In this paper, we propose a new scheduling algorithm to achieve fault tolerance in multiprocessor systems. This algorithm first partitions a parallel program into subsets of tasks, based on the notion of height of a task graph. For each subset, the algorithm then duplicates and schedules the tasks in the subset successively. We prove that schedules obtained by the proposed algorithm can tolerate a single processor failure and show that the computational complexity of the algorithm is O(|V|4) where V is the set of nodes of a task graph. We conduct simulations by applying the algorithm to two kinds of practical task graphs (Gaussian elimination and LU-decomposition). The results of this experiment show that fault tolerance can be achieved at the cost of small degree of time redundancy, and that performance in the case of a processor failure is improved compared to a previous algorithm.

A micro reactor array for biochemical or biomedical use was developed. Conceptof this development is to get as much as biological data at the same time. Ninety-six micro reaction wells, volume of each well was 1.5 µl, were integrated in the array. The micro reactor array was fabricated on 1 mm thick silicon wafer and twelve pairs of a temperature sensor and a heater were formed on the backside. A tiny transparent window for optical measurement was formed at the center of bottom wall on each well. Several temperature gradients were applied to the array by means of few heaters and compared with simulation results to optimize the parameters. Finally, performance of the array was evaluated by basic DNA reaction. Advantages of the array system are the fast thermal response due to the small heat capacity and easy to make several reaction conditions in parallel.

Multilayer feedforward neural network (MFNN) trained by the backpropagation (BP) algorithm is one of the most significant models in artificial neural networks. MFNNs have been used in many areas of signal and image processing due to high applicability. Although they have been implemented as analog, mixed analog-digital and fully digital VLSI circuits, it is still difficult to realize their hardware implementation with the BP learning function efficiently. This paper describes a special BP algorithm for the logic oriented neural network (LOGO-NN) which we have proposed as a sort of MFNN with quantized weights and multilevel threshold neurons. Both weights and neuron outputs are quantized to integer values in LOGO-NNs. Furthermore, the proposed BP algorithm can reduce high precise calculations. Therefore, it is expected that LOGO-NNs with BP learning can be more effectively implemented as digital type circuits than the common MFNNs with the classical BP. Finally, it is shown by simulations that the proposed BP algorithm for LOGO-NNs has good performance in terms of the convergence rate, convergence speed and generalization capability.

In this report, a design method of neural networks for limit cycle generator is described. First, the constraint conditions for the synaptic weights, which are given by the linear inequalities, are derived from the dynamics of neural networks. Next, the linear inequalities are solved by the linear programming method. The synaptic weights and other parameters are determined by the above solutions. Furthermore, neuro-based limit cycle generator is designed with analog electronic circuits and simulated by Spice. Finally, we confirm that our design method is efficient and practical for the design of neuro-based limit cycle generator.

High-temperature superconducting (HTS) films have low surface resistance in the microwave region. This indicates that HTS passive microwave devices, for example, filters and antennas, are promising for commercial use. For superconducting filters, low insertion loss, sharp skirt property and high attenuation value out of band are important, and for superconducting antennas, electrically small antennas and super-directive high-gain antennas are useful. It is known that the elliptic filter has a sharp skirt property compared with that of a Chebyshev filter. We examined a cross-coupled filter which is one type of elliptic filter. The insertion loss of the filter was approximately 0.3 dB. The skirt property of the four-resonator cross-coupled filter was almost equal to that of an eight-resonator hairpin filter. We also examined a cryocooler cooling system for superconducting antenna and measured the relative gain, directivity and power capability of an HTS antenna. The gain of the HTS patch antenna was approximately 10 dB larger than that of a normal conducting antenna. The large gain of the HTS antenna was caused by the low surface impedance and enhancement effect of the dielectric window of the cooling system.

