In recent years, the edge computing paradigm enables the movement of processing units and storage nearer to the data available locations. The mechanism completes the computation in a short span of time in minimum bandwidth. Edge ecosystem is a type of distributed computing that is sensitive to topology and geography; the Internet of Things is a prime instance of this. Rather than referring to a single technology, Edge computing refers to an architecture. This paper proposes a resource allocation methodology that will enliven the situation between users and edge servers. By creating continuous control at the edge servers to determine resource allocation, edge computing improves reaction time, provides high security with decreased risk, scalability, lowers transmission costs, and versatility (offload targets, migration bandwidth and computing resources). The Deterministic Policy Gradient, Deep learning and Quality Network concepts are combined in the proposed system. The continuous action space is achieved by a deterministic policy gradient. The experience relay includes a quality network. In the proposed system, the actor-critic network produces a single continuous action instead of resulting probability based actions. The critic-part uses Q-value from a quality network based on current status and activity. The goal of the proposed system is to develop a Deep Deterministic Policy Gradient methodology to allocate servers for mobile users with the help of the Edge computing while taking computation resources, offloading goals and migration bandwidth into consideration. The simulation result indicates that deterministic policy gradients integrated deep learning models improve the system performance compared with Game theory.