摘要
A maximum entropy method of power spectrum estimation, originally suggested by Burg(l967), has been proven to be a powerful technique for spectrum analysis, primarily due to itshigh resolution properties. A two-dimensional form of maximum entropy spectrum estimationproblem had been presented by Burg in an unpublished report. A general solution method for themaximum entropy spectrum estimation with multi-dimensional was described by MeClellan andLang (1983). The two-dimensional maximum entropy problem, however, is highly non1inear, andso far no closed-form solution has been yet found. An algorithm researched by Lim and Malik(1981 ) to solve that problem would be an iterative "alternating projections" type of one. Here, itwould be solved that a fast method of maximum entropy spectrum estimation for multi-dimensionalsignal spectrum analysis by logging data. The result of a nonlinear programming for the maximumentropy with multi-dimensional signal problem and a dualistic nonlinearity one would have beenfound. The maximum entropy spectrum estimation with multi-dimensional might then be obtainedas solution to a dual optimization problem. Therefore, it would be only considered a "correlation-matching" problem in multi-dimentaonal signal for the maximum entropy spectrum estimation bylogging data.
A maximum entropy method of power spectrum estimation, originally suggested by Burg(l967), has been proven to be a powerful technique for spectrum analysis, primarily due to itshigh resolution properties. A two-dimensional form of maximum entropy spectrum estimationproblem had been presented by Burg in an unpublished report. A general solution method for themaximum entropy spectrum estimation with multi-dimensional was described by MeClellan andLang (1983). The two-dimensional maximum entropy problem, however, is highly non1inear, andso far no closed-form solution has been yet found. An algorithm researched by Lim and Malik(1981 ) to solve that problem would be an iterative "alternating projections" type of one. Here, itwould be solved that a fast method of maximum entropy spectrum estimation for multi-dimensionalsignal spectrum analysis by logging data. The result of a nonlinear programming for the maximumentropy with multi-dimensional signal problem and a dualistic nonlinearity one would have beenfound. The maximum entropy spectrum estimation with multi-dimensional might then be obtainedas solution to a dual optimization problem. Therefore, it would be only considered a "correlation-matching" problem in multi-dimentaonal signal for the maximum entropy spectrum estimation bylogging data.
