To defend against the eavesdropping and spoofing attacks, the physical layer authentication technique utilizes the unique attributes of channel or device for identifying attackers. In this paper, we investigate an analytical method for the physical layer authentication technique utilizing channel phase. Specifically, we simplify the expressions of the complex binary hypothesis variables which are available for calculating probability density functions (PDFs). Based on the PDFs, we further derive the closed-form expressions of expectation and variance under the binary hypothesis test of $\mathcal{H}_{0}$ and $\mathcal{H}_{1}$ as well as the receiver operator characteristic (ROC) curve for performance evaluation. Simulation results show that the theoretical curves are tight even under very low signal-to-noise ratio (SNR) regions, which outperforms the benchmark.