Ad Hoc network traffic prediction based on the Elman neural network

For complex Ad Hoc network environment, based on Elman neural network prediction model, using particle swarm optimization algorithm to optimize the original Elman BP training can find the global optimization thresholds and weights of the neural network layers, and a modified Elman neural network model is proposed to predict the Ad Hoc network node traffic. According to Ad Hoc network node traffic data obtained by simulation in NS-2, network traffic predict experiment results show that the modified Elman neural network model has improved compared to the previous Elman neural network model and has a lower error and higher prediction accuracy.