In WLAN position fingerprint localization algorithms , K-means clustering algorithm is always used to cluster the RPs in the offline training phase. Aiming at the fact that it is very sensitive to noise and outliers, this thesis uses Fast K-medoids algorithm to cluster the RP. In the offline phase, the Fast K-medoids RP clustering algorithm selects initial medoids among the RP fingerprints first , then finds the medoid of each cluster, which is the fingerprint minimizing the total distance to other fingerprints in its cluster by using iterative method. Finally , the experiment indicates that Fast K-medoids based position fingerprint localization algorithm has greater robustness and localization accuracy than K-means based position fingerprint localization algorithm in the view of average error , STD and CDF.%在WLAN位置指纹定位技术中,K-means聚类算法一直被用于离线训练阶段的参考点聚类,文中针对该法对噪声数据和孤立点数据非常敏感等缺点,采用快速K-medoids聚类算法来对定位区域内的参考点进行聚类.快速K-medoids参考点聚类算法先选取初始类中心参考点,再通过迭代方式在每一类中选取与其他位置指纹信息距离之和最小的那条位置指纹信息对应的参考点作为类中心参考点.最后通过实验数据分析表明,相比K-means参考点聚类算法,从平均误差、标准差和累积误差曲线图3个方面可以看出快速K-medoids参考点聚类算法在去除噪声数据和孤立点数据上具有更好的鲁棒性,可有效地提升定位精度.
展开▼
