The indoor fingerprinting localization technology has received more attention in recent years due to the increasing demand of the indoor location based services (LBSs). However, a high quality of the LBS requires a positioning solution with high accuracy and low computational complexity. The particle swarm optimization (PSO) technique, which emulates the social behavior of a flock of birds to search for the optimal solution of a special problem, can provide attractive performance in terms of accuracy, computational efficiency and convergence rate. In this paper, we adopt the PSO algorithm to estimate the location information. First, our system establishes a Bayesian-rule based objective function. It then applies PSO to identify the optimal solution. We also propose a hybrid access point (AP) selection method to improve the accuracy, and analyze the effects of the number and the initial positions of particles on the localization performance. In order to mitigate the estimation error, we use the Kalman Filter to update the initial estimated location via the PSO algorithm to track the trail of the mobile user. Our analysis indicates that our method can reduce the computational complexity and improve the real-time performance. Numerous experiments also demonstrate that our proposed localization and tracking system achieve higher localization accuracy than existing systems.

Wireless LANs, which consist of access points and wireless stations, have widely spread in recent years. Routing in wireless LANs suffers the problem that each wireless station selects an access point and a wired path to its destination station. It is desired to design an adaptive routing protocol for wireless LANs since throughputs of communications are dynamically affected by selections of other wireless stations and external environmental changes. In this paper, we propose a routing protocol for wireless LANs based on attractor selection. Attractor selection is a biologically inspired approach, and it has high adaptability to dynamic environmental changes. By applying attractor selection, each wireless station can adaptively select its access point and wired path with high throughput against environmental changes. In addition, we design the protocol with a new technique: combination of multiple attractor selections. The technique is useful because it enables us to divide a problem into several simpler problems. To the best of our knowledge, our protocol is the first one designed around a combination of multiple attractor selections. We show the effectiveness and adaptability of our protocol by simulations.