Transcranial alternating current stimulation (tACS) is considered to have a positive effect on the rehabilitation of Alzheimer's disease (AD) as an intervention method that matches stimulation frequency to neurogenesis frequency. However, when tACS intervention is delivered to a single target, the current received by brain regions outside the target may be insufficient to trigger neural activity, compromising the effectiveness of stimulation. Therefore, it is worth studying how single-target tACS restores gamma-band activity in the whole hippocampal-prefrontal circuit during rehabilitation. We used Sim4Life software to conduct finite element methods (FEM) on the stimulation parameters to ensure that tACS intervened only in the right hippocampus (rHPC) and did not activate the left hippocampus (lHPC) or prefrontal cortex (PFC). We stimulated the rHPC by tACS for 21 days to improve the memory function of AD mice. We simultaneously recorded local field potentials (LFPs) in the rHP, lHPC and PFC and evaluated the neural rehabilitative effect of tACS stimulation with power spectral density (PSD), cross-frequency coupling (CFC) and Granger causality. Compared to the untreated group, the tACS group exhibited an increase in the Granger causality connection and CFC between the rHPC and PFC, a decrease in those between the lHPC and PFC, and enhanced performance on the Y-maze test. These results suggest that tACS may serve as a noninvasive method for Alzheimer's disease rehabilitation by ameliorating abnormal gamma oscillation in the hippocampal-prefrontal circuit.