In response to the difficulties in setting the peak detection threshold for signal acquisition and the decrease in acquisition accuracy of satellite navigation receivers in dynamic scenarios or scenarios with weak navigation signal strength, a satellite navigation signal acquisition peak detection method based on improved Support Vector Machine (SVM) is proposed. This method first reduces the dimensionality of sample features through Principal Component Analysis (PCA), then classifies the acquisition correlation results of satellite navigation signals, and finally determines whether the navigation signal is successfully acquired by judging whether there is a peak in the correlation results. Simulation results show that, compared with existing traditional threshold setting methods, standard SVM methods, and logistic regression classification learning methods, the detection method proposed in this paper has the advantages of low false alarm rate and high true alarm rate, and the acquisition success rate is also better than existing methods.