Tone detection for signals sent through a vocoder

• US 9,208,797 B2
• Filed: 04/18/2008
• Issued: 12/08/2015
• Est. Priority Date: 04/18/2008
• Status: Active Grant

• Current Assignee: General Motors LLC (General Motors Company)
• Original Assignee: General Motors LLC (General Motors Company)
• Inventors: Madhavan, Sethu K., Yin, Jijun, Jiang, Qin, Van Buer, Darrel James
• Primary Examiner(s): He, Jialong
• Application Number:
  US12/106,261
• Publication Number:
  US 20090265173A1
• Time in Patent Office: 2,790 Days
• Field of Search: 379/31, 379/283, 379/386, 704500-503
• US Class Current: 1/1
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10
• CPC Class Codes: H04Q 1/46 comprising means for distinguishing between a signalling current of predetermined frequency and a complex current containing that frequency, e.g. speech current

Abstract

A tone detector and associated method for use with EVRC-B and GSM vocoders to enable reliable detection of system connect tones over a wireless communication system. The tone detection method examines a number of sequential data frames of the signal received from the vocoder and determines that the tone is present if the spectral energy at frequencies around the tone is much higher than that at neighboring frequencies and if the calculated center frequency of the data frames is at or near the frequency of the tone.

Claims

1. A method for detecting a tone in a signal received from a vocoder, comprising carrying out the following steps using a processor-based tone detector:

performing a spectral energy analysis on the signal;

determining a center frequency for the signal; and

determining that tone is present if a result of the spectral energy analysis meets at least one selected criterion and if the center frequency is within a selected range of frequencies that includes a primary frequency of the tone.

2. A method for detecting a tone in a signal received from a vocoder, comprising carrying out the following steps using a processor-based tone detector:

(a) carrying out the following steps (a1) through (a3) for each of plurality of different portions of the signal;

(a1) determining an energy characteristic of a portion of the signal at each of a first plurality of selected frequencies that together comprise a first group of frequencies;

(a2) determining an energy characteristic of the portion of the signal at each of a second plurality of selected frequencies that together comprise a second group of frequencies;

(a3) determining a relationship between a combination of the energy characteristics for the first group of frequencies and a combination of the energy characteristics for the second group of frequencies; and

(b) determining a center frequency for the first group of frequencies; and

(c) determining that the tone is present in the signal if the relationship meets at least one selected criterion and if the center frequency is within a selected range of frequencies that includes a primary frequency of the tone.

3. The method of claim 2, wherein the first group of frequencies includes the primary frequency of the tone.

4. The method of claim 2, wherein the primary frequency and the frequencies contained in the first group of frequencies are all within a range of 2,000 to 3,000 Hz and wherein the second group of frequencies includes frequencies under 2,000 Hz and frequencies over 3,000 Hz.

5. The method of claim 4, wherein the primary frequency is 2225 Hz and constitutes the only frequency of the tone.

6. The method of claim 2, wherein steps (a1) and (a2) further comprise determining the energy contained in the portion of the signal at each of the selected frequencies.

7. The method of claim 6, wherein the energy is determined using a Goertzel algorithm.

8. The method of claim 6, wherein step (a3) further comprises the steps of combining the determined energies of the first plurality of frequencies into a first energy value, combining the determined energies of the second plurality of frequencies into a second energy value, and determining an energy ratio using the first and second energy values.

9. The method of claim 8, further comprising the step of setting the energy ratio to zero if the first energy value is below a selected value.

10. The method of claim 8, further comprising the step of smoothing the energy ratio.

11. The method of claim 8, wherein step (c) further comprises determining that the tone is present if the energy ratio is above a pre-established threshold and if the center frequency is within the selected range of frequencies.

12. The method of claim 11, wherein step (b) further comprises determining the center frequency only if the energy ratio is above the pre-established threshold.

13. The method of claim 2, wherein step (b) further comprises determining a single center frequency associated with the plurality of different portions of the signal.

14. The method of claim 13, wherein step (b) further comprises determining the center frequency from an estimated center frequency of each of the different portions of the signal.

15. The method of claim 2, wherein step (b) further comprises determining the center frequency using the energy characteristics determined for the first group of frequencies.

16. The method of claim 2, wherein the different portions of the signal comprise sequential data frames.

17. The method of claim 2, further comprising, prior to step (a), the step of receiving the signal via a vocoder using an Enhanced Variable Rate Codec-B (EVRC-B) or Global System for Mobile Communications (GSM) codec.

18. A method for detecting a tone in a signal received from an Enhanced Variable Rate Codec-B (EVRC-B) or Global System for Mobile Communications (GSM) vocoder, comprising carrying out the following steps using a processor-based tone detector:

(a) providing a first frequency set containing a plurality of discrete frequencies located near the tone;

(b) providing a second frequency set containing a plurality of discrete frequencies, some of which are located below all the frequencies in the first frequency set and others of which are located above all the frequencies in the first frequency set;

(c) carrying out the following steps (c1) through (c3) for each of plurality of sequential data frames of the signal;

(c1) determining an energy value for the first frequency set;

(c2) determining an energy value for the second frequency set;

(c3) determining an energy ratio using the energy values for the first and second frequency sets;

(d) determining a center frequency for the sequential data frames; and

(e) determining that the tone is present in the signal if the energy ratio meets at least one selected criterion and if the center frequency is within a selected range of frequencies that includes a primary frequency of the tone.

19. The method of claim 18, wherein the primary frequency and the frequencies contained in the first frequency set are all within a range of 2,000 to 3,000 Hz and wherein the second frequency set includes frequencies under 2,000 Hz and frequencies over 3,000 Hz.

20. The method of claim 19, wherein:

step (c1) further comprises determining a first energy value by determining the energy contained in the data frame at each of the discrete frequencies in the first frequency set using a Goertzel algorithm and then summing those determined energies together;

step (c2) further comprises determining a second energy value by determining the energy contained in the data frame at each of the discrete frequencies in the second frequency set using a Goertzel algorithm and then summing those determined energies together;

step (c3) further comprises determining the ratio of the first energy value to the second energy value;

step (d) further comprises determining a single center frequency for the plurality of sequential data frames; and

step (e) further comprises determining that the tone is present in the signal if the energy ratio is above a pre-established threshold and if the center frequency is within a selected range of frequencies.