Adaptive threshold setting for discontinuous transmission detection
A method of adaptively determining a threshold for determining whether an error associated with a frame is an erasure or the result of a discontinuous transmission (DTX) uses a sample set of recently received frames yet avoids undesirable influences on the threshold that otherwise might be caused by the number of erasures in the sample set. A disclosed example uses a histogram-based approach and a simulated sort through the histogram to determine an appropriate threshold value. In a disclosed example the lesser of a bad frame threshold guideline value and a good frame threshold guideline value provide the threshold used to distinguish between a DTX and an erasure.
1. A method of adaptively determining a threshold for distinguishing between a DTX and an erasure based upon at least one determined guideline value from a histogram of a sample set of recently received frames, comprising the steps of:
determining the histogram of the sample set by determining a good frame histogram of the good frames in the sample set and a bad frame histogram of the bad frames in the sample set;
arranging each histogram into a plurality of bins each having a bin size based on a range of the corresponding histogram and a number of the plurality of bins;
placing each one of the corresponding sample set frames into one of the bins corresponding to an energy value of each frame;
selecting a value of one of the bins as an indicator of a desired value of the good frame histogram and the bad frame histogram, respectively;
determining a good frame threshold value based on the desired value of the good frame histogram;
determining a bad frame threshold value based on the desired value of the bad frame histogram; and
selecting the lower of the determined good frame threshold value or the determined bad frame threshold value as the adaptively determined threshold for distinguishing between a DTX and an erasure.
2. The method of claim 1 , comprising predetermining a decision array and using a corresponding entry in the decision array for selecting the value of the one of the bins as the indicator of the desired value.
3. The method of claim 1 , comprising selecting the one of the bins based on a distribution of corresponding frames from the sample set in the corresponding histogram and a predetermined selection array that identifies the one of the bins useful for providing the indicator of the desired value.
4. The method of claim 3 , comprising using the selected value of the one bin associated with the good frames to determine the good frame threshold guideline;
using the selected value of the one bin associated with the bad frames to determine the bad frame threshold guideline; and
selecting the lesser of the good frame threshold or the bad frame threshold as the adaptively determined threshold.
5. The method of claim 1 , where the desired value corresponds to an approximation of a median value.
6. The method of claim 1 , comprising selecting a first period of time for defining the sample set of recently received frames to include a first number of frames and using the determined threshold for distinguishing a DTX frame an erasure for a second, lower number of incoming frames during an ensuing second, shorter period of time.
7. The method of claim 6 , wherein the first number of frames is fifty and the second number of frames is 10.
8. The method of claim 1 , comprising using the adaptively determined threshold to determine whether a subsequently received bad frame is an erasure.
9. The method of claim 8 , comprising implementing a power control adjustment responsive to determining that the received bad frame is an erasure.