Traffic condition matrix estimation via weighted Spatio-Temporal Compressive Sensing for unevenly-distributed and unreliable GPS data

Traffic condition monitoring is important to nowadays metropolitan. A recent trend is to exploit the prevalence of Global Positioning System (GPS) embedded in public vehicles. The collected data forms a two dimensional traffic condition matrix (TCM), i.e., time slot and road segment. The problem is that the TCM directly obtained from the probed data is incomplete. Traffic estimation can complete the TCM by filling the missing entries. We find that in practice it is challenging to reliably estimate a TCM. First, The distribution of probed data is uneven among road segments. Second, most entries of probed data are unreliable since they are the average of only a few reports. Our approach is Weighted Spatio-Temporal Compressive Sensing. Demonstrated by extensive large scale computational experiments, the estimation error of our approach reduces to just half of the baseline approach.