In this paper, 3 classic suffix type algorithms: QS, Tuned BM and BMHq were improved by reducing the average cost of basic operations. Firstly, the multi-windows method was used to let the calculations of the jump distance run in parallel and pipelining. Secondly, the comparison unit was increased to integer to reduce the total number and the average cost of comparisons. Especially for BMHq, the jump distance was increased by good prefix rule and the operations to get the jump distance were simplified by unaligned integer read. Thus, 3 algorithms named QSMI, TBMMI and BMHqMI were presented. These algorithms are faster than other known algorithms in many cases.
CITATION STYLE
Fan, H., Shi, S., Zhang, J., & Dong, L. (2016). Suffix type string matching algorithms based on multi-windows and integer comparison. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9543, pp. 414–420). Springer Verlag. https://doi.org/10.1007/978-3-319-29814-6_35
Mendeley helps you to discover research relevant for your work.