Research of synchronization and training sequence design for cooperative D2D communications underlaying hyper-cellular networks

In this paper, new training sequence structure with adjustable parameters is proposed for the multiple timing offsets and frequency offsets estimation of cooperative device-to-device (D2D) systems underlaying hyper-cellular networks. By using the proposed training sequences, timing offsets can be estimated using a cross-correlation approach with good performance, and frequency offsets can be estimated using a signal subspace decomposition approach with low complexity. By designing the parameters of the proposed training sequences properly, estimation outliers can be effectively avoided.