access networks that can support any mobile service or radio air-interface. As one example of the proposed network, the optical TDMA network for radio is analyzed and results of some theoretical calculations are shown.

A new configuration is proposed for an optoelectronic network (OEN) using microwave frequency mixing and multiplexing. The mn OEN consists of m optical sources, m-parallel n-stage cascaded optical intensity modulators, and m-photodetectors. The mn OEN matrix is theoretically discussed, and 12, 22 and 33 OENs are analyzed in detail. The 22 OEN, which mixes and multiplexes microwaves, is further investigated and the theoretical prediction derived from OEN equations is experimentally confirmed.

This paper proposes fiber-optic radio highway network using code division multiple access (CDMA) method which is universally applicable for various type of personal radio services and radio air interfaces. The proposed system can asynchronously open the radio-free space among any microcells. The outage probability and the number of connectable radio base stations are theoretically analyzed and compared with these in using conventional subcarrier multiplexing (SCM) method. Analysis results show that the reduction effect of the optical signal beat noise, due to spread spectrum of optical signal, improves the number of the active RBSs in CDMA radio highway network.

This paper describes a finite element method to obtain an accurate solution of the scalar Helmholtz equation with field singularities. It is known that the spatial derivatives of the eigenfunction of the scalar Helmholtz equation become infinite under certain conditions. These field singularities under mine the accuracy of the numerical solutions obtained by conventional finite element methods based on piecewise polynomials. In this paper, a regularized eigenfunction is introduced by subtracting the field singularities from the original eigenfunction. The finite element method formulated in terms of the regularized eigenfunction is expected to improve the accuracy and convergence of the numerical solutions. The finite element matrices for the present method can be easily evaluated since they do not involve any singular integrands. Moreover, the Dirichlet-type boundary conditions are explicitly imposed on the variables using a transform matrix while the Neumann-type boundary conditions are implicitly imposed in the functional. The numerical results for three test problems show that the present method clearly improves the accuracy of the numerical solutions.

An invertible length preserving transducer is called a weakly invertible finite automaton (WIFA for short). If the first letter of any input string of length τ + 1 is uniquely determined by the corresponding output string by a WIFA and its initial state, it is called a WIFA with delay τ. The composition of two WIFAs is the natural concatenation of them. The composition is also a WIFA whose delay is less than or equal to the sum of the delays of the two WIFAs. In this paper we derive various results on a decomposition of a WIFA into WIFAs with smaller delays. The motivation of this subject is from theoretical interests as well as an application to cryptosystems. In order to capture the essence of the decomposability problem, we concentrate on WIFAs such that their input alphabets and their output alphabets are identical. A WIFA with size n of the input and output alphabet is denoted by an n-WIFA. We prove that for any n > 1, there exists an n-WIFA with delay 2 which cannot be decomposed into two n-WIFAs with delay 1. A one-element logic memory cell is a special WIFA with delay 1, and it is called a delay unit. We show that for any prime number p, every strongly connected p-WIFA with delay 1 can be decomposed into a WIFA with delay 0 and a delay unit, and that any 2-WIFA can be decomposed into a WIFA wiht delay 0 and a sequence of k delay units if and only if every state of the 2-WIFA has delay k.

This paper proposes a Japanese continuous speech recognition mechanism in which a full-sentence-level context-free-grammar (CFG) and one kind of semantic constraint called dependency relationships between two bunsetsu (a kind of phrase) in Japanese" are used during speech recognition in an integrated way. Each dependency relationship is a modification relationship between two bunsetsu; these relationships include the case-frame relationship of a noun bunsetsu to a predicate bunsetsu, or adnominal modification relationships such as a noun bunsetsu to a noun bunsetsu. To suppress the processing overhead caused by using relationships of this type during speech recognition, no rigorous semantic analysis is performed. Instead, a simple matching with examples" approach is adopted. An experiment was carried out and results were compared with a case employing only CFG constraints. They show that the speech recognition accuracy is improved and that the overhead is small enough.