초록 ▼

An adaptive signal processor including a network that adjusts itself on the basis of power maximization to resolve a raw composite input signal structure at least approximately into its separate statistically independent self-correlated components, or eigenvectors, and automatic gain control circuit

An adaptive signal processor including a network that adjusts itself on the basis of power maximization to resolve a raw composite input signal structure at least approximately into its separate statistically independent self-correlated components, or eigenvectors, and automatic gain control circuits for equalizing the power levels, or eigenvalues, of these components. The equalized component signals may be used as inputs to conventional adaptive systems, providing improved settling speed and dynamic range, or may in some cases be used directly, without further adaptive processing.

대표청구항 ▼

The method of processing a composite of raw input signals having common components in separately variable proportions to produce an output wherein one of said components predominates, and at least one other output wherein said one component does not predominate, comprising the steps of: a. diverting

The method of processing a composite of raw input signals having common components in separately variable proportions to produce an output wherein one of said components predominates, and at least one other output wherein said one component does not predominate, comprising the steps of: a. diverting portions of all said raw inputs and passing the remainders thereof, b. combining said diverted portions to produce a first output, c. adjusting the phases of said raw inputs to maximize said first output, and d. combining said remainders of said raw inputs to produce at least one further output.