Class IntegerBootstrapping

  • All Implemented Interfaces:
    IntegerLeastSquareSolver

    public class IntegerBootstrapping
    extends LambdaMethod
    Bootstrapping engine for ILS problem solving. This method is base on the following paper: Success probability of integer GPs ambiguity rounding and bootstrapping by P. J. G. Teunissen 1998 and Influence of ambiguity precision on the success rate of GNSS integer ambiguity bootstrapping by P. J. G. Teunissen 2006.

    This method is really faster for integer ambiguity resolution than LAMBDA or MLAMBDA method but its success rate is really smaller. The method extends LambdaMethod as it uses LDL' factorization and reduction methods from LAMBDA method. The method is really different from LAMBDA as the solution found is not a least-square solution. It is a solution which asses a probability of success of the solution found. The probability increase with the does with LDL' factorization and reduction methods.

    If one want to use this method for integer ambiguity resolution, one just need to construct IntegerBootstrapping only with a double which is the minimal probability of success one wants. Then from it, one can call the solveILS method.

    Since:
    10.2
    Author:
    David Soulard
    • Constructor Detail

      • IntegerBootstrapping

        public IntegerBootstrapping​(double prob)
        Constructor for the bootstrapping ambiguity estimator.
        Parameters:
        prob - minimum probability acceptance for the bootstrap
    • Method Detail

      • discreteSearch

        protected void discreteSearch()
        Compute the solution by the bootstrap method from equation (13) in P.J.G. Teunissen November 2006. The solution is a solution in the distorted space from LdL' and Z transformation.
        Overrides:
        discreteSearch in class LambdaMethod