A single origin and single destination highway traffic network with multiple routes and sections is considered. For dynamic routing of traffic, accurate measurement of traffic densities is needed, which requires the placement of many sensors, one for each section, which can be a costly solution. In this work, a simpler solution is presented based on a single sensor placed at the destination node and the use of a state dependent Riccati equation filter for estimation of traffic densities in each section of every route. Two sets of simulations are provided for illustration. In addition, a comparative robustness study with extended Kalman filter involving performance under different noise level is included.