The Personalized Priority Routing Algorithm In Publish/Subscribe Network

As the various urgency and importance of the messages in the pub/subnetwork, this paper proposes a feedback-based personalized priority routingalgorithm in the purpose of service differentiation. It integrates with thecontent-cover-based routing algorithm and allows the dynamic change of themessage priority. Moreover, the user can have different weight as their rightto access to different level of service. And they can make personalizedsubscriptions as well, which makes those subscriptions of more subscribersand importance to achieve higher priority (or transmission quality). Besides,in case some users might make high priority subscriptions greedily, thefeedback-based priority scheduling algorithm is introduced to ensure thefairness. And the priority-considered super node selection algorithm is alsointroduced to minimize the side-effect of feedback-based priority schedulingmechnism.The experiments show that the super node selection algorithmdoes select qualified super nodes and maintain priority balance. And with priority routing algorithm, high priority messages do have much betterperformance. In addition, the feedback-based priority scheduling is proved tobe able to maintain system fairness as well, which eventually will allow thepractical applicability of the personalized priority routing algorithm.