Methods and apparatus for job scheduling in a programmable mixed-radix DFT/IDFT processor

1. A method of processing discrete Fourier transform (“DFT”) or inverse DFT (“IDFT”) for data transmission in a communication network, the method comprising:

receiving a data stream containing a plurality of DFT jobs wherein a first portion of DFT jobs has a first DFT point size and a second portion of DFT jobs has a second DFT point size;

generating a first mega-job configured to bundle at least a portion of the DFT jobs with the first DFT point size in a predefined sequence;

storing the first mega-job at a ping half-side of a ping-pong memory bank before allowing the first mega-job to be fed to a vector pipeline for radix computation; and

determining a radix factorization for calculating DFT in response to the first DFT point size.

2. The method of claim 1 , further comprising generating a second mega-job configured to bundle at least a portion of the DFT jobs with the second DFT point size in a predefined second sequence.

3. The method of claim 1 , further comprising storing the second mega-job at another ping half-side of the ping-pong memory bank before allowing the second mega-job to be fed to the vector pipeline for radix computation.

4. The method of claim 1 , further comprising determining a second radix factorization for calculating DFT in response to the second DFT point size.

5. The method of claim 1 , further comprising retrieving the first mega-job from the first half-side of a ping-pong memory bank upon determination of the radix factorization based on the first DFT point size.

6. The method of claim 1 , further comprising loading the first mega-job into a vector pipeline for scaling and twiddle multiplication.

7. The method of claim 1 , further comprising generating one or more radix engine control signals to program a mixed radix engine for providing DFT calculations based on the radix factorization.

8. The method of claim 7 , wherein the generating one or more radix engine control signals further includes configuring a mixed radix engine to perform a selected one of radix3, radix4, radix5, and radix6 computations for each stage.

9. A method of processing discrete Fourier transform (“DFT”) for a high-speed data transmission in a communication network, the method comprising:

identifying a first mega-job stored in a first half-side of a ping-pong memory bank in a programmable vector processor (“PVP”);

determining multiple DFT jobs bundled in the first mega-job from and identifying a first point size associated with the multiple DFT jobs within the first mega-job;

determining a radix factorization for calculating DFT in response to the first point size; and

generating one or more radix engine control signals to program a mixed radix engine for providing DFT calculations based on the radix factorization.

10. The method of claim 9 , further comprising configuring a configurable vector mixed-radix engine to facilitate a first radix computation during a first stage of the radix factorization for the first mega-job.

11. The method of claim 10 , further comprising retrieving the first mega-job from the first half-side of a ping-pong memory bank.

12. The method of claim 11 , further comprising loading the first mega-job into a vector pipeline for scaling vector data and calculating twiddle multiplication.

13. The method of claim 12 , further comprising performing a programmed radix computation in response to the first mega-job and the first point size.

14. The method of claim 9 , further comprising:

identifying a second mega-job stored in a second half-side of a ping-pong memory bank;

determining multiple DFT jobs bundled in the second mega-job from and identifying a second point size associated with the multiple DFT jobs within the second mega-job; and

determining a second radix factorization for calculating DFT in response to the second point size.

15. A method of processing discrete Fourier transform (“DFT”) for wireless data transmission in a communication network, the method comprising:

configuring a configurable vector mixed-radix engine for performing a first radix computation during a first stage for a radix factorization of a first mega-job which bundles a plurality of DFT jobs having a first point size;

retrieving the first mega-job from the first half-side of a ping-pong memory bank;

loading the first mega-job into a vector pipeline for scaling and twiddle multiplying; and

performing a programmed radix computation in response to the first mega-job and the first point size.

16. The method of claim 15 , further comprising receiving a data stream containing a plurality of DFT jobs, wherein receiving a data stream further includes identifying a first point size for a first portion of DFT jobs and identifying a second point size for a second portion of DFT jobs.

17. The method of claim 15 , further comprising generating a first mega-job configured to bundle at least a portion of the DFT jobs with the first DFT point size in a predefined sequence.

18. The method of claim 15 , further comprising storing the first mega-job at the first half-side of the ping-pong memory bank before allowing the first mega-job to be fed to a vector pipeline for radix computation.

19. The method of claim 15 , further comprising determining a radix factorization for calculating DFT in response to the first DFT point size.

20. The method of claim 15 , further comprising identifying a first mega-job stored in a first half-side of a ping-pong memory bank.

21. The method of claim 15 , further comprising determining multiple DFT jobs bundled in the first mega-job from and identifying a first point size associated with the multiple DFT jobs within the first mega-job.

22. The method of claim 15 , further comprising determining a radix factorization for calculating DFT in response to the first point size.

23. The method of claim 15 , further comprising generating one or more radix engine control signals to program a mixed radix engine for providing DFT calculations based on the radix factorization.

24. The method of claim 23 , wherein the generating one or more radix engine control signals further includes configuring a mixed radix engine to perform a selected one of radix3, radix4, radix5, and radix6 computations for each stage.

25. An apparatus for processing discrete Fourier transform (“DFT”) for data transmission comprising:

means for receiving a data stream containing a plurality of DFT jobs wherein a first portion of DFT jobs has a first DFT point size and a second portion of DFT jobs has a second DFT point size;

means for generating a first mega-job configured to bundle at least a portion of the DFT jobs with the first DFT point size in a predefined sequence;

means for storing the first mega-job at a ping half-side of a ping-pong memory bank before allowing the first mega-job to be fed to a vector pipeline for radix computation; and

means for determining a radix factorization for calculating DFT in response to the first DFT point size.

26. The apparatus of claim 25 , further comprising means for generating a second mega-job configured to bundle at least a portion of the DFT jobs with the second DFT point size in a predefined second sequence.

27. The apparatus of claim 25 , further comprising means for storing the second mega-job at another ping half-side of the ping-pong memory bank before allowing the second mega-job to be fed to the vector pipeline for radix computation.

28. An apparatus of processing discrete Fourier transform (“DFT”) for wireless data transmission in a communication network, the method comprising:

means for configuring a configurable vector mixed-radix engine for performing a first radix computation during a first stage for a radix factorization of a first mega-job which bundles a plurality of DFT jobs having a first point size;

means for retrieving the first mega-job from the first half-side of a ping-pong memory bank;

means for loading the first mega-job into a vector pipeline for scaling and twiddle multiplying; and

means for performing a programmed radix computation in response to the first mega-job and the first point size.

29. The apparatus of claim 28 , further comprising means for receiving a data stream containing a plurality of DFT jobs, wherein means for receiving a data stream further includes means for identifying a first point size for a first portion of DFT jobs and means for identifying a second point size for a second portion of DFT jobs.

30. The apparatus of claim 28 , further comprising means for generating a first mega-job configured to bundle at least a portion of the DFT jobs with the first DFT point size in a predefined sequence.

31. The apparatus of claim 28 , further comprising means for storing the first mega-job at the first half-side of the ping-pong memory bank before allowing the first mega-job to be fed to a vector pipeline for radix computation.

32. The apparatus of claim 28 , further comprising means for determining a radix factorization for calculating DFT in response to the first DFT point size.

33. The apparatus of claim 28 , further comprising means for identifying a first mega-job stored in a first half-side of a ping-pong memory bank.