High-temperature superconductor (HTS) receiving filter subsystem for mobile telecommunication base station has been developed. An 11-pole HTS filter using YBa2Cu3O7-δ (YBCO) thin films and a low noise amplifier were cooled to 70 K by a small cryocooler. Total noise figure of this subsystem was measured to be 0.5 dB. Furthermore the effect of using the subsystem in the receiver front-end of Code Division Multiple Access (CDMA) cellular base station was investigated. The transmitting power reduction of handy terminal was estimated to be about 35%.

In this paper, we introduce a lower bound for the generalized Hamming weights, which is applicable to arbitrary linear code, in terms of the notion of well-behaving. We also show that any [n,k] linear code C over a finite field F is the t-th rank MDS for t such that g(C)+1 t k where g(C) is easily calculated from the basis of Fn so chosen that whose first n-k elements generate C. Finally, we apply our result to Reed-Solomon, Reed-Muller and algebraic geometry codes on Cab, and determine g(C) for each code.

In this paper, we give lower bounds for the generalize Hamming weights of linear codes constructed on affine algebraic varieties. By introducing a number g*, which is determined by a given affine variety, we show that when the order t of generalized Hamming weights is greater than g*, the proposed lower bound agrees with their true generalize Hamming weights. Moreover, if we use the notion of well-behaving, we can obtain a more precise bound. Finally, we compare the proposed bound and the conventional one for algebraic geometric code on the curve Cab.

A new or future system of mobile telecommunication is built by new digital technologies to provide an improved and more consistent quality of service for the customers. These digital systems can provide greater number of transmission channel allocation for their subscribers and security. On the digital communication system, distortion of transmitted signal should be eliminated as much as possible for high communication quality. However, the need to both minimize distortion of signal amplifiers and continue to provide good filtering protection can become difficult to achieve with conventional high power amplifiers and filters. In this paper, the application of high-Tc superconducting (HTS) power filters on such digital communication systems and recent progress of filter device developments for those are discussed.

Large-area high-temperature superconducting films and damage-free processing techniques have been developed to fabricate low insertion loss and sharp skirt filters for mobile telecommunication. An off-axis-type dc sputtering method was employed to deposit Y-Ba-Cu-O films on both sides of the substrate. The surface resistance of the films was about 0. 35 mΩ(at 70 K and 10 GHz). An 11-pole bandpass receiving filter for the IS-95 telecommunication system was designed and fabricated using 60 mm 50 mm YBCO films on a 0. 5-mm-thick MgO substrate. The passband insertion loss at 70 K was about 0. 1 dB with 0. 1 dB ripple. The third-order intercept point of the filter was 49. 5 dBm. We have assembled the filter and a low-noise amplifier in a dewar with a cryocooler. Ultralow-noise performance (noise figure: 0. 5 dB at 70 K) was presented by the cryogenic filter subsystem.

Considering the pattern classification/recognition tasks, the influence of the activation function on fault tolerance property of feedforward neural networks is empirically investigated. The simulation results show that the activation function largely influences the fault tolerance and the generalization property of neural networks. It is found that, neural networks with symmetric sigmoid activation function are largely fault tolerant than the networks with asymmetric sigmoid function. However the close relation between the fault tolerance and the generalization property was not observed and the networks with asymmetric activation function slightly generalize better than the networks with the symmetric activation function. First, the influence of the activation function on fault tolerance property of neural networks is investigated on the XOR problem, then the results are generalized by evaluating the fault tolerance property of different NNs implementing different benchmark problems.

A new learning algorithm is proposed to enhance fault tolerance ability of the feedforward neural networks. The algorithm focuses on the links (weights) that may cause errors at the output when they are open faults. The relevances of the synaptic weights to the output error (i.e. the sensitivity of the output error to the weight fault) are estimated in each training cycle of the standard backpropagation using the Taylor expansion of the output around fault-free weights. Then the weight giving the maximum relevance is decreased. The approach taken by the algorithm described in this paper is to prevent the weights from having large relevances. The simulation results indicate that the network trained with the proposed algorithm do have significantly better fault tolerance than the network trained with the standard backpropagation algorithm. The simulation results show that the fault tolerance and the generalization abilities are improved.