ISSN : 2319-7323


Open Access


Title : A Survey on Prediction Methodologies
Authors : Lokesh M R, Y.S. Kumaraswamy, Smaran N S
Keywords : Next state prediction, Markov Models, Hidden Markov models, Avionics systems
Issue Date : July 2015
Abstract : This paper covers increasing need for next state prediction in the field of the avionics system that will predict and overcome the faults caused by the avionics systems. “Next State Prediction” is used for the predication of the possible states that the system can exist. This paper explains about how Markov models and hidden markov models can be used for the prediction of the state in which the system exist based on which it could be stated if it is in a safe state. In the current scenario it is not possible to have 100% test coverage for all the test cases. There will be instances where some of the portions that will not be covered by the test coverage and may lead to catastrophic faults. A very good process of verification and validation is carried out before the software is operational. However, all these are carried out before the software is commissioned on the target for its stipulated use. Further, the V&V is carried out in a simulated environment. It is very difficult to simulate the environment in to. There could be some environmental conditions which have not been simulated and which could trigger the software state to an unsafe condition. Hence we use this methodology so as to predict if the current state or any further states in which an aircraft could exist would be safer one , if not, necessary steps could be carried out so as to make it safer.
Page(s) : 116-121
ISSN : 2319-7323
Source : Vol. 4, No.4