GB2346040A - Method of slope detection in bursty RF signals - Google Patents

Abstract

A method for slope detection in RF signals in the presence of noise, for example on close GSM power bursts, looks at a cluster of power points to make a slope determination, as opposed to a simple change between points A and B, which is prone to noise errors. The slope detection is moved across the signal in order to find the rising and falling edges of the burst so that once these edges are determined, the burst is identified and extracted. Filtering or averaging is not needed and detection takes place only when the power is above a specified threshold.

Description

METHOD OF SLOPE DETECTION IN RF SIGNALS The present invention relates to slope detection in RF signals, for example for the detection of slope and identification of power bursts such as, in one preferred embodiment, close GSM power bursts.

GSM, or Global System for Mobile Communications, is the worldwide standard for digital mobile telephones. One problem with GSM applications has been an efficient method to determine the slope of a signal with potential noise, or an RF envelope, or any other such signal. Traditional methods are prone to noise errors. It would be desirable to provide an improved method of detection of RF signals in the presence of noise.

According to an aspect of the present invention, there is provided a method of slope detection in an RF signal comprising the steps of : determining first and second power cluster points A and B to be examined on an RF signal burst of the RF signal; calculating the power under first and second power cluster areas A and B on the RF signal power burst under examination; and calculating the slope of the respective first and second power cluster areas A and B.

The preferred embodiment can provide a method for slope detection in RF signals in the presence of noise, more particularly for detecting slope on close RF power bursts such as GSM power bursts. Once the positive and negative slopes have been determined, the burst can be identified and extracted. One preferred embodiment is applicable to GSM applications. However, other embodiments could be applicable to the general class of Time Division Multiple Access (TDMA) signals of which GSM, PDC (Pacific Digital Cellular), NADC (North American Digital Cellular) and the like are a part.

In one preferred embodiment, the method differs from other edge detectors by looking at a cluster of power points to make a slope determination, as opposed to a simple A between points A and B, which is prone to noise errors. The computation of the cluster of points is an integration. This integration acts like a selective moving average filter. Among the novel contributions are: 1) filtering or averaging is not needed on the entire signal prior to slope detection, and only occurs in areas above a specified threshold (the selective integration provides computational efficiency) ; 2) the slope detection is moved across the signal in order to find the rising and falling edges of the burst so that once these edges are determined, the burst is identified and extracted. A specified change in power must occur, further rejecting false detection of slope in the presence of noise.

The preferred method determines the slope of a signal at a given point in the presence of noise. Rather than take a simple delta between the points, the novel method computes the slope by comparing the power around the desired point. The width of the comparison is a key factor of rejecting noise. Too much width could reject the signal. Another key factor is the amount of change (i. e., delta) in power.

An embodiment of the present invention is described below, by way of example only, with reference to the accompanying drawings, in which: Figure 1 shows a timing diagram of a GSM power burst; Figure 2 shows a timing diagram illustrating an embodiment of detection method; Figures 3A and 3B show timing diagrams illustrating particular examples of the embodiment of Figure 2; Figure 4 shows a timing diagram illustrating an embodiment of burst extraction.

Figure 1 shows a timing diagram of a GSM power burst in which the y-axis is a measure of GSM RF power and the x-axis is measured over time. It can be seen that the GSM signal is a series of recurring bursts.

Figure 2 shows a timing diagram illustrating an embodiment of detection method which includes the following series of steps : In Figure 2, the method detecting the slope on close GSM power bursts such as shown in Figure 1.

The first step in Figure 2 is based on level threshold, pick a point to examine.

The second step is to calculate power under areas A and B shown in Figure 2. The width of power region is a parameter At (i. e., width), that is set by the application software.

The next step is to calculate the slope as follows: (a) if | A-B [ < e, assume the slope is flat (i. e. 0-see sample 3 in Figure 2), where e is some minimum epsilon greater than 0.0, or change in power, and is specified by the application software.

