| Popis: |
Currently, the threshold control for the monitor and management of network equipment functionality adopts both static and dynamic threshold strategies. In traditional dynamic threshold computation methods, there is no support for error data correction and data sampling time interval is long. With these methods, the computed baseline value keeps unchanged within each sampling time interval. Hence, the calculated baseline value is not good at describing the changing trend of network equipment functionality. In other words, it has worse support to applications with real time and dynamic requirement. In this paper, a novel approach for computing dynamic threshold is proposed based on error correction. It improves the traditional dynamic threshold computation model. With the developed system for the monitor and management of network equipment functionality, we prove the feasibility of the dynamic threshold computation based on error correction. Through the experiment, we demonstrate that the dynamic threshold computation based on error correction can improve the threshold predicting accuracy with better support to real time and dynamic requirement. It has excellent performance in describing the changing trend of network equipment functionality. |
