Improved and Comparative End to End Delay Analysis in CBS and TAS using Data Compression for Time Sensitive Network
Md. Mehedi Hasan, Feng He, Tanvir Hasan, Bipro Gain, Md Ibrahim Ullah, Shahrukh Khan · 2021 3rd International Conference on Applied Machine Learning (ICAML) · 2021
TSN, or Time-Sensitive Network, is a set of standards that allows an Ethernet network to be extended to serve real-time applications while ensuring safety-critical hardware delivery. TSN uses a variety of shaping and scheduling techniques, like Credit-Based Shaping, Asynchronous Traffic Shaping (ATS), Burst-Limiting Shaper (BLS), and Time Aware Shaper (TAS), to improve real-time guarantees. In modern industries, the increasing number of traffic and data are causing the latency to increase significantly. To minimize the end-to-end delay, it is highly required to have lossless compress data. This paper proposes some pre-existing lossless compression algorithms (Huffman, LZW, RLE), which are well suited for the TSN network. It can improve the end-to-end delay despite increasing data and is applicable for the worst-case scenarios. So, the focus of this research is confined to CBS and CBS+TAS shaping mechanism. For the validation of our statement, we have simulated various traffics such as Control Data Traffic (CDT), Audio & Video Bridging (AVB) class A, AVB class B and best-effort by OMNeT++. Comparisons are shown in terms of end-to-end delay between the compressed and uncompressed data and end-to-end delay have been improved for all classes.