(b) if (A > B), the slope is negative-see sample 1 in Figure 2 where A = power on points before the point to test and B = power on points after the point to test (given a point to test, A is the region of power before the point, and B is the region of power after the point).

(c) if (A < B) slope is positive + see sample 2 in Figure 2.

(d) variances in At and e (the change in power) can affect the level of noise rejection. ç s method can be implemented using a common programming langage (e. g. C, C-H-, etc.) or programmable instructions for a DSP IC using arrays of data representing time vs amplitude.

This method differs from other edge detectors in that it looks at a cluster of power points to make the slope determination, as opposed to a simple A between points A and B which is prone to noise errors. Among the contributions are: 1) filtering or averaging is not needed on the entire signal prior to slope detection, and only occurs in areas above a specified threshold (the selective integration provides computational efficiency); 2) the change in power (delta) exceeds some value (epsilon) as provide by the application software; 3) slope detection is moved across the signal in order to find the rising and falling edges of the burst so that once these edges are determined, the burst is identified and extracted..

Figures 3A and 3B show timing diagrams illustrating traditional methods compared with the method described above. Figure 3A shows a potential false negative slope using traditional methods and Figure 3B shows the method described above.

The method shown in Figure 3b determines Power A to be < Power B at slope point which indicates a generally rising edge. Traditional methods show falling edge, as shown in Figure 3A.

The described method deternires theslope of a signal at a given point in the presence of noise. Rather than take a simple delta between the points, the method computes the slope by comparing the power around the desired point. The width of the comparison is a key factor of rejecting noise. Too much width could reject the signal. The amount of change in power is another key factor.

-Figure 4 shows a timing diagram illustrating an embodiment of burst extraction.

In Figure 4, the RF signal prior to time Tl is below a threshold so that the slope is not computed prior to time T1. At time T1, the burst is above the threshold and the delta power slope detection method described above is applied on the RF signal to find the positive or rising edge. Similarly, at time T2, the delta power slope detection method is applied on the RF signal to find the negative or falling edge. A scan of the slope detector across the signal between times T1 and T2 provides for identifying the burst so that the burst can be extracted. After time T2, the RF signal falls below the threshold, so that the slope is not computed after time T2.

The foregoing descriptions of specific embodiments of the present invention have been presented for purposes of illustration and description. They are not intended to be exhaustive or to limit the invention to the precise forms disclosed, and it should be understood that many modifications and variations are possible in light of the above teaching. For example, one preferred embodiment is applicable to GSM applications. However, other embodiments could be applicable to the general class of Time Division Multiple Access (TDMA) signals of which GSM, PDC (Pacific Digital Cellular), NADC and the like are a part. The embodiments were chosen and described in order to best explain the principles of the invention and its practical application, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated.

The disclosures in United States patent application numbers 60/106,665 and 09/196,516, from which this application claims priority, and in the abstract accompanying this application are incorporated herein by reference.

Claims (7)

1. CLAIMS 1. A method of slope detection in an RF signal comprising the steps of : determining first and second power cluster points A and B to be examined on an RF signal burst of the RF signal ; calculating the power under first and second power cluster areas A and B on the RF signal power burst under examination ; and calculating the slope of the respective first and second power cluster areas A and B.
2. 2. The method as in Claim 1 wherein the RF signal is one of a class of Time Division Multiple Access Signals, of which GSM, PDC, NADC and the like are a part.
3. 3. The method as in Claim 2 including the step of determining that the slope is flat if the difference between first and second calculated power cluster areas is less than epsilon where epsilon is a specified change in power.
4. 4. The method as in Claim 3 including the step of determining that the calculated slope is negative if the power cluster area A is greater than B.
5. 5. The method as in Claim 3 or 4including the step of determining that the calculated slope is positive if the power cluster A is less than B.
6. 6. The method as in Claim 5 including the step of detecting the rising and falling edges of the RF signal burst to provide for identification of the signal burst.
7. 7. The method as in Claim 6 including the step of extracting the identified signal burst.