A bearings-only target location bibliography

There are many articles on the bearings-only target problem. The following articles and books are the ones that I used directly during the development of DFLib.

1. R.G. Stansfield (1947), "Statistical Theory of DF Fixing," J. IEE, Vol. 94, Part IIIA, No. 15, pp. 762-770; 1947. This is the first paper on the subject, and it is referenced by just about every other. It is rather opaque to read.
2. C.J. Ancker (1958), "Airborne Direction Finding---The Theory of Navigation Errors," IRE Trans. Aero. Nav. Elect., Vol. ANE-5, No. 4, pp. 199-210; December 1958. This paper includes a nice restatement of Stansfield's work in a more readable format with more explicit statement of the assumptions involved.
3. D. Koks, "Numerical Calculations for Passive Geolocations Scenarios," Austrailian Government Department of Defence Defence Science and Technology Organization, DSTO-RR-0319, Edinburgh, S. Aust. : DSTO, 2007. This paper can be accessed on-line at http://catalogue.nla.gov.au/Record/4367609. It is very readable and provides some good explanations of how to use Stansfield's formulas in practical application. It also provides several other more sophisticated statistical methods for processing DF bearing data.
4. K. Dogancay, "Bearings-only target localization using total least squares," Signal Processing 85(2005) pp. 1695-1710, doi:10.1016/j.sigpro.2005.03.007. To find the document on the web using the DOI in this entry, go to http://dx.doi.org/ and paste the DOI citation to the left (starts with "doi:" and ends with "007" not including the following period). It is not a free download unless you happen to work somewhere that gets free downloads from Elsevier as I do. This paper provides nice, clean descriptions of the Least Squares and Maximum Likelihood estimators using a formalism that is easy to implement. Just ignore his use of "steepest descents" to solve the ML problem — it is an appalling choice for real use. DFLib uses conjugate gradients instead.
5. S. Burnett et al., "Probabilistic Position-Fixing," Final report of the 1985-86 Claremont Graduate School Claremont McKenna College Mathematics Clinic. http://www.dtic.mil/cgi-bin/GetTRDoc?Location=U2&doc=GetTRDoc.pdf&AD=ADA190397. This paper highlights an error in Stansfield's original equations for the confidence region axes. It also provides a fairly readable exposition of the DF problem and its statistical treatment.
6. J.R. Shewchuk, "An Introduction to the Conjugate Gradient Method Without the Agonizing Pain," http://www.cs.cmu.edu/~jrs/jrspapers.html. A very readable explanation of the conjugate gradient method that is being used under the hood by DFLib in computation of the ML fix.
7. W.H. Press et al., Numerical Recipes in C(2nd edition): the art of scientific computing. ISBN:0-521-43108-5. A great starting place for developing codes that do numerical solution. The actual C code in the book can be a little sketchy due to its Fortrash history ("A determined programmer can write Fortran in any language"), so you have to modify them a lot, but a very good start.