DSSS signal is a technical means used for spread spectrum communication. Due to its unique frequency characteristics, it has great anti-interference and multipath resistance capabilities, and has many advantages such as a wide spectrum, making it an important enabling technology for spread spectrum communication. However, the relevant characteristics of spread spectrum signals pose challenges in detecting spread spectrum signals from conventional signals. Based on this, this article selects a feature value based on quadratic power spectrum to distinguish between conventional signals and spread spectrum signals after binary phase shift keying (BPSK) modulation, and selects decision boundaries through a large number of training sets, and tests the model. The results show that the spread spectrum signal detection method based on the secondary power spectrum has good detection accuracy and performance, with a discrimination accuracy of 99.44% and 99.6% for the training and testing sets, respectively, verifying the feasibility of this detection method.