A Real-Time C-V2X Beamforming Selector Based on Effective Sequence to Sequence Prediction Model Using Transitional Matrix Hard Attention

For C-V2X systems, the selection of best beam in a real-time mode becomes an increasingly critical and yet open topic. Most existing approaches adopt either conventional ARIMA or ANN. Recently, there has been research on adopting sequence-to-sequence (Seq2Seq) predictors with attentions to extract time series features emphasis information achieve data prediction. In this paper, Seq2Seq predictor integrating Transitional Matrix based Hard attention is presented validated through artificial test dataset predefined transitional states. At first, transition probability matrix generated from previous fed into “hard” module determine weights during training phase. Secondly, was implemented adopted predict beams beamforming selector built up by authors. Experiments were conducted captured used validate performance predictor. When compared baseline models, can enhanced prediction accuracy gain 10-12%.