Based on the resource slicing strategy of deep reinforcement learning, this paper proposes a method framework for emergency Internet of Things, No.EIoT) slice resource reservation and multi-heterogeneous slice resource sharing and isolation. In view of the differentiated service quality requirements of multiple network slices, and the different requirements of different slices for speed and delay indicators, a shape-based two-dimensional knapsack problem is used and heuristic algorithm is used to numerically solve it. Experimental results show that, compared with the traditional NVS, No.network virtualization substrate) and Netshare algorithms, the Dueling DQN algorithm is better, effectively balancing the performance of heterogeneous coexisting slices.

Cluster analysis can find not only potential and valuable structured information in the data set, but also provide pre-processing functions for other data mining algorithms, and then can refine the processing results to improve the accuracy of the algorithm. Therefore, cluster analysis has become one of the hot research topics in the field of data mining. K-means algorithm, as a clustering algorithm based on the partitioning idea, can compare the differences between the data set classes and classes. We can use the K-means algorithm to mine the clustering results and further discover the potentially valuable knowledge in the data set. Help people make more accurate decisions. This paper summarizes and analyzes the traditional K-means algorithm, summarizes the improvement direction of the K-means algorithm, fully considers the dynamic change of information in the K-means clustering process, and reduces the standard setting value for the termination condition of the algorithm to reduce The number of iterations of the algorithm reduces the learning time; the redundant information generated by the dynamic change of information is deleted to reduce the interference in the dynamic clustering process, so that the algorithm achieves a more accurate and efficient clustering effect. Experimental results show that when the amount of data is large, compared with the traditional K-means algorithm, the improved K-means algorithm has a greater improvement in accuracy and execution efficiency.

No abstract

In autonomous driving path planning, ensuring the computational efficiency and safety of planning is an important issue. The Dyna framework in reinforcement learning can solve the problem of planning efficiency. At the same time, the Sarsa algorithm in reinforcement learning can be effective in guaranteeing the safety of path planning. This paper proposes a path planning algorithm based on Sarsa-Dyna for autonomous driving, which effectively guarantees the efficiency and safety of path planning. The results show that the number of steps planned in advance is proportional to the convergence speed of the reinforcement learning algorithm. The Sarsa-Dyna will be proposed. The analysis of convergence speed and collision times has been done between the proposed Sarsa-Dyna, Q-learning, Sarsa and Dyna-Q algorithm. The proposed Sarsa-Dyna algorithm can reduce the number of collisions effectively, ensure safety during driving, and at the same time ensure convergence speed.