Spectral density for third order cumulants under strong mixing conditions (Q1587360)

Let \(X=\{X_v\), \(v \in \mathbb{Z}^d\}\) be a strictly stationary real-valued random field. Three types of conditions are used: (A) means that \(X_v\) is a sum of certain indicator functions, (B) describes the decreasing rate of \(\alpha\)-mixing coefficient and (C) involves an estimate of some other mixing coefficient (using conditional distributions). Conditions (A) and (C) are discussed and examples of their validity are provided. If (A), (B) and (C) are satisfied, then there exists a continuous spectral density \(f(\lambda_1,\lambda_2)\) for the third order cumulant \(\text{cum}(X_a,X_b,X_c)\). It is mentioned that higher-order cumulants properties are important in signal analysis.