With the development of digital wireless communication technol-ogy, the wireless signal identification has been suffering from increasingly complex electromagnetic environment and higher spectrum utilization. In this paper, we propose a wireless signal identification method based on interference cleaning and convolutional neural network (CNN) in 230MHz Band. The method firstly analyzes the received signal in time domain, building feature data sets combined with amplitudes, phases, in-phase components and orthogonal components. The method then generalizes singular value decomposition(SVD) and subspace division to preserve signal subspace, eliminate noise subspace and interference compress subspace. Finally, it utilizes the data set to train the CNN and make the wireless signals' identification through the well-trained the CNN. The experimental results with different kinds of modulation show that this method can achieve high recognition accuracy and strong anti-noise ability.

In multi-cell multi-user massive MIMO systems, pilot contamination(PC) caused by pilot reuse will reduce the performance of channel estimation. Aiming at the drawbacks of the weighted graph coloring based pilot decontamination (WGC-PD) algorithm, the pilot assignment ignores the users of lower potential pilot contamination, which allocates the pilot of lower interference for the user of higher potential pilot contamination sequentially, and a joint power control and pilot assignment algorithm is proposed. The proposed algorithm improves the WGC-PD algorithm, sets a threshold in pilot assignment, determines whether assign small interference pilot to this user, and establishes the optimization model by combining pilot power control after pilot assignment, further reduces pilot contamination and maximizes the uplink achievable sum-rate. The simulation results show that, compared with the WGC-PD algorithm, the proposed algorithm can guarantee the performance of the users with small potential pilot contamination and obtain higher uplink achievable sum-rate.