工业物联网中基于多维特征业务的网络切片匹配算法

工业物联网中的业务具有数量大、种类多、特征维度高等特点，为了实现每一个业务都能根据自身的多维特征有效的匹配到具有特定服务质量的网络切片以获取相应的服务，提出了一种多维特征聚类的业务与网络切片匹配算法。该算法可以对工业物联网设备中提取的业务多维特征进行聚类，根据聚类结果将所定义切片的优先级因子添加到相应的业务中，进而实现业务与网络切片的匹配。实验结果表明，该算法在实现业务与网络切片匹配的过程中，与对比算法相比，既实现了高准确率，又加快了匹配完成的速度。

Network slices matching algorithm based on multi-dimensional feature service for industrial Internet of Things

The services in the Industrial Internet of Things are characterized by a large number, variety, and high feature dimensions. In order to realize that each service can effectively match to the network slicing with specific service quality according to its own multi-dimensional features to obtain the corresponding services, this paper proposed a multi-dimensional feature clustering services and slices matching algorithm. The algorithm clustered the multi-dimensional features of services extracted from industrial Internet of Things devices, and added the priority factor of the defined slices to the corresponding services according to the clustering results, so as to realize the matching between services and slices. Experimental results show that the algorithm not only achieves high accuracy, but also accelerates the speed of matching compared with the comparison algorithm.