Meta-heuristic Search For Radio Frequency Identification Network Planning And Charging

Radio frequency identification(RFID)is a non-contact automatic identification technology with features such as high speed,and strong anti-interference ability.A typical RFID system usually consists of three parts: readers,tags,and a central control system.In practical applications,since the power radius of the reader is limited and there are many tags distributed in the working space,the layout of readers need to be determined.RFID network planning is proposed to study how to lay out readers and determine the number of readers.Our purpose is to effectively use resources and improve the efficiency and stability of RFID systems.Due to the inefficiency of existing methods which can only deal with small-scale data,this thesis discusses the RFID network planning problem(RNP)and wireless rechargeable sensor network problem(WRSNs).The main works are summarized as follows:(1)An efficient and fast kinematics-based algorithm for RFID network planning is proposed.We design a curling algorithm for RFID network planning from the viewpoint of kinematics.Each reader is considered as a curling which can slide on the working space for searching a property location,and tags on the working space generate friction to stop readers.We design the reader movement operator,the reader collision operator and the revised strategy of friction.Each time a reader is put into the working space,then it can move and stop because of the friction of tags.More readers are put into the working space until enough tags have been covered.We also design a redundant elimination operator to improve the efficiency of our algorithm.The results illustrate the effectiveness of curling algorithm which has good performance in searching locations and number of readers.(2)A memetic algorithm for RFID Network Planning is designed.The main methods for RFID network planning are evolutionary algorithm and swarm intelligence algorithm.Due to the differences in individuals,both of these kinds of algorithms can promote the evolution of populations by taking advantage of some excellent individuals.It is considered that there may exist some useful information in each individual for self-improvement.Each individual contains information of several readers.According to this idea,we design two local search operators,i.e.the attractive operator and the repulsive operator.We combine these two operators with genetic algorithms to optimize the planning of RFID network.The results illustrate that our local search strategy can improve the performance of original algorithms and can find the solution with better convergence and higher retrieval efficiency.(3)A memetic algorithm for optimal charging in wireless rechargeable sensor networks is designed.To elongate the lifecycle of wireless sensor networks,we apply a mobile RFID reader to charge all sensors in the working space.We design a memetic algorithm to solve this problem.An individual modification operator and a local search operator are designed to improve the performance of our algorithm.We simultaneously optimize the charging delay,total cost of charging energy and the maximum number of fully charged nodes.Simulation results are offered to demonstrate the effectiveness of our algorithm.