Font Size: a A A

A Particle Swarm Optimization Algorithm For RFID Network Planning With Load Balancing Constraint

Posted on:2015-10-31Degree:MasterType:Thesis
Country:ChinaCandidate:Y L LiFull Text:PDF
GTID:2308330464968666Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
In recent years, RFID(Radio Frequency Identification) techno logy attracts more and more researchers’ attention as a representative of networking technology. It also involves many aspects of technical issue in the RFID system, such as network planning, collision and anti- collis ion safety techno logy, frequency selection, antenna techno logy, low- power technology, packaging technology, location tracking and so on. Radio Frequency Identification(RFID) has a widespread application in reality, and RFID network planning(RNP) problem is a core challenge in the deployment of RFID networks. The RFID network planning optimization problem involves many constraints and objectives, and has been proven to be NP- hard problem. The aim of RFID network planning optimizat ion problems is to find the bes t locations of readers, and the locations of readers must satisfy constrains, such as tag coverage, the number of readers, interference and load balancing. The main work of this paper is as follows:(1) By analys is of the RFID network planning optimization problem, we know that RFID network planning optimizat ion problem actually is a comb ination of a mult i-objective optimization problem, where the side constraints need to consider three objectives, namely, coverage, interference label reader, the number of readers. According to the optimizat ion proble m of RFID network planning, we propose a method based on partic le swarm optimization algorit hm Mult i- group symb iosis. This method effectively controls the premature of PSO algorit hm. Meanwhile, for the selection of the number of RFID network planning optimization, we introduce Tentative Reader Elimination operator. The operator is able to inte lligent ly adjust the number of readers deployed in the optimizat ion process. Finally, experiments demonstrate the effect iveness of the optimizat ion method based on populat ions symbiotic particle swarm algorit hm proposed by us, and it makes the RFID network planning optimizat ion problems adjust the number of readers of the real needs adaptive ly.(2) The basic problem o f RFID network planning optimizat ion invo lves three main objectives constraints : the number of tag co verage, reader interference, the number of readers. But wit h the cont inuous development of RFID technology, RFID network scale is also growing. In such a large system, each RFID reader communicates with mult iple tags in the vic init y simultaneously. If the accident of some load imbalance happens, it will result in a failure to communicate wit h mult ip le tags at the same time and a sharp decrease in Tag coverage. Therefore, taking into account of RFID network planning optimization problem and load balancing constraints is essentia l. Thus, in the second work, we will introduce the balance constraint of network into the RFID network planning optimizat ion proble m. Aiming at the optimization of RFID network planning based on load balancing, we use the standard PSO optimization algorithms and temptations eliminate operator(TRE), which can dynamically adjust the number of readers. We find the best configuration of the RFI D reader deployed, the optima l tag coverage, the number of reader, the load balancing and the constraint to interference. The experimenta l simulat ion shows that the algorithm can achieve s high coverage while using relatively fewer readers, high load balanc ing and low interference.
Keywords/Search Tags:Radio Frequency Identific ation, Multi-group symbiotic algorithm, Tentative Reader Elimination operator, Particle swarm optim ization algorithm, Load balancing
PDF Full Text Request
Related items