원문정보
초록
영어
In different environments, there are some differences characteristics of the wireless channel. These differences are called "fingerprint". Extracting the "fingerprint" characteristics of different channels in different environment is very important to the development of wireless communication. This paper is focus on the problem of "fingerprint extraction". Through wireless channel signal inversion and K-means clustering and compressed sensing, establishing an adaptive clustering model. And then establish a reasonable "fingerprint" feature with the existing data and relevant physical background. Use MATLAB to solve the model and verify the accuracy. Computer verification and comparison analysis show that this model can b used to identify the wireless channels sith high accuracy.
목차
1. Introduction
2. Feature Extraction of Wireless Channel
3. Adaptive Clustering Model Establish
3.1.Wireless Channel Signal Based on Compressed Sensing
3.2 The fitting of ideal signal
3.3. Analysis of the Rationality of Fingerprint Characteristics
3.4. K-Nearest Neighbor (KNN) Classification Model Based on Local Weighted Mean
4. Model Verification
5. Conclusion
References