Technique for soft bit scaling
AssigneeTelefonaktiebolaget L M Ericsson (publ)
InventorUdo Wachsmann
Life of this patent
Estimating the term…
Abstract
A technique for determining in a fading channel environment a scaling factor for soft bit quantization is proposed. In a method realisation, the technique includes repeatedly determining an SNR value for the fading channel, calculating a weighted average based on a plurality of the SNR values, wherein weighting is performed such that contributions of extremal SNR values are attenuated, and determining the scaling factor for soft bit quantization dependent on the weighted average.
Claims15
A method of determining in a fading channel environment a scaling factor for soft bit quantization, the method implemented by a decoder stage comprising steps of:
- repeatedly determining a Signal-to-Noise Ratio (SNR) value for the fading channel;
- calculating a weighted average based on a plurality of the SNR values, wherein the weighting of the SNR values is performed such that contributions of extremal SNR values are attenuated;
- determining the scaling factor for soft bit quantization dependent on the weighted average;
- and
- wherein the weighting is performed by a transformation configured to transform each SNR value into a code rate domain to obtain an associated code rate value, and
- wherein the method further comprises averaging the resulting code rate values to obtain the weighted average.
A non-transitory computer-readable medium which has an executable computer program product stored thereon, the computer program product comprising program code portions to be processed by a computing device for determining in a fading channel environment a scaling factor for soft bit quantization by:
- repeatedly determining a Signal-to-Noise Ratio (SNR) value for the fading channel;
- calculating a weighted average based on a plurality of the SNR values, wherein the weighting of the SNR values is performed such that contributions of extremal SNR values are attenuated;
- determining the scaling factor for soft bit quantization dependent on the weighted average;
- and
- wherein the weighting is performed by a transformation configured to transform each SNR value into a code rate domain to obtain an associated code rate value, and wherein the method further comprises averaging the resulting code rate values to obtain the weighted average.
A decoder stage configured to determine in a fading channel environment a scaling factor for soft bit quantization, comprising:
- a first component configured to repeatedly determine a Signal-to-Noise Ratio (SNR) value for the fading channel;
- a processor configured to calculate a weighted average based on a plurality of the SNR values, wherein the weighting of the SNR values is performed such that contributions of extremal SNR values are attenuated;
- a second component configured to determine the scaling factor for soft bit quantization dependent on the weighted average;
- and
- wherein the processor is configured to perform the weighting by transforming each SNR value into a code rate domain to obtain an associated code rate value, and to average the resulting code rate values to obtain the weighted average.
A decoder stage configured to determine in a fading channel environment a scaling factor for soft bit quantization, comprising
- a first component configured to repeatedly determine an SNR value for the fading channel;
- a first processing unit configured to transform each SNR value into a code rate domain to obtain an associated code rate value;
- a second processing unit configured to calculate an average based on a plurality of the transformed SNR values, and perform a weighting such that in the calculated average contributions of extremal SNR values are attenuated, and
- a second component configured to determine the scaling factor for soft bit quantization dependent on the average.
Dependencies are inferred from the claim text.