Equalization System Design Based on CPU+GPU Heterogeneity
Guangkun Luo, Yumeng Liu, Yangbin Wei, Jiawei Cao, Hao Li, Zhigang Wang · 2023
With the development of digital communication systems, higher requirements have been put forward on how to eliminate channel noise interference. This paper adopts blind equalization algorithms and combines the idea of adaptive equalization to design heterogeneous blind equalization software processing schemes for different modulation formats. The scheme uses CPU to calculate the equalization coefficient, GPU relies on stable equalization coefficients to complete parallel equalization correction, and combines EVM parameters to determine whether the equalization coefficient needs to be updated, achieving a blind equalization software architecture implementation of CPU+GPU heterogeneity. Compared to traditional serial equalization methods, its computing speed has been greatly improved.