Sixth Generation (6G) network is meant to allow wireless networking and computing by digitalizing and sharing everything, by providing a computer image of the actual network world. Mobile edge computation as one of the key factors in allowing mobile downloads faces unparalleled obstacles because the 6G network environment is incredibly dynamic and unforeseeable. In the latest literature on mobile edge computing, the implications of user mobility and the volatile mobile edge computing world are still ignored. In this paper, we propose a new methodology for the Digital Twin mirror that offers training data to offload decisions for digital edge servers to evaluate the edge servers’ status and the Digital Twin for the whole edge computing environment. In the wireless twin edge networks, the proposed system is to reduce the download delay in the face of the cumulative expense of relocation from the accessed service Mobility for consumers. The Lyapunov approach’s Optimization is used to simplify the cost constraint of Long-term transformation to an intra-functional enhancement challenge, which is then resolved by profoundly enhanced Actor-Critic (AC) learning. Replications demonstrate that, as opposed to benchmark systems, our proposed arrangements effectively decrease the average offload delay, discharge failure rate and operation migration rate and save device costs with Digital Twin help.