<p>Mobile Edge Computing (MEC) and Network Function Virtualization (NFV), combined as complementary paradigms, are deemed as a scalable solution for high-quality and low-delay services provision and user service experience enhancement. Different users’ requests demand different network services with different resource and delay requirements which are associated with different Virtualized Network Functions (VNFs). Currently, in order to ensure the quality of service, researchers focus more on the quality of service issues and ignore the key factor of server utilization. Nevertheless, delay-sensitive and computation-intensive service requests pose a challenge to provide high-quality and low-latency services which meet users’ delay requirements and server utilization enhancement at the same time. In this paper, we first expound on user satisfaction with services in the MEC network and formulate a novel server utilization enhancement problem in the awareness of user satisfaction. We focus on the optimization of edge server utilization when providing virtualized network function services to mobile users in MEC, while considering user mobility and service delay requirements. We propose an Edge Server Utilization Optimization Online Algorithm (USP-SUO) based on user satisfaction perception and demonstrate through simulation experiments that USP-SUO can effectively improve the utilization of edge servers and keep the utilization in a good state.</p>