To mitigate the negative impact of varying satellite signal-to-noise ratio (SNR) on the vector lock loop (VLL) in a Global Navigation Satellite System (GNSS) receiver, an innovation based adaptive Kalman filter is employed to real-time adjust VLL measurement noise covariance matrix. Intermediate frequency data from a Global Positioning System (GPS) signal generator are post-processed using a software GPS receiver to verify the proposed method. The experimental result shows that VLL with the proposed adaptive Kalman filter outperforms under lower SNR conditions or with even insufficient amount of visible satellites comparing with the conventional scalar lock loop (SLL). Both carrier frequency tracking results and navigation solutions from the proposed method are more accurate than those from either fixed-parameter VLL or conventional SLL.