{"title":"Statistical Models of Network Traffic","authors":"Barath Kumar, Oliver Niggemann, Juergen Jasperneite","country":null,"institution":"","volume":37,"journal":"International Journal of Computer and Information Engineering","pagesStart":177,"pagesEnd":186,"ISSN":"1307-6892","URL":"https:\/\/publications.waset.org\/pdf\/1610","abstract":"Model-based approaches have been applied successfully\r\nto a wide range of tasks such as specification, simulation, testing, and\r\ndiagnosis. But one bottleneck often prevents the introduction of these\r\nideas: Manual modeling is a non-trivial, time-consuming task.\r\nAutomatically deriving models by observing and analyzing running\r\nsystems is one possible way to amend this bottleneck. To\r\nderive a model automatically, some a-priori knowledge about the\r\nmodel structure\u2013i.e. about the system\u2013must exist. Such a model\r\nformalism would be used as follows: (i) By observing the network\r\ntraffic, a model of the long-term system behavior could be generated\r\nautomatically, (ii) Test vectors can be generated from the model,\r\n(iii) While the system is running, the model could be used to diagnose\r\nnon-normal system behavior.\r\nThe main contribution of this paper is the introduction of a model\r\nformalism called 'probabilistic regression automaton' suitable for the\r\ntasks mentioned above.","references":null,"publisher":"World Academy of Science, Engineering and Technology","index":"Open Science Index 37, 2010"}