Research on the Algorithm of LAN Performance Evaluation Based on Deep Learning

Abstract

A network performance evaluation algorithm is presented based on deep learning framework RNN on this paper. The algorithm collects 8 index parameters in the private LAN and realizes the automatic evaluation of the three levels of the private LAN: excellent, good, and unqualified. The results of the example verification and analysis show that the algorithm proposed in this paper evaluates the performance of the private area network using K-fold cross-validation. The correctness of the algorithm is over 90%. The robustness and high efficiency of the algorithm, the computing time of milliseconds, can be effectively applied to the real-time performance evaluation of the local area network.