Fluctuation of the received signal strength (RSS) is the key performance-limiting factor for Wi-Fi indoor positioning schemes. In this study, the Manhattan distance was used in the weighted K-nearest neighbour (WKNN) algorithm to improve positioning accuracy. Reference point (RP) intervals were optimised to reduce the complexity of the system. Specifically, two new positioning schemes are proposed in this paper. Scheme 1 uses the cellular network to refine the fingerprint database, while Scheme 2 uses the cellular network positioning to locate the node a priori, then uses the Wi-Fi network to further improve accuracy. The experimental results showed that the average positioning error of Scheme 1 was 1·60 m, a reduction of 12% compared with the existing Wi-Fi fingerprinting schemes. In Scheme 2, when double cellular networks were used, RP usage was reduced by 64% and the calculating time was 0·24 s, a reduction of up to 69·5% compared with the Manhattan-WKNN algorithm. These proposed schemes are suitable for high accuracy and real-time positioning situations, respectively.

中文翻译：

---

一种新的蜂窝和 Wi-Fi 网络室内定位算法

接收信号强度 (RSS) 的波动是 Wi-Fi 室内定位方案的关键性能限制因素。在这项研究中，曼哈顿距离被用于加权ķ-最近邻（WKNN）算法，以提高定位精度。参考点 (RP) 间隔进行了优化，以降低系统的复杂性。具体来说，本文提出了两种新的定位方案。方案一利用蜂窝网络对指纹库进行细化，方案二利用蜂窝网络定位先验定位节点，再利用Wi-Fi网络进一步提高精度。实验结果表明，方案一的平均定位误差为1·60 m，与现有的Wi-Fi指纹方案相比降低了12%。在方案2中，当使用双蜂窝网络时，RP的使用减少了64%，计算时间为0·24 s，与Manhattan-WKNN算法相比减少了69·5%。