A Novel Source Localization Scheme In The Urban Environment

In this paper, a novel scheme of source localization in the urban environment is proposed. It firstly employs unitary ESPRIT (UESPRIT) algorithm to estimate the angles (DOA) and delays (TOA) of the arrival rays radiated from the source and secondly uses a back ray tracing technique to recover the signal propagation paths with the obtained DOA, TOA and city electronic maps, at last, obtains the source's position. As the estimated results by each path may not be identical, the final estimated position is determined by averaging all the estimated positions.In the back ray tracing processing some paths may be misleading because of the estimated accuracy of the UESPRIT. In order to discard those invalid paths a valid-range selection criterion is proposed. On the other hand, we introduce a path length weighting factor to reduce the estimated errors caused by the terrain data inaccuracy.This article focuses on studies of the application of our method in two-dimensional and three-dimensional environment, and gives a specific formula derivation. A series of numerical simulation results demonstrate the practicality and effectiveness of our method. At last this approach will also be extended to blind source localization, and some progress is made.