The　growing　number　of　vehicles　has　put　a　lot　of　pressure　on　the　transportation　system.　Intelligent　Transportation　System　(ITS)　faces　a　great　challenge　of　traffic　congestion.　Traffic　density　displays　the　congestion　of　current　traffic　which　reflects　explicitly　about　traffic　status.　With　the　development　of　communication　technology,　people　use　mobile　stations　(MSs)　at　any　time　and　cellular　signals　are　everywhere.　Different　from　traditional　traffic　information　estimation　methods　based　global　positioning　system　(GPS)　and　vehicle　detector　(VD),　this　paper　resorts　to　Cellular　Floating　Vehicle　Data　(CFVD)　to　estimate　the　traffic　density.　In　this　paper,　Hand　over　(HO)　and　call　arrival　(CA)　cellular　signals　are　essentials　to　estimate　traffic　flow　and　traffic　speed.　In　addition,　mixture　probability　density　distribution　generator　is　adopted　to　assist　estimating　the　probabilities　HO　and　CA　events.　Through　accurate　traffic　flow　and　traffic　speed　estimations,　precise　traffic　density　is　achieved.　In　the　simulation　experiments,　the　proposed　method　achieves　estimation　MAPEs　11.92%,　13.97%　and　16.47%　for　traffic　flow,　traffic　speed　and　traffic　density,　respectively.　©　2022　ACM.