Procedure for the removal of the DC component inherent in any radio frequency chain
View Patent ↗A procedure for compensating the DC component inherent in any radio frequency chain making it possible from a single measurement, generally located in the digital stage, to determine a set of multiple compensation values that must be applied to multiple compensation points of the analogue chain. The compensation values are calculated by an iterative process converging towards a cancellation of the DC component and avoid saturating amplification components and components of the analogue digital converter.
1. Procedure for the removal of the DC component inherent in any radio frequency chain within a radio frequency reception device composed of at least one analogue stage, said analogue stage comprising a multiplicity of analogue components likely to insert a DC component in the signal passing through the stage, said stage comprising a multiplicity of compensation points, a DC compensation signal obtained from the analogue-digital conversion of a compensation value being added to each compensation point, all these compensation points forming a compensation vector, the addition of all the signals to the compensation points aiming to cancel the said DC component inherent to the radio frequency chain being measured at the end of the said analogue stage, which is characterised in that it comprises the following steps:
a) a computation step ( 3 . 1 ) of an approximate gain vector with a number of components corresponding to the number of compensation points and of which each component is the product of gain and attenuation values of analogue components located downstream of the corresponding compensation point in the analogue stage;
b) a computation step ( 3 . 2 ) of the corresponding unit gain vector by dividing the gain vector by its Euclidean norm;
c) a computation step ( 3 . 3 ) of a first term, where said value measured at the end of the said analogue stage to be compensated is divided by the gain vector norm and making the product of this result with the unit gain vector, wherein the first term is multiplied by a weighting value;
d) a computation step ( 3 . 4 ) of a new compensation vector by adding to the first term a second term, constituted by the current compensation vector or of the scalar product of this current compensation vector with the unit gain vector multiplied by this unit gain vector, wherein the second term is multiplied by a weighting value;
e) a lowering step ( 3 . 5 ) of a component of the new compensation vector to a maximum admissible value of analogue-digital converters used to generate the compensation signals when a value of the obtained component of the new compensation vector is higher than the admissible maximum value.
2. Procedure according to claim 1 , characterised in that phases c), d) and e) are repeated iteratively.
3. Radio frequency reception device comprising:
at least one analogue stage, comprising a multiplicity of analogue components which can insert a DC component in the signal passing through the stage, characterised in that the said analogue stage comprises a multiplicity of compensation points, a DC compensation signal obtained from the analogue-digital conversion of a compensation value being added to each compensation point, all these compensation points forming a compensation vector, the addition of all the signals to the compensation points aiming to cancel the said DC component measured at the end of the said analogue stage, and
a module for producing the vector of compensation values, the module causing the device to perform:
a) calculating a gain vector with a number of components corresponding to the number of compensation points and of which each component is the product of gain and attenuation values of analogue components located downstream of the corresponding compensation point in the analogue stage;
b) calculating the corresponding unit gain vector by dividing the gain vector by its Euclidean norm;
c) calculating a first term where the value to be compensated is divided by the gain vector norm and making the product of this result with the unit gain vector, wherein the first term is multiplied by a weighting value;
d) calculating a new compensation vector by adding to the first term a second term, composed of the standard constituted by the current compensation vector or of the scalar product of this current compensation vector with the unit gain vector multiplied by this unit gain vector, wherein the second term is multiplied by a weighting value;
e) lowering the value of a component of the new compensation vector obtained to the maximum admissible value of analogue-digital converters used to generate the compensation signals when a value of the obtained component of the new compensation vector is higher than the admissible maximum value.