Instruction and method for fused rake-finger operation on a vector processor

• US 9,276,778 B2
• Filed: 01/31/2014
• Issued: 03/01/2016
• Est. Priority Date: 01/31/2014
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Khan, Raheel, Hsu, Steve Hengchen, Singaravelu, Gayatri
• Primary Examiner(s): LAM, KENNETH T
• Application Number:
  US14/170,234
• Publication Number:
  US 20150222453A1
• Time in Patent Office: 760 Days
• US Class Current: 1/1
• CPC Class Codes: G06F 9/30 Arrangements for executing machine instructions, e.g. instruction decode for executing microinstructions G06F9/22
• CPC Class Codes: G06F 9/30036 Instructions to perform operations on packed data, e.g. vector, tile or matrix operations
• CPC Class Codes: H04B 1/7115 Constructive combining of multi-path signals, i.e. RAKE receivers
• CPC Class Codes: H04B 1/7117 Selection, re-selection, allocation or re-allocation of paths to fingers, e.g. timing offset control of allocated fingers
• CPC Class Codes: H04L 25/0204 of multiple channels
• CPC Class Codes: H04L 25/0212 of impulse response
• CPC Class Codes: H04L 25/0224 using sounding signals

Abstract

Systems and methods for performing a rake-finger operation on a vector processor are described herein. In one embodiment, a method for rake-finger processing comprises loading samples from a register into an execution unit, performing a rake-finger operation on the samples in the execution unit, and writing results from the rake-finger operation to the register. Performing the rake-finger operation comprises performing a finite impulse response (FIR) filter operation, and performing a despread operation, wherein filtered samples from the FIR filter operation are input to the despread operation without going through the register.

Claims

1. A method for rake-finger processing, comprising:

loading samples from a register into an execution unit;

performing a rake-finger operation on the samples from the register in the execution unit, wherein performing the rake-finger operation comprises;

performing a finite impulse response (FIR) filter operation on the samples from the register to produce first filtered samples and second filtered samples, the first filtered samples being offset in time from the second filtered samples; and

performing a despread operation on the first filtered samples and the second filtered samples, wherein the first filtered samples and the second filtered samples from the FIR filter operation are input to the despread operation without going through the register, and performing the despread operation comprises;

despreading the first filtered samples with a first code into first despreaded samples; and

despreading the second filtered samples with a time-shifted version of the first code into second despreaded samples; and

writing results from the rake-finger operation to the register.

2. The method of claim 1, wherein performing the rake-finger operation further comprises performing a rotation operation on the samples from the register before the samples are input to the FIR filter operation.

3. The method of claim 2, wherein the rotation operation compensates for a carrier frequency offset in the samples from the register.

4. The method of claim 3, wherein performing the rotation operation comprises:

generating phases, and

rotating the samples from the register by the generated phases to compensate for the carrier frequency offset.

5. The method of claim 1, wherein performing the despread operation comprises

despreading the second filtered samples with a second code into third despreaded samples.

6. The method of claim 5, wherein the second code comprises a pilot code.

7. The method of claim 1, wherein performing the despread operation comprises:

accumulating the first despreaded samples over a length of the first code to generate a first symbol; and

accumulating the second despreaded samples over a length of the time-shifted version of the first code to generate a second symbol.

8. The method of claim 7, wherein the first and second symbols are spaced a half a chip apart or one chip apart.

9. The method of claim 7, further comprising delaying the first code by one chip to generate the time-shifted version of the first code.

10. The method of claim 1, further comprising configuring the execution unit to perform the rake-finger operation using a single instruction.

11. An apparatus for rake-finger processing, comprising:

means for loading samples from a register into an execution unit;

means for performing a rake-finger operation on the samples from the register in the execution unit, wherein the means for performing the rake-finger operation comprises;

means for performing a finite impulse response (FIR) filter operation on the samples from the register to produce first filtered samples and second filtered samples, the first filtered samples being offset in time from the second filtered samples; and

means for performing a despread operation on the first filtered samples and the second filtered samples, wherein the first filtered samples and the second filtered samples from the FIR filter operation are input to the despread operation without going through the register, and the means for performing the despread operation comprises;

means for despreading the first filtered samples with a first code into first despreaded samples; and

means for despreading the second filtered samples with a time-shifted version of the first code into second despreaded samples; and

means for writing results from the rake-finger operation to the register.

