自适应高斯遍历和声搜索物联网射频识别均衡优化

Adaptive Optimization of Radio Frequency Identification with Gauss Traversal and Harmony Search Network in Internet of Things

针对物联网射频识别过程中存在的数据量过大、传统算法计算复杂度较高和识别准确率较低的问题,提出了自适应高斯遍历和声搜索(Gauss traversal and harmony search algorithm,GTHS)物联网射频识别优化算法.首先,基于和声搜索算法进行网络优化设计,针对标准HS在优化精度和计算复杂度等方面存在的问题,利用高斯函数的遍历特性对算法即兴创作过程引入控制参数,提高前后期搜索的针对性,并给出参数选取的理论分析;其次,对物联网射频识别优化模型进行研究,提出改进的自适应优化目标,实现性能指标的均衡优化;最后,将该算法与RPSOAS、CDE以及C-MC算法进行了实验对比分析,结果表明,所提GTHS算法在区域大小为1 000 m×1 000 m、标签数量为100 000的大型物联网RFID(radio frequency identification network)实验对象中,收敛精度为7.215 6,收敛精度提高29.6%以上.

To handle the large amount of data and complexity in the radio frequency identification process of the Internet of things, and overcome the disadvantages of lower complexity and recognition accuracy in the traditional algorithms, an algorithm was proposed to achieve the adaptive optimization of radio frequency identification with Gauss traversal and harmony search network in internet of things. First, the harmony search algorithm was used to optimize the design of the network. In order to solve the problems of the low optimizing accuracy and high computational complexity, the ergodicity of Gaussian function was applied to introducing control parameter in the improvisation process of the algorithm, which improves the pertinence for different evolution periods. The theoretical analysis for parameter selection was presented as well. Secondly, the optimization model of the radio frequency identification for the Internet of things was studied, and the improved adaptive optimization objective was proposed to achieve the equilibrium optimization of the performance index. Finally, the proposed algorithm was compared with the RPSOAS, CDE and C-MC algorithm, showing that, in the RFID （radio frequency identification network） experiment with 1 000 m × 1 000 m region and the 100 000 tags, the convergence rate is 7. 215 6, increased by more than 29.6%.