Localization is crucial for the monitoring applications of cities, such as road monitoring,\nenvironment surveillance, vehicle tracking, etc. In urban road sensor networks, sensors are\noften sparely deployed due to the hardware cost. Under this sparse deployment, sensors cannot\ncommunicate with each other via ranging hardware or one-hop connectivity, rendering the existing\nlocalization solutions ineffective. To address this issue, this paper proposes a novel Traffic Lights\nSchedule-based localization algorithm (TLS), which is built on the fact that vehicles move through\nthe intersection with a known traffic light schedule. We can first obtain the law by binary vehicle\ndetection time stamps and describe the law as a matrix, called a detection matrix. At the same time,\nwe can also use the known traffic light information to construct the matrices, which can be formed as\na collection called a known matrix collection. The detection matrix is then matched in the known\nmatrix collection for identifying where sensors are located on urban roads. We evaluate our algorithm\nby extensive simulation. The results show that the localization accuracy of intersection sensors can\nreach more than 90%. In addition, we compare it with a state-of-the-art algorithm and prove that it\nhas a wider operational region.
Loading....