12. The apparatus of claim 11, wherein the means for performing the rake-finger operation further comprises means for performing a rotation operation on the samples from the register before the samples are input to the means for performing the FIR filter operation.

13. The apparatus of claim 12, wherein the rotation operation compensates for a carrier frequency offset in the samples from the register.

14. The apparatus of claim 13, wherein the means for performing the rotation operation comprises:

means for generating phases, and

means for rotating the samples from the register by the generated phases to compensate for the carrier frequency offset.

15. The apparatus of claim 11, wherein the means for performing the despreading operation comprises

means for despreading the second filtered samples with a second code into third despreaded samples.

16. The apparatus of claim 15, wherein the second code comprises a pilot code.

17. The apparatus of claim 11, wherein the means for performing the despread operation comprises:

means for accumulating the first despreaded samples over a length of the first code to generate a first symbol; and

means for accumulating the second despreaded samples over a length of the time-shifted version of the first code to generate a second symbol.

18. The apparatus of claim 17, wherein the first and second symbols are spaced a half a chip apart or one chip apart.

19. The apparatus of claim 17, further comprising means for delaying the first code by one chip to generate the time-shifted version of the first code.

20. An execution unit, comprising:

a read circuit configured to load samples from a register into the execution unit;

a rake-finger circuit configured to perform a rake-finger operation on the samples from the register in the execution unit, wherein the rake-finger circuit comprises;

a finite impulse response (FIR) filter circuit configured to perform an FIR filter operation on the samples from the register to produce first filtered samples and second filtered samples, the first filtered samples being offset in time from the second filtered samples; and

a despread circuit configured to perform a despread operation on the first filtered samples and the second filtered samples, wherein the first filtered samples and the second filtered samples from the FIR filter circuit are input to the despread circuit without going through the register, and the despread circuit comprises;

a first despreader configured to despread the first filtered samples with a first code into first despreaded samples; and

a second despreader configured to despread the second filtered samples with a time-shifted version of the first code into second despreaded samples; and

a write circuit configured to write results from the rake-finger circuit to the register.

21. The execution unit of claim 20, wherein the rake-finger circuit further comprises a rotator configured to perform a rotation operation on the samples from the register before the samples are input to the FIR filter circuit.

22. The execution unit of claim 21, wherein the rotation operation compensates for a carrier frequency offset in the samples from the register.

23. The execution unit of claim 20, wherein the despread circuit comprises

a third despreader configured to despread the second filtered samples with a second code into third despreaded samples.

24. The execution unit of claim 23, wherein the second code comprises a pilot code.

25. The execution unit of claim 20, wherein the despread circuit comprises:

a first accumulator configured to accumulate the first despreaded samples over a length of the first code to generate a first symbol; and

a second accumulator configured to accumulate the second despreaded samples over a length of the time-shifted version of the first code to generate a second symbol.

26. The execution unit of claim 25, wherein the first and second symbols are spaced a half a chip apart or one chip apart.

27. The execution unit of claim 20, wherein the FIR filter circuit comprises:

a delay line comprising a plurality of delay elements, wherein the samples from the register propagate down the delay line;

a first multiply and merge circuit coupled to a first plurality of the delay elements, wherein the first multiply and merge circuit is configured to produce each one of the first filtered samples by receiving samples from the first plurality of the delay elements during a respective one of a plurality of clock cycles, multiplying the received samples by respective coefficients to produce respective products, and summing the respective products to produce the one of the first filtered samples; and

a second multiply and merge circuit coupled to a second plurality of the delay elements, wherein the second multiply and merge circuit is configured to produce each one of the second filtered samples by receiving samples from the second plurality of the delay elements during a respective one of the plurality of clock cycles, multiplying the received samples by respective coefficients to produce respective products, and summing the respective products to produce the one of the second filtered samples.

28. The execution unit of claim 27, wherein the first plurality of the delay elements and the second plurality of the delay elements overlap.

29. The execution unit of claim 28, wherein the offset in time is half a chip period.

30. The execution unit of claim 20, wherein the despread circuit further comprises a third despreader configured to despread the first filtered samples with the time-shifted version of the first code into third despreaded samples.

31. The execution unit of claim 30, wherein the offset in time is half a chip period and the time-shifted version of the first code is generated by delaying the first code by one chip period.