| Summary: | In <xref ref-type="bibr" rid="ref1">[1]</xref>, the calculation of <xref ref-type="disp-formula" rid="deqn8">(8)</xref> was erroneous. To be specific, as shown in <xref ref-type="disp-formula" rid="deqn8">(8)</xref> of <xref ref-type="bibr" rid="ref1">[1]</xref>, the average transmit power of a specially-designed signal <inline-formula> <tex-math notation="LaTeX">${x_{m} = - {{\sqrt {P_{S}} {h_{Sm}}h_{Am}^{*}{x_{S}}}}/{{{{\left |{ {{h_{Am}}} }\right |}^{2}}}}}$ </tex-math></inline-formula> was obtained as <disp-formula> <tex-math notation="LaTeX">\begin{equation*}{P_{m} = \frac {{\sigma _{Sm}^{2}}}{{\sigma _{Am}^{2}}}}P_{S},\end{equation*} </tex-math></disp-formula> where <inline-formula> <tex-math notation="LaTeX">$\sigma _{Sm}^{2} = E(|{h_{Sm}}{|^{2}})$ </tex-math></inline-formula> and <inline-formula> <tex-math notation="LaTeX">$\sigma _{Am}^{2} = E(|{h_{Am}}{|^{2}})$ </tex-math></inline-formula>. This calculation was obtained by mistaking an expectation of <inline-formula> <tex-math notation="LaTeX">$1/|h_{Am}|^{2}$ </tex-math></inline-formula> for <inline-formula> <tex-math notation="LaTeX">$E(1/|h_{Am}|^{2})=1/E(|h_{Am}|^{2})= 1/\sigma _{Am}^{2}$ </tex-math></inline-formula>, which is erroneous to be corrected. Actually, considering that <inline-formula> <tex-math notation="LaTeX">$|h_{Am}|^{2}$ </tex-math></inline-formula> is an exponentially distributed random variable with a mean of <inline-formula> <tex-math notation="LaTeX">$\sigma ^{2}_{Am}$ </tex-math></inline-formula>, the expectation of <inline-formula> <tex-math notation="LaTeX">$1/|h_{Am}|^{2}$ </tex-math></inline-formula> tends toward infinity, rather than <inline-formula> <tex-math notation="LaTeX">$1/\sigma _{Am}^{2}$ </tex-math></inline-formula>.
|
|---|