Informatica

Identifying legal and illegal states significantly reduces computational complexity of ATPG. A unified framework for identification of the legal and illegal states is presented. Most known methods for identification of the legal and illegal states are interpretable within this framework. New theorems and the resulting procedures for identifying exact collection of legal or illegal states of a circuit are presented. Experimental results demonstrate that exact collection of legal states for some circuits is significantly smaller than collections obtained by backward state search algorithm and by algorithm based on combinational ATPG theorems. The use of the exact collection of legal states allows identifying more undetectable faults. The proposed procedure for identifying of the exact collection of legal states starts from any state of the circuit, builds initially an enlarged collection of legal states and converges rapidly to the exact solution.

A partially blind signature scheme allows the signer to inoculate a non‐removable common information into his blind signature. This common information may represent the date or the amount of e‐cash. Due to its un‐traceablility and partial blindness property, the partially blind signature plays an important role in many e‐commerce applications. Based on the RSA scheme, we propose a partially blind threshold signature with low‐computational load for the client.

This work describes a realistic performance prediction tool for the parallel block LU factorization algorithm. It takes into account the computational workload, communication costs and the overlapping of communications by useful computations. Estimation of the tool parameters and benchmarking are also discussed. Using this tool we develop a simple heuristic for scheduling LU factorization tasks. Results of numerical experiments are presented.

We consider a generalized model of neural network with a fuzziness and chaos. The origin of the fuzzy signals lies in complex biochemical and electrical processes of the synapse and dendrite membrane excitation and the inhibition mechanism. The mathematical operations included into fuzzy neural network modeling are: the scalar product between inputs of layers and synaptic weights is replaced by a fuzzy logic multiplication, the sum of products changes to the fuzzy logic sums, and the operators such as supremum, maximum, and minimum are presented for a fuzzy description. The algorithm of varying membership functions, built basing on a backpropagation paradigm and a method of fuzzy neural optimization, has been considered. Both fuzzy properties and a chaos phenomenon are analyzed basing upon experimental computations.

In an internet environment, such as UNIX, a remote user has to obtain the access right from a server before doing any job. The procedure of obtaining acess right is called a user authentication protocol. User authentication via user memorable password provides convenience without needing any auxiliary devices, such as smart card. A user authentication protocol via username and password should basically withstand the off‐line password guessing attack, the stolen verifier attack, and the DoS attack. Recently, Peyravian and Zunic proposed one password transmission protocol and one password change protocol. Later, Tseng et al. (2001) pointed out that Peyravian and Zunic's protocols can not withstand the off‐line password guessing attack, and therefore proposed an improved protocol to defeat the attack. Independently, Hwang and Yeh also showed that Peyravian and Zunic's protocols suffer from some secury flaws, and an improved protocol was also presented. In this paper, we show that both Peyravian and Zunic's protocols and Tseng et al.'s improved protocol are insecure against the stolen verifier attack. Moreover, we show that all Peyravian and Zunic's, Tseng et al.'s, and Hwang and Yeh's protocols are insecure against DoS attack.

Sun's nonrepudiation threshold proxy signature scheme is not secure against the collusion attack. In order to guard against the attack, Hwang et al. proposed another threshold proxy signature scheme. However, a new attack is proposed to work on both Hwang et al.'s and Sun's schemes. By executing this attack, one proxy signer and the original signer can forge any valid proxy signature. Therefore, both Hwang et al.'s scheme and Sun's scheme were insecure.

This paper discusses the linear periodically time‐varying (LPTV) system parameter estimation using a block approach. An block algorithm is proposed for optimal estimation of the parameters of LPTV system from the input sequence and the output sequence corrupted by additive Gaussianly distributed noise. In the proposed method, the least squares error criterion has been used.The algorithm provides a useful computational tool based on an appropriate theoretical foundation for parameter estimation of linear time‐invariant (LTI) systems from input and output data. Simulation results are presented that demonstrate the performance of the approach.

A quick gradient training algorithm for a specific neural network structure called an extra reduced size lattice‐ladder multilayer perceptron is introduced. Presented derivation of the algorithm utilizes recently found by author simplest way of exact computation of gradients for rotation parameters of lattice‐ladder filter. Developed neural network training algorithm is optimal in terms of minimal number of constants, multiplication and addition operations, while the regularity of the structure is also preserved.

Reinforcement learning addresses the question of how an autonomous agent can learn to choose optimal actions to achieve its goals. Efficient exploration is of fundamental importance for autonomous agents that learn to act. Previous approaches to exploration in reinforcement learning usually address exploration in the case when the environment is fully observable. In contrast, we study the case when the environment is only partially observable. We consider different exploration techniques applied to the learning algorithm “Utile Suffix Memory”, and, in addition, discuss an adaptive fringe depth. Experimental results in a partially observable maze show that exploration techniques have serious impact on performance of learning algorithm.

This paper investigates a class of sharp integral means inequalities involving fractional calculus operators of certain analytic functions with negative coefficients. Some consequences of the main result are also mentioned.