From: Locally optimal detector design in impulsive noise with unknown distribution
Step 1: Given P and \(\breve {g}[x_{i}]\), calculate \(\widehat {\boldsymbol {A}}\) by formulas (16)-(19). |
Step 2: Solve the function \(\sum \limits _{p=0}^{P} (p+1)\widehat {A}_{p} x^{p} =0, \,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,(23) \) |
and evaluate x n e and x p o by the largest negative root and the smallest positive root respectively. |
Step 3: Compute the two parameters |
\( B_{ne} = \sum \limits _{p=0}^{P} \widehat {A}_{p} x_{ne}^{p+1}, B_{po} = \sum \limits _{p=0}^{P} \widehat {A}_{p} x_{po}^{p+1} \,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,(24) \) |
Step 4: Obtain \(\widehat {g}(x)\) by (21). |