• PDF / 1,021,365 Bytes
• 9 Pages / 595.276 x 790.866 pts Page_size
• 6 Downloads / 253 Views

DOWNLOAD

REPORT

---

(0123456789().,-volV)(0123456789(). ,- volV)

An improved Cuckoo search localization algorithm for UWB sensor networks Xiaofeng Qin1 • Bin Xia1

•

Tian Ding1 • Lei Zhao1

Ó Springer Science+Business Media, LLC, part of Springer Nature 2020

Abstract Conventional Cuckoo search (CS) localization method can obtain good positioning results that are highly accurate and robust. However, its positioning performance is constrained by the limited distance information available between the unknown node and the anchor node. In order to further enhance the positioning accuracy of the CS method, an improved CS localization algorithm is proposed that can take advantage of all of the distance information available. In the positioning process, an objective function which contains the distance information among the unknown nodes is given. Firstly, we use this distance information between the anchor node and the unknown node to determine an initial position through the conventional CS method. Then, based on the initial position, all of the distance information available is used to compute a more precise position. The simulation results demonstrate that the proposed algorithm can enhance the positioning accuracy in comparison with the conventional CS algorithm. Keywords Cuckoo search  Localization algorithm  Non line of sight

1 Introduction An ultra-wide band(UWB) sensor network [1–3] is a network of cheap and simple processing UWB senor nodes that are spatially distributed in an area of interest in order to cooperatively monitor physical or environmental phenomena. Localization in UWB sensor networks is an important area that attracts significant research interest. It is required in many sensor network applications, such as target localization [4], medicine [5, 6], robotics [7], and so on. Therefore, many range-based localization algorithms are proposed in the literature. Range-based localization requires measuring the distance between nodes, and using this distance to determine the position of the unknown node. Distance measurement methods generally include the received signal strength indicator [8–10], angle of arrival [11] , and time of arrival (TOA) [12]. Because the pulse signal in the UWB sensor network has a high time and multipath resolution, the

distance is measured by the TOA value of the signal. In reliance on the TOA measurement data, there are two major localization methods, i.e. linear and nonlinear positioning. The linear methods, such as Taylor algorithms [13, 14], transform a nonlinear positioning equation into a linear one. The nonlinear methods, such as Cuckoo search (CS) algorithm [15, 16] and gradient descent (GD) algorithm [17, 18], regard the positioning problem as an optimization one. Among all the above methods, the CS algorithm is commonly employed; it has the advantages of being both accurate and robust. In the CS method, the objective function only utilizes the distance information between the anchor node and the unknown node. It does not utilize the distance information available among the unknown n