Method for creating a representation of a calculation result linearly dependent upon a square of a value

US8037114B2 · kind B2 · utility

Assignee

• Fraunhofer-Gesellschaft zur förderung der angewandten Forschung e.V. · DE

Inventors

• Marc Gayer · Erlangen, DE
• Manfred Lutzky · Nuremberg, DE
• Markus Lohwasser · Hersbruck, DE
• Sascha Disch · Bad Mergentheim, DE
• Johannes Hilpert · Nuremberg, DE
• Stefan Geyersberger · Erlangen, DE
• Bernhard Grill · Lautenbächle, DE

Key dates

• Filing date: Jun 13, 2007
• Grant date: Oct 11, 2011
• Expiry date: Aug 9, 2030

Classification

• Technology area (CPC G): Physics
• CPC primary: G10L19/02
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

In the transition into the logarithmic range, not the entire bit width of the result linearly dependent upon the square of the value must be considered. Rather, it is possible to scale the result of a value with x bits such that a representation with less than x bits of the result is sufficient to receive the logarithmic representation based thereon. The effect of the scaling factor on the resulting logarithmic representation may be compensated for by adding or subtracting a correction value received by the logarithm function applied to the scaling factor to or from the scaled logarithmic representation without any loss of dynamics. This way, a method and an apparatus for creating a representation of a result linearly dependent upon a square of a value are provided so that the calculation is simple and/or possible with little hardware expenditure.