N-version programming

Previous research on software component failure dependencies seems to have been done primarily for parallel components, typically related to diverse and redrmdant components in fault tolerant designs such as N-version programming. These situations are characterised by components that are subject to the same input. We argue that failure dependencies must be viewed more generally, and that possible causes of dependent failure behaviour are more complex than current methods consider. 

Chen, L., Avizienis, A. N-version programming A fault-tolerance approach to reliability of software operation. In International Symposium on Fault-Tolerant Computing, FTCS (1978)... 

N-version programming This is similar to static redundancy. Here, independently written programs (versions) for the same functions are executed parallel and their outputs are voted at special checkpoints. Naturally, the voted data may not be exactly the same. So, criteria must be used to identify and reject faulty versions and to determine all good versions for use. See Fig. XI/1.1.5-2A. 

Software fault tolerance methods. FT, fault tolerance. (A) Software fault n version programming, (B) software fault - recovery block, and (C) software fault - recovery block schematic. 

N-version programming is a system design technique that involves producing two or more software components that provide the same function, but in different ways. This approach is intended to avoid sources of common errors between redundant components. N-version programming is also referred to as multi-version software, dissimilar software, or software diversity. 

The Swedish State Railways experience with n-version programmed systems. 

This paper contains a resume of the Swedish experience with n-version programmed systems the last 12 years ideas and attitudes, the growth of experience, current understanding, and future goals.The two control applications Computer controlled Interlocking System (CCI) and Automatic Train Control (ATQ will be explained. 

N version This program is about what types of plants survive on different planets. For each planet, a plant will be designed. The goal is to learn what type of roots, stem, and leaves allow plants to survive in each environment. Some hints are provided throughout the program. 

Static redundancy In this system n versions of the programs are written and run for performing the same function. There will be a few checkpoints where each of the output from different versions will be voted to be selected as final output. 

Knight, J.C., Leveson, N.G. A Large Scale Experiment In A-Version Programming. In Proceedings of the Fifteenth International Symposinm on Fanlt-Tolerant Computing, Ann Arbor, MI, pp. 135-139 (1985)... 

Davies, C. A., E. Freedman, D. J. Frurip, G. R. Hertel, W. H. Seaton, and D. N. Treweek (1990). CHETAH Version 4.4 The ASTM Chemical Thermodynamic and Energy Release Evaluation Program. 2nd Edition. Philadelphia, PA American Society for Testing and Materials. 

The first widely used molecular mechanics program was developed by Professor N. L. Allinger of the University of Georgia and was known in its various versions as MM2, MM3, and so on. They have been refined to the extent that many structural features can be calculated more easily and more accurately than they can be measured experimentally. 

We show in Figs. 2 and 3 the speed up for a Davidson iteration obtained with the PVM [17] ( Parallel Virtual Machine ) and PVMe [18] ( PVM enhanced ) message passing interfaces respectively. Speed-up factors are here relative to the sequential version of the program (n=l), and the theoretical mciximum has been defined according to the expression s n) — n — I appropriate to the used master/slave model. 

With the molecular descriptors as the X-block, and the senso scores for sweet as the Y-block, PLS was used to calculate a predictive model using the Unscrambler program version 3.1 (CAMO A/S, Jarleveien 4, N-7041 Trondheim, Norway). When the full set of 17 phenols was us, optimal prediction of sweet odour was shown with 1 factor. Loadings of variables and scores of compounds on the first two factors are shown in Fig es 1 and 2 respectively. Figure 3 shows predicted sweet odour score plotted against that provid by the sensory panel. Vanillin, with a sensory score of 3.3, was an obvious outlier in this set, and so the model was recalculated without it. Again 1 factor was r uired for optimal prediction, shown in Figure 4. 

---