Paper Title
Selecting Alternatives Combination for NVP By TOPSIS

Abstract
N version programming (NVP) techniques use two or more versions of the same software module that meet design diversity requirements.Different teams, coding languages or different algorithms can be used to minimize the possibility of failure in common mode in different versions.An N-version software (NVS) unit is a fault-tolerant software unit based on a general decision algorithm to determine a consensus result of the results given by NVS units.This decision algorithm is generally applied as a general electorate to arbitrate between the results.In practice, the number of spare modules is low and is generally considered to be three or five. In terms of functionality, voting algorithms can be divided into three main categories: general voting algorithms, hybrid voting algorithms, and intended voters. Majority voters are well known methods in general voting algorithms. This algorithm uses a decision rule which at least selects from among n variants of which more than one half of the variant results accepts.One of the weaknesses of voting algorithms, which are likely to agree on incorrect outcome variants, is the consensus on identical mispredictions.Hybrid voting algorithms use some additional information, such as the level of reliability of variants, to improve the accuracy of the vote.In a fault tolerant system based on software diversity and voting strategy, there is a difference between accuracy and agreement.In other words, correct and accepted false outputs agreed on these algorithms can be accepted. An NVP plan with a definite voter and large output area is proposed.In this diagram, if the version errors are indeed statistically independent, only the agreement between the two variables is sufficient. This algorithm uses a decision rule which at least selects from among n variants of which more than one half of the variant results accepts.One of the weaknesses of voting algorithms, which are likely to agree on incorrect outcome variants, is the consensus on identical mispredictions. Hybrid voting algorithms use some additional information, such as the level of reliability of variants, to improve the accuracy of the vote.In a fault tolerant system based on software diversity and voting strategy, there is a difference between accuracy and agreement.In other words, correct and accepted false outputs agreed on these algorithms can be accepted.An NVP plan with a definite voter and large output area is proposed.In this diagram, if the version errors are indeed statistically independent, only the agreement between the two variables is sufficient. It is used in decision making systems as well as in which versions to choose in addition to this booster selection. It is the right and the most appropriate alternative for any decision that multi-criteria decision making (MCDM) takes.The reliability or objectivity of decisions that play a crucial role in the design phase.If the decisions are taken superficially and only on the basis of judgment, NVP will cause the design to fail.Therefore, changing traditional decision making methods with modern decision making methods will be advantageous in terms of reliability and availability of the cost system. As TOPSIS (Technique for Order Performance by Similarity to Ideal Solution), one of the most important features of modern decision making methods, it helps to make decisions by considering many criteria for a more objective performance evaluation. The TOPSIS method is widely used in the literature because it is a popular method in MCDM. In this study, N version programming and safety critical applications will be discussed about Voting algorithms, Example NVP Problem definition, TOPSIS algorithm from MCDM methods, Example NVP problem example and Results. Index terms - TOPSIS,MCDM, N-version software, software reliability, fault tolerance