To enhance the efficiency and expediency of issuing e-licenses within the power sector, we must confront the challenge of managing the surging demand for data traffic. Within this realm, the network imposes stringent Quality of Service (QoS) requirements, revealing the inadequacies of traditional routing allocation mechanisms in accommodating such extensive data flows. In response to the imperative of handling a substantial influx of data requests promptly and alleviating the constraints of existing technologies and network congestion, we present an architecture for QoS routing optimization with in Software Defined Network (SDN), leveraging deep reinforcement learning. This innovative approach entails the separation of SDN control and transmission functionalities, centralizing control over data forwarding while integrating deep reinforcement learning for informed routing decisions. By factoring in considerations such as delay, bandwidth, jitter rate, and packet loss rate, we design a reward function to guide the Deep Deterministic Policy Gradient (DDPG) algorithm in learning the optimal routing strategy to furnish superior QoS provision. In our empirical investigations, we juxtapose the performance of Deep Reinforcement Learning (DRL) against that of Shortest Path (SP) algorithms in terms of data packet transmission delay. The experimental simulation results show that our proposed algorithm has significant efficacy in reducing network delay and improving the overall transmission efficiency, which is superior to the traditional methods.

Federated Learning (FL) is a framework for machine learning on a large-scale distributed dataset, enabling the training of a collaborative model across multiple parties while preserving the privacy of user data. However, in cases where data are distributed in a non-independent and identically distributed (non-iid) manner, the convergence speed of the federated collaborative model and its prediction accuracy on client nodes can be significantly affected. Therefore, personalized FL methods have emerged to further adapt to the data characteristics of different clients. In response to the data heterogeneity issue, this paper presents a multi-task clustering-based personalized federated learning algorithm, which is applied to the prediction of carbon emissions in different regions and enterprises. This algorithm partitions nodes with similar data distributions and aggregates local models within the same cluster to form cluster models. It introduces the concept of multi-task learning, dividing the lower layers of cluster models into expert layers. These expert layers of different cluster models are then weighted and aggregated for the acquisition of global knowledge. Additionally, adaptive weight is applied to control the aggregation of expert layers, thereby achieving model personalization at the local level. Simulation experiments conducted on carbon emission prediction data demonstrate that the proposed algorithm performs better in various evaluation metrics compared with the Federated Averaging (FedAvg) algorithm and traditional clustering personalized federated learning algorithm. It also exhibits excellent experimental results and performance when dealing with different quantities of heterogeneous data distributions.

The internet is an abundant source of news every day. Thus, efficient algorithms to extract keywords from the text are important to obtain information quickly. However, the precision and recall of mature keyword extraction algorithms need improvement. TextRank, which is derived from the PageRank algorithm, uses word graphs to spread the weight of words. The keyword weight propagation in TextRank focuses only on word frequency. To improve the performance of the algorithm, we propose Semantic Clustering TextRank (SCTR), a semantic clustering news keyword extraction algorithm based on TextRank. Firstly, the word vectors generated by the Bidirectional Encoder Representation from Transformers (BERT) model are used to perform k-means clustering to represent semantic clustering. Then, the clustering results are used to construct a TextRank weight transfer probability matrix. Finally, iterative calculation of word graphs and extraction of keywords are performed. The test target of this experiment is a Chinese news library. The results of the experiment conducted on this text set show that the SCTR algorithm has greater precision, recall, and F1 value than the traditional TextRank and Term Frequency-Inverse Document Frequency (TF-IDF) algorithms.