Multi-reader's Collision Avoidance Model And Its Intelligent Optimizer

With the rapid development of the Internet of Things(IOT),RFID reader as the bottom equipment of IOT plays an important role for accurate data collection,data analysis and applications.It has become a key problem to design applicable reader collision avoidance planning models and effective algorithms.Therefore,for the problem of RF signal interference and collision caused by multiple interactive readers with overlapped coverage areas and competitively shared media in static or dynamic environments,this thesis develops readers' collision avoidance and scheduling models,while exploring the related bi-level hybrid intelligent optimization algorithms.The acquired achievements are summarized as follows: 1.For the scheduling problem of multiple readers in static environments,a bi-level and multi-reader scheduling model is established based on the total area of readers' recognition range as the performance index.An immune fish swarm optimization algorithm is developed,after the crowding degree and rear-end behavior of a fish swarm algorithm are embedded into an immune optimization algorithm in order to strengthen the ability of local exploitation.Hereafter,a hybrid intelligent optimization algorithm is obtained to solve the scheduling model,after the acquired algorithm as an operator module is embedded into the genetic algorithm.Comparative experiments show that the obtained model is rational while the latter algorithm has obvious advantages over the compared approaches with the aspects of search effect,stability,convergence speed and the scheme's rationality.2.For the problem of readers' collision avoidance and scheduling of known in multi-slot environments,a bi-level planning model,which takes the total recognition area of the readers as the performance index within the specified time slot,is designed to formulate the scheduling problem of known readers with collision avoidance in multi-slot environments.It includes some constraints which the readers need to have minimal recognition ranges.Subsequently,on the basis of the above immune fish swarm optimization algorithm and the genetic algorithm,a hybrid optimization algorithm for solving the gained model is established by introducing the strategies of fitness sharing and chaos variation.Comparative experiments show that the designed model is reasonable while the algorithm performs well over the compared approaches and also the obtained reader scheduling scheme is rational.3.A bi-level programming model is developed to formulate the problem of collision avoidance of readers in dynamic environments,in which the total recognition area of the readers is regarded as the performance index.It involves the factor of which the number of the readers may keep variable.Then,on the basis of the above immune fish swarm optimization and genetic algorithm,a hybrid optimization algorithm is designed to tackle the acquired dynamic scheduling model,depending on the strategies of environment detection and reverse learning mechanism.Comparative experiments validates that the derived model is reasonable while the algorithm can detect the change of the environment and also behave well over the compared approaches.