Vol. 22 · No. 5

We introduce to readers a new matrix C for textttMixColumn operation for AES algorithm for discussion. This matrix has significantly larger multiplicative order, ord(C) = 340, than the used one which is 4 only. This makes so called XSL attack less effective. It is possible to find such a matrix due to our new Euler-Fermat-like theorem and its corollaries for regular circulant matrices over GFps.

In this paper, we consider memory-mapping problems in High-Level Synthesis. We focus on the port mapping, bit-width mapping and word mapping, respectively. A 0-1 Integer Linear Programming (ILP) technique is used to solve the mapping problems, which synthesizes the source memory using one or more memory modules from a target memory library at a higher level. This method can not only perform bit-width mapping and word mapping, but it can also perform port mapping at the same time. Experimental results indicate that ILP approach is an effective method for memory reuse in high-level synthesis.

A process functional paradigm is based on applications of processes and functions instead of assignments. The imperative computation and the functional evaluation are clearly separated in a well-balanced manner, providing the strong feedback about the implementation to a user. In this paper we present the unified approach to explicit and implicit environments in PFL – an experimental process functional language, as a uniform basis for approved implementation extensible by additional specification. PFL environmental concept is the generalized implementation substance, which can be found in many programming languages exploiting the large variety of programming paradigms. Environment variables do not occur in expressions, being still visible to a programmer either in PFL textual form or in an equivalent form of control driven data flow graphs. The approach is promising for reasoning about the functional correctness and predicting the behavior of systems.

Based on the works citekosko2, song, zadeh a fuzzy time series model is proposed and applied to predict chaotic financial process. The general methodological framework of classical and fuzzy modelling of economic time series is considered. A complete fuzzy time series modelling approach is proposed which includes: determining and developing of fuzzy time series models, developing and calculating of fuzzy relations among the observations, calculating and interpreting the outputs. To generate fuzzy rules from data, the neural network with SCL-based product-space clustering is used.

In a ubiquitous information environment, massive number of users carrying their portable computers can retrieve information anywhere and anytime using wireless mobile computing technologies. Wireless data broadcasting as a way of disseminating information to the large number of clients, has an inherent advantage by providing all types of users global access to information. An adaptive access method, which tolerates the access failure, has been proposed in an error-prone mobile environment. However the influence of version bits to deal with the updates of the broadcast data has not been exploited for the broadcast with modified but the same size and structure update. The basic idea is to distinguish the type of update that does not influence the change in the size and structure of the broadcast has been introduced. To deal with the types of updates, we classified the users in mobile computing environment into the users in system and the new users. In the proposed continuous algorithms, the user in systems record the previous result and use it efficiently to access the desired records with less number of probes in the broadcast, which is updated by a stream of same size and structure bits. In the performance analysis, the experimental results show that the proposed modified progression method has the best performance, as it requires the minimum cost to access the broadcast data.