In this Proposed work, the problem of resource al location for device-to-device (D2D) communication in a Third Generation Partnership Project (3GPP) Long-Term Evolution Advanced (LTE-A) network. The users in the network can operate either in a traditional cellular mode, communicating with each other via the evolved NodeB (eNB), or in a D2D mode, communicating with each other without traversing the eNB. In the considered model, the D2D users and cellular users share the same radio resources. Particularly, each resource block (RB) within the available bandwidth can be occupied by one cellular and several D2D users. Hence, the problem of interference management is crucial for effective performance of such a network. The twofold aim of the proposed algorithm is to 1) mitigate the interference between cellular and D2D users and 2) improve the overall user-perceived quality of service (QoS). To control the interference, for each user, we define a certain target interference level and constrain the interference from the other users to stay below this level. The corresponding optimization problem maximizes the QoS of the users by minimizing the size of the buffers of user equipments (UEs). The performance of the algorithm has been evaluated by using the OPNET-based simulations. The algorithm shows improved performance in terms of mean packet end-to-end delay and loss for UEs when compared to other relevant schemes.