摘要
根据位置指纹室内定位算法的理念,提出了一种旨在减小计算量的定位方法,并将此方法应用于KNN算法中。以KNN算法为例,理论上分析了其计算量优化的情况,并在此优化算法的基础上,通过仿真比较了K的取值、AP节点的位置及数量对定位精度的影响。结果表明该算法不但能够保证位置指纹室内定位的精度,而且还能有效的减小定位过程中的计算量。该方法同样可以推广到其他位置指纹定位算法中,能在理论上解决位置指纹定位算法的计算量问题。
According to the concept of location algorithm for fingerprint-based indoor position. A novel computing method is applied to k-nearest neighbors （KNN） algorithm for reducing the computational complexity. Take the KNN algorithm for example, analyzed the computational complexity theoretically, and compared the different value of K, the position and number of AP on the influence to the precision of positioning by simulating on the basic of this algorithm. Simulation results indicate that the better location performance can be achieved and the computational complexity is reduced by proposed KNN algorithm, and this algorithm is also suitable for other location systems to reduce the computational complexity.
出处
《电子设计工程》
2013年第7期44-46,共3页
Electronic Design Engineering
基金
国家自然科学基金资助项目(61271279
61201157)
关键词
室内定位
位置指纹
计算量
KNN
indoor location
location fingerprint
computational
KNN
