A Distributed Publish-Subscribe Algorithm Based on Spatial Text Information Flow

With the rapid development of society and the popularity of smart devices, the volume of information sent and received is increasing day by day. It has become very difficult to accurately and efficiently match a large number of events with a large number of subscriptions, and the event and subscription matching speed can no longer meet demand. To speed up the matching speed of events and subscriptions, this paper uses similarity and correlation to optimize the clustering operation and increase the data transfer and data throughput of the category fusion strategy. Firstly, the clustering operation is performed on the subscription messages, and the category to which the events belong is found according to the clustering result. Subsequently, in the category, the subscriptions matching the events are found. An on-the-fly subscription publishing algorithm is proposed to coordinate spatial information and event attribute information to handle not only the matching operation of events and subscriptions on-the-fly but also to perform subscription updates and category updates on the distributed environment on-the-fly. It can also perform clustering operations and matching operations instantly without prior knowledge. We design a distributed system for the publish-subscribe algorithm and propose a load balancing strategy for this algorithm on the distributed system. Subsequently, we experimentally validate the proposed publish-subscribe algorithm in this paper by building our own cluster and using real data. perform clustering operations and matching operations instantly without prior knowledge. We design a distributed system for the publish-subscribe algorithm and propose a load balancing strategy for this algorithm on the distributed system. Subsequently, we experimentally validate the proposed publish-subscribe algorithm in this paper by building our own cluster and using real data.