With the popularity of wireless networks and smart devices, indoor positioning has been rapidly developed. In indoor positioning, the fingerprint-based positioning method has gradually become a research hotspot because it does not require external facilities and strong anti-interference. The development of deep learning in recent years has brought new opportunities for improving the accuracy of fingerprint positioning algorithms. This paper proposes a convolutional neural network (CNN)-based fingerprint location algorithm to improve the construction of the fingerprint database. First, the collected CSI and magnetic field data is processed through CNN, and the CNN model parameter values are used at each reference point as fingerprint. Then a probabilistic method is utilized for the final fingerprint matching. Experimental results show that the proposed positioning algorithm has better robustness and higher positioning accuracy than the traditional fingerprint positioning algorithm.