A PTDOA-DRSS hybrid factor graph-based unknown radio wave geolocation
We propose a hybrid Pythagorean Time Difference of Arrival and Differential Received Signal Strength based Factor Graph (PTDOA-DRSS-FG) to estimate the location of unknown radio wave emitter in outdoor environments. The term of “unknown” indicates that the knowledge of neither time of departure (TOD) nor absolute transmit power of the radio wave emitter are required. The PTDOA-FG can eliminate the necessity of TOD knowledge of the target signal transmission and DRSS-FG can eliminate the necessity of the knowledge of target absolute transmit power. However, PTDOA-FG alone requires perfect time synchronism between sensors, which is difficult in practice. On the other hand, DRSS-FG alone requires the most suitable monitoring spots. In this paper, PTDOA-FG is used to provide the rough estimation of the target position to select the most suitable monitoring spots for DRSS-FG technique. It is shown that, in terms of root mean square error (RMSE) vs. iteration times, the achieved RMSE of the proposed technique is better than the PTDOA-FG alone and very close to the DRSS reference, i.e., the DRSS-FG technique with the idealistic monitoring spots. Performing the proposed technique in the framework of factor graph (FG) does not require excessive computational complexity due to the fact that using the Gaussian distribution model, it uses mean and variance only with the sum-product algorithm.