Application of ELM Algorithm Incorporating AE Principles in Wireless Communication Signal Detection

Abstract

With the development of society, the requirements for communication are getting higher and higher, but the detection performance of wireless communication signals is insufficient. To solve this problem, this paper combines the principle of Auto-encoder (AE) and the feature classification function of Extreme Learning Machine (ELM) algorithm to obtain AE-ELM algorithm, and builds a new wireless communication signal detection model based on this algorithm, hoping to improve the detection performance of wireless communication signals. The performance comparison experiment of the proposed AE-ELM algorithm shows that the bit error rate and average relative time complexity of the algorithm are 0.0004 and 0.5292s respectively, which is better than the comparison algorithm. Additionally, the effectiveness of the fusion algorithm-based signal detection model for wireless communication is investigated, and the findings demonstrate that this model outperforms the others with the lowest detection error rate of 0.0031. The above results demonstrate the high application potential of the novel wireless communication signal detection model and its ability to significantly improve the performance of wireless communication signal detection. Its application to wireless communication signal detection can greatly promote the development of the signal detection field in China.

Keywords

• AE
• ELM
• wireless communication
• signal detection
• deep learning