The evaluation and performance testing of a least squares emitter locating algorithm

This thesis is an evaluation of a least squares algorithm developed by North American Rockwell (NAR). The NAR algorithm was previously implemented for locating radar emitters in airborne data collection systems, and was considered for implementation by the Tactical Electronic Reconnaissance Processing Evaluation Segment (TERPES) of Marine Air Ground Intelligence System (MAGIS). Because the NAR algorithm was fast and supposedly proven, it was a natural candidate for TERPES. However, subsequent testing indicated the presence of a serious systematic error. This paper is concerned with evaluating the NAR algorithm to (1) formally substantiate the existence of a systematic error, (2) determine the source of the error, and (3) correct the error or develop an alternate approach (See more in text.)