A Novel Double-Time-Scale Evaluation Method for Ethernet Performability

Abstract

Generally, there are two ways in Ethernet study, i.e., structural reliability analysis and performance evaluation. The former cannot evaluate the ability of Ethernet to perform particular functionalities precisely, whereas the later ignores random failures, and thus cannot reflect the long-term performance degradation of Ethernet. This article proposes a novel double-time-scale evaluation method to overcome the above disadvantages. The study of Ethernet is conducted in Macro and Micro levels, i.e., the relationship between the number of operational workstations and the random failures is formulated in Macro scale, and the performance is modeled with the number of operational workstations in Micro scale. This method can synthesize the effect of reliability and performance, and evaluate the Ethernet in a systematic way effectively. It reflects the actual influence of random failures on Ethernet performance degradation. Because the impact of main Ethernet design parameters is included in the analytical model of Ethernet performability, this method offers a powerful tool for design optimization of the Ethernet.

Keywords:

• Double time scale
• Ethernet
• Performability
• Performance
• Reliability

Mathematics Subject Classification:

• 68M10

Acknowledgments

The research was result of a project financed by National Natural Science Foundation of China (No. 61004022) and 111 Project of China Education Department.