Paper Info
Title
Comparative performance analysis of various artificial immune networks applied to RFID reader-to-reader collision avoidance
Abstract
The reader-to-reader anti-collision model is translated as an optimization problem.A summary review of six artificial immune networks is proposed.The major immune operators are redesigned by following the initial spirits.The effects of multiple variables are investigated.The algorithmic computational efficiency is examined. Within a radio frequency identification (RFID) system, the reader-to-reader collision problem may occur when a group of readers operate simultaneously. The scheduling-based family, as one branch of RIFD reader collision avoidance methods, focuses on the allocation of time slots and frequency channels to RFID readers. Generally, the RFID reader-to-reader collision avoidance model can be translated as an optimization problem related with the communication resource allocation by maximizing the total effective interrogation area. Artificial immune networks are emerging heuristic evolutionary algorithms, which have been broadly applied to scientific computing and engineering applications. Since the first version of artificial immune networks for optimization occurred, a series of revised or derived artificial immune networks have been developed which aim at capturing more accurate solutions at higher convergence speed. For the RFID reader-to-reader collision avoidance model, this paper attempts to investigate the performance of six artificial immune networks in allocating communication resources to multiple readers. By following the spirits of artificial immune networks, the corresponding major immune operators are redesigned to satisfy the practice of RFID systems. By taking into account the effects of time slots and frequency channels, respectively, two groups of simulation experiments are arranged to examine the effectiveness of different artificial immune networks in optimizing the total effective interrogation area. Besides, a group of examination is executed to investigate the performance of six algorithms in solving different dimensionality of solution space in reader collision avoidance model. Meanwhile, a single group of simulation experiments are arranged to examine the computational efficiency of six artificial immune networks. The results demonstrate that six artificial immune networks perform well in searching the maximum total effective interrogation and are suitable to solve the RFID reader-to-reader collision avoidance model.
Year
DOI
Venue
2015
10.1016/j.asoc.2015.04.006
Applied Soft Computing
Keywords
Field
DocType
Artificial immune network,Comparative analysis,Optimization,Reader-to-reader collision avoidance,Resource allocation,RFID system
Heuristic,Evolutionary algorithm,Scheduling (computing),Collision,Resource allocation,Artificial intelligence,Collision problem,Optimization problem,Radio-frequency identification,Machine learning,Mathematics
Journal
Volume
Issue
ISSN
32
C
1568-4946
Citations 
PageRank 
References 
2
0.39
26
Authors
5
Name
Order
Citations
PageRank
Zhonghua Li1384.60
Jianming Li220.39
Shanjin Wang320.72
Chengpei Tang4434.45
Minghua Xia520.72