Run-time coefficient generation for digital filter with slewing bandwidth
View Patent ↗A method for slewing a bandwidth characteristic of a digital filter within a range between a first bandwidth characteristic and a second bandwidth characteristic. The digital filter includes a plurality of cascaded filter sections and each of the sections has respective coefficients. The method establishes a polynomial having an order greater than or equal to two for modeling a changing value of one of the coefficients at the bandwidth characteristics within the range. An index value is determined for identifying a desired bandwidth characteristic within the range. The index value is input into the polynomial to obtain a corresponding coefficient value for the one coefficient. The coefficient value is then loaded into a respective filter section.
1. A method of slewing a bandwidth characteristic of a digital filter within a range between a first bandwidth characteristic and a second bandwidth characteristic, said digital filter including a plurality of cascaded filter sections, each of said plurality of cascaded filter sections having respective coefficients, said method comprising the steps of:
establishing a polynomial having an order greater than or equal to two for modeling a changing value of one of said coefficients at said first and second bandwidth characteristics within said range;
determining an index value for identifying a desired bandwidth characteristic within said range;
inputting said index value into said polynomial to obtain a corresponding coefficient value for said one of said coefficients; and
loading said corresponding coefficient value into a respective filter section;
wherein each of said plurality of cascaded filter sections is comprised of a second order filter having coefficients a 1 , a 2 , b 0 , b 1 , and b 2 , and wherein coefficient values for said a 1 and a 2 coefficients are modeled by said polynomial
a 1 =p 0,1 +( p 1,1 ●i )+( p 2,1 ●i 2 ), and
a 2 =p 0,2 +( p 1,2 ●i )+( p 2,2 ●i 2 ),
where i is said index value and p terms are predetermined constants.
2. The method of claim 1 wherein a plurality of polynomials is established for modeling respective coefficients in each respective filter section, wherein said index value is input into each polynomial to obtain respective coefficient values, and wherein said respective coefficient values are loaded into their respective filter sections.
3. The method of claim 1 wherein said first bandwidth characteristic corresponds to a narrow bandwidth filter and said second bandwidth characteristic corresponds to a wide bandwidth filter, and wherein said index value is adapted to slew between said narrow and wide bandwidth filters.
4. The method of claim 1 wherein said index value is determined according to the steps of:
initializing said index value corresponding to a predetermined bandwidth characteristic;
detecting whether said predetermined bandwidth characteristic is to be increased or decreased; and
incrementing said index value if said predetermined bandwidth characteristic is to be increased, otherwise decrementing said index value.
5. The method of claim 4 wherein said digital filter comprises a channel filter in a broadcast radio receiver for selecting a desired radio signal, and wherein said step of detecting whether said predetermined bandwidth characteristic is to be increased or decreased comprises detecting the presence of undesired radio signals at adjacent frequencies, whereby said predetermined bandwidth characteristic is decreased when said undesired radio signals are present.
6. The method of claim 4 wherein said digital filter comprises a channel filter in a broadcast radio receiver for selecting a desired radio signal, and wherein said step of detecting whether said predetermined bandwidth characteristic is to be increased or decreased comprises detecting a weak signal condition of said desired radio signal, whereby said predetermined bandwidth characteristic is decreased when said weak signal condition is present.
7. The method of claim 4 wherein said index value is slewed by a first predetermined step size.
8. The method of claim 7 wherein said first predetermined step size is substantially equal throughout said range.
9. The method of claim 7 wherein said first predetermined step size varies nonlinearly throughout said range.
10. The method of claim 9 wherein said first predetermined step size increases exponentially within said range.
11. The method of claim 4 wherein said index value is incremented by a first predetermined step size and is decremented by a second predetermined step size.
12. The method of claim 11 wherein said first and second predetermined step sizes are respective constants.
13. The method of claim 12 wherein said second predetermined constant is greater than said first predetermined constant so that said digital filter slews from a wide bandwidth to a narrow bandwidth more quickly than from said narrow bandwidth to said wide bandwidth.
14. The method of claim 11 wherein said first and second predetermined step sizes vary nonlinearly throughout said range.
15. The method of claim 1 further comprising the steps of:
deriving coefficient values for said b 0 , b 1 , and b 2 coefficients in response to said coefficient values for said a 1 and a 2 coefficients.
16. The method of claim 15 wherein said b 0 , b 1 , and b 2 coefficients are given values providing unity gain for the respective filter section.
17. A digital signal processing unit comprising:
a reconfigurable digital filter including a plurality of cascaded filter sections, each section adapted to receive respective filter coefficients for providing selectable bandwidth characteristics within a range between a first bandwidth characteristic and a second bandwidth characteristic;
a memory storing at least one polynomial having an order greater than or equal to two for modeling a changing value of at least one of said filter coefficients at said first and second bandwidth characteristics within said range; and
a controller determining an index value for identifying a desired bandwidth characteristic within said range, inputting said index value into said at least one polynomial to obtain a corresponding coefficient value for said at least one of said filter coefficient, and loading said corresponding coefficient value into a respective filter sections;
wherein each of said plurality of cascaded filter sections is comprised of a second order filter, wherein said respective filter coefficients comprise the terms a 1 , a 2 , b 0 , b 1 , and b 2 , and wherein coefficient values for said a 1 and a 2 coefficients are modeled by said polynomial according to
a 1 =p 0,1 +( p 1,1 ●i )+( p 2,1 ●i 2 ), and
a 2 =p 0,2 +( p 1,2 ●i )+( p 2,2 ●i 2 ),
where i is said index value and p terms are predetermined constants.
18. The digital signal processing unit of claim 17 wherein said unit is a radio receiver for receiving a plurality of broadcast radio signals, wherein said unit further comprises an interference detector for detecting interference signals affecting a desired radio signal, and wherein said controller determines said index value in response to said interference detector.
19. The digital signal processing unit of claim 18 wherein said first bandwidth characteristic corresponds to a narrow bandwidth filter and said second bandwidth characteristic corresponds to a wide bandwidth filter, and wherein said index value is adapted to slew between said narrow and wide bandwidth filters.
20. The digital signal processing unit of claim 19 wherein said index value varies in predetermined constant step sizes.
21. The digital signal processing unit of claim 19 wherein said index value varies in predetermined nonlinear step sizes.