Presenting a Method for Line Selection on SCTP-CMT Protocol Using Adaptive Neuro-Fuzzy Inference System (ANFIS)

Abstract

The issue of line selection and the amount of data to be streamed in SCTP_CMT with dissimilar lines is one of the important challenges to increaseSCTP-CMT protocol efficiency. Line selection with lower probability of failing in data streaming from several simultaneous paths is possible by the use of line properties such as bandwidth and line congestion. In this study, a controlling method has been presented using ANFIS for an optimal line selection and the number of data appropriate with each line properties. Using a random data set, the neuro-fuzzy system was evaluated in this study and finally the idleness of data streaming lines was calculated using the proposed neuro-fuzzy system compared to the random selection of data streaming lines. Results of evaluations indicate a reduction of idleness of the data streaming lines in the proposed method compared to the random line selection method in SCTP-CMT. The reduction in idleness of data streaming linesdirectly influences the time of data streaming and can reduce the time required for data streaming.

Keywords: Communication protocol, Sctp-Cmt, Fuzzy logic, Neural network, Fuzzy system, Bandwidth, line congestion.