Indoor visible light localization based on compressed sensing under matrix filling recovery

In the study of indoor visible light localization, a matrix-based fingerprint library construction method is used to recover the data in the offline phase. While in the online matching and positioning phase, the Stagewise Orthogonal Matching Pursuit algorithm in compressed sensing theory is improved and introduced into indoor visible light localization. After simulation, the average error is 0.027 m. Then we verify the advancement of this algorithm by comparing with other algorithms applied to visible light localization. The experimental verification shows that the error is only 0.029 m.