Acceleration analysis of the quadratic sieve method based on the online matrix solving

Abstract

<p>The algorithm for the<strong> </strong>online matrix solving<strong> </strong>is proposed. The rate of acceleration of the basic quadratic sieve method based on the online matrix solving<strong> </strong>is investigated. Acceleration of the quadratic sieve method will reduce the runtime, the complexity of the algorithm and expand the set of numbers, where this algorithm is the best.</p><p>It is shown that the modified algorithm has increased the number of successful decompositions. That is, the number of cases where the basic quadratic sieve (standard sieving interval and size of the factor base) failed to form a matrix to obtain a solution was reduced. This became possible due to the fact that in the modified algorithm there is no need to obtain all L<sup>a</sup>+2 B-smooth numbers prior to diagonalization of the matrix, as in the case of the basic method. Among other important characteristics of this method, it should be noted that when used, the same operations as in the basic quadratic sieve method are performed, only their order is changed. The computing complexity decreases if the set of B-smooth numbers, for which the power matrix vectors form a linearly dependent system, are found quickly.</p>According to the data obtained, the modified QS method, based on<strong> </strong>the online matrix solving, provides an acceleration of about 5.45 percent for numbers of 10<sup>130 </sup>in size. It is shown that improvements associated with solving the matrix cannot lead to a significant increase in the sieving interval. After all, the rate of acceleration decreases with increasing number N. Further improvement to the quadratic sieve method should be related to methods aimed at a significant reduction of the sieving interval and the size of the factor base, which in relative terms should be the greater, the higher N

Authors and Affiliations

Stepan Vynnychuk, Vitalii Misko

Keywords

Related Articles

Development of a technique for the reconstruction and validation of gene network models based on gene expression profiles

<p>We have developed a technique for the reconstruction and validation of models of gene networks based on the gene expression profiles derived in the course of DNA microchip experiments or by the method of RNA molecules...

Modification of the PERT method for project time evaluation taking into account unexpected delays

<p>The PERT-based method for estimation of project implementation time, which uses Rayleigh distribution instead of β-distribution, was proposed. The modification of the PERT, which made it possible to involve the qualit...

Determining the influence of the composition of milk from cows of different breeds on quality indicators for the dutch-type cheese

<p>The paper reports a comprehensive study into the impact of feeding rations and breed of cows on the technological properties of milk and quality characteristics of the Dutch type of cheese. Despite the abundance of in...

Thin­walled structures: analysis of the stressed­strained state and parameter validation

<p>The approach is developed to substantiate technical solutions for thin-walled machine building structures. It implies that the problem is considered in the space of generalized parameters. These parameters combine des...

Development of a criteria­based approach to agroecological assessment of slope agrolandscapes

<p class="a">For the forecast and management of erosion processes in order to protect the environment, information is needed on the state of its components and impact factors as well as the results of this impact. The ex...

Download PDF file
  • EP ID EP527871
  • DOI 10.15587/1729-4061.2018.127596
  • Views 75
  • Downloads 0

How To Cite

Stepan Vynnychuk, Vitalii Misko (2018). Acceleration analysis of the quadratic sieve method based on the online matrix solving. Восточно-Европейский журнал передовых технологий, 2(4), 33-38. https://europub.co.uk/articles/-A-527871