Gauss-Markov-based mobile anchor localization (GM-MAL) algorithm based on local linear embedding optimization in internet of sensor networks

In the wireless sensor networks localization algorithm, the position of anchor node decides the node localization precision. Gauss-Markov-based mobile anchor localization (GM-MAL) algorithm is proposed. GM-MAL algorithm proposes mobile path planning of self-adaptive anchor node based on Gaussian-Markove mobile mode, and then estimates the position of node using alternating minimization algorithm (AMA). Concretely speaking. At the path planning stage, the algorithm plans the path through speed adjustment strategy, perpendicular bisector strategy, virtual repulsion strategy and virtual gravity strategy. At the location stage, the algorithm transforms the non-convex problem to bi-convex form, and then utilizes AMA for solving and obtaining shorter anchor node movement path. The experiment data indicate that the introduction of virtual gravity strategy improves the path planning precision and covers more monitoring area. In addition, compared with linear algorithm, the localization precision of GM-MAL is enhanced.