Performance of visible light communication (VLC) depends highly on channel conditions, which are sensitive to changes in user position. This paper presents estimation schemes of VLC parameters using a Kalman filter (KF), based on angular parameters of user position. The angular dynamic model is established so that the estimation process is directly in accordance with the Lambertian model of VLC channel. The use of angular model also gave way to use two parameters to describe a three-dimensional position. Estimations based on angular position are formulated, that is the KF estimation of position parameters, and the extended Kalman filter (EKF) where channel gain is estimated and also serves as a state parameter. The performance is observed in simulation and compared to reference models of Cartesian based estimation. The proposed angular model EKF with the channel gain as the state parameter showed comparably higher error than the Cartesian model EKF of 3:2 in comparison but required remarkably less processing time of 1:5 to the referred model.