The long short-term memory (LSTM) network is especially suitable for dealing with time series-related problems, which has led to a wide range of applications in analyzing stock market quotations and predicting future price trends. However, the selection of hyperparameters in LSTM networks was often based on subjective experience and existing research. The inability to determine the optimal values of the parameters results in a reduced generalization capability of the model. Therefore, we proposed a sparrow search algorithm-optimized LSTM (SSA-LSTM) model for stock trend prediction. The SSA was used to find the optimal hyperparameters of the LSTM model to adapt the features of the data to the structure of the model, so as to construct a highly accurate stock trend prediction model. With the Shanghai Composite Index stock data in the last decade, the mean absolute percentage error, root mean square error, mean absolute error, and coefficient of determination between stock prices predicted by the SSA-LSTM method and actual prices are 0.0093, 41.9505, 30.5300, and 0.9754. The result indicates that the proposed model possesses higher forecasting precision than other traditional stock forecasting methods and enhances the interpretability of the network model structure and parameters.

With the increasing number of fibre access users, P2P file sharing download traffic has gradually become one of the enormous bandwidth consumptions for the access network. Localizing traffic could save bandwidth and reduce latency, but it caused optical line terminal (OLT) overload problems. A cloud‐edge collaboration technology (IRS‐CECT) based P2P redirection strategy is proposed for this problem. OLT assumes the role of the cloud computing centre in the strategy. Optical network unit (ONU) serves as an edge computing node to share traffic load for the cloud computing centre. ONU intercepts the upstream query packets for OLT. This strategy can reduce the problem of OLT overload caused by increasing users. Simulation results indicate that the IRS‐CECT strategy can reduce the delay by nearly half and reduce the traffic load on the cloud computing centre OLT.

In the development of ethernet passive optical networks (EPONs), quality of service (QoS) support and fairness per optical network unit (ONU) are crucial issues. However, making an elaborate analysis of the existing prediction-based bandwidth allocation algorithm, light load penalty, low prediction precision are pointed out. We present an improved dynamic bandwidth pre-allocation algorithm (R-DBA), which employs recurrent neural network (RNN) to predict the high-priority service traffic in EPON. And we introduce mixed integer linear programming (MILP) for optimally building DBA algorithm. This algorithm achieves the prediction of the high-priority service traffic by RNN during the waiting time and supports bandwidth pre-allocation, thus ensuring the fairness of the bandwidth allocation.