With mobile cloud computing system architecture gradually replacing the traditional Client-Server model, one of the most critical issues that needs to be addressed is the resource allocation deciding problem to resolve how a cloud could efficiently handle the cloud resources to satisfy the requirements of mobile device for cloud services. Simoultaneously, this is in effort to obtain the maximal utilization rate of cloud resources and system rewards of cloud. The study first proposes a novel cloud computing resource allocation model based on semi-Markov decision process (SMDP) to achieve the optimal resource allocation scheme in terms of maximal system reward, cloud resource utilization, and the QoS of mobile device while capturing the dynamics of resource request arrivals and departures. Finally, the performance of the proposed model is evaluated by the simulation results, which show that the obtained theoretic results are consistent with this simulation results.