Traffic Prediction-based Dynamic Time Slot Allocation TDMA Protocol

Considering the channel waste resulted by the fixed or certainty traffic-based time slot assignment in uncertainty traffic flows TDMA networks, a traffic prediction-based dynamic time slot allocation TDMA protocol (PD-TDMA) is proposed. Coordinator collects the transmission queue length information which sanded by all nodes with the CSMA mechanism, and then it predicts the real time arrival traffic with the measure of wavelet neural network and makes a dynamic time slot allocation according to greedy algorithm. The simulation results show that the proposed protocol can guarantee better performance on predicting real-time traffic, improving the channel utilization and reducing delay than the conventional CSMA/CA and hybrid MAC protocol.