RadarCaponDoAROI (Direction of Arrival)#

Macros#

PVA_DOA_CAPON_ROI_MAX_ROIS

Maximum number of ROIs supported by the Capon ROI DoA operator.

Functions#

NVCVStatus pvaDoaCaponROICreate(NVCVOperatorHandle *handle, NVCVTensorRequirements const *const snapshotsTensorRequirements, NVCVTensorRequirements const *const steeringVectorsTensorRequirements, NVCVTensorRequirements const *const peakPowerTensorRequirements, NVCVTensorRequirements const *const azimuthBinIndicesTensorRequirements, NVCVTensorRequirements const *const elevationBinIndicesTensorRequirements, PVADoaCaponROIParams const *const params)

Constructs the Capon ROI DoA operator.

NVCVStatus pvaDoaCaponROISubmit(NVCVOperatorHandle handle, cudaStream_t stream, NVCVTensorHandle const snapshotsROI, NVCVTensorHandle const steeringVectorsROI, NVCVTensorHandle peakPower, NVCVTensorHandle azimuthBinIndices, NVCVTensorHandle elevationBinIndices)

Submits the Capon ROI DoA operator to a CUDA stream.

NVCVStatus pvaDoaCaponROISubmit(NVCVOperatorHandle handle, cupvaStream_t stream, NVCVTensorHandle const snapshotsROI, NVCVTensorHandle const steeringVectorsROI, NVCVTensorHandle peakPower, NVCVTensorHandle azimuthBinIndices, NVCVTensorHandle elevationBinIndices)

Submits the Capon ROI DoA operator to a cuPVA stream.

Functions#

NVCVStatus pvaDoaCaponROICreate(
NVCVOperatorHandle *handle,
NVCVTensorRequirements const *const snapshotsTensorRequirements,
NVCVTensorRequirements const *const steeringVectorsTensorRequirements,
NVCVTensorRequirements const *const peakPowerTensorRequirements,
NVCVTensorRequirements const *const azimuthBinIndicesTensorRequirements,
NVCVTensorRequirements const *const elevationBinIndicesTensorRequirements,
PVADoaCaponROIParams const *const params,
)#

Constructs the Capon ROI DoA operator.

The operator allocates its own device scratch for the intermediate L_inv matrices passed between its internal stages; this scratch is released by nvcvOperatorDestroy.

Parameters

Parameters:
  • handle – [out] Pointer to storage for the operator instance handle.

    • Must not be NULL.

  • snapshotsTensorRequirements – [in] Pointer to the NVCVTensorRequirements structure that describes the snapshots tensor.

    • Data Layout: [NHW]

    • Tensor Rank: 3

    • Tensor Shape: [numROIs, numDopplerBinsCovMat, svDim]

    • Data Type: NVCV_DATA_TYPE_2S32

    • The two S32 channels store the fixed-point real and imaginary components of each snapshot sample.

    • Each ROI must be contiguous; padding between ROIs is permitted.

  • steeringVectorsTensorRequirements – [in] Pointer to the NVCVTensorRequirements structure that describes the steering-vector tensor.

    • Data Layout: [NCHW]

    • Tensor Rank: 4

    • Tensor Shape: [numROIs, svDim, numElBins, numAzBins]

    • Data Type: NVCV_DATA_TYPE_2S32

    • The two S32 channels store the fixed-point real and imaginary components of each steering-vector element.

    • The tensor must be contiguous.

  • peakPowerTensorRequirements – [in] Pointer to the NVCVTensorRequirements structure that describes the peak-power output tensor.

    • Data Layout: [HW]

    • Tensor Rank: 2

    • Tensor Shape: [numROIs, numPeaksPerROI]

    • Data Type: NVCV_DATA_TYPE_F32

    • Each ROI’s values must be contiguous; padding between ROIs is permitted.

  • azimuthBinIndicesTensorRequirements – [in] Pointer to the NVCVTensorRequirements structure that describes the azimuth-bin-index output tensor.

    • Data Layout: [HW]

    • Tensor Rank: 2

    • Tensor Shape: [numROIs, numPeaksPerROI]

    • Data Type: NVCV_DATA_TYPE_F32

    • Each ROI’s values must be contiguous; padding between ROIs is permitted.

  • elevationBinIndicesTensorRequirements – [in] Pointer to the NVCVTensorRequirements structure that describes the elevation-bin-index output tensor.

    • Data Layout: [HW]

    • Tensor Rank: 2

    • Tensor Shape: [numROIs, numPeaksPerROI]

    • Data Type: NVCV_DATA_TYPE_F32

    • Each ROI’s values must be contiguous; padding between ROIs is permitted.

  • params – [in] Pointer to the PVADoaCaponROIParams structure that contains the operator configuration.

Return values:
  • NVCV_ERROR_INVALID_ARGUMENT – An argument is null or violates the operator requirements.

  • NVCV_ERROR_OUT_OF_MEMORY – Not enough memory to create the operator.

  • NVCV_SUCCESS – Operator created successfully.

NVCVStatus pvaDoaCaponROISubmit(
NVCVOperatorHandle handle,
cudaStream_t stream,
NVCVTensorHandle const snapshotsROI,
NVCVTensorHandle const steeringVectorsROI,
NVCVTensorHandle peakPower,
NVCVTensorHandle azimuthBinIndices,
NVCVTensorHandle elevationBinIndices,
)#

Submits the Capon ROI DoA operator to a CUDA stream.

All tensors must match the requirements passed to pvaDoaCaponROICreate.

Note

CUDA stream support requirements:

  • PVA SDK 2.7.0 or later

  • Jetpack 7 or later

  • DriveOS 7 or later

  • x86 Emulator is not supported

Parameters:
  • handle – [in] Handle to the operator instance returned by pvaDoaCaponROICreate.

    • Must not be NULL.

  • stream – [in] Handle to a valid CUDA stream.

  • snapshotsROI – [in] Input tensor handle containing the complex snapshot ROIs.

  • steeringVectorsROI – [in] Input tensor handle containing the complex steering vectors.

  • peakPower – [out] Output tensor handle containing the peak powers.

  • azimuthBinIndices – [out] Output tensor handle containing the azimuth bin indices.

  • elevationBinIndices – [out] Output tensor handle containing the elevation bin indices.

Return values:
  • NVCV_ERROR_INVALID_ARGUMENT – An argument is null or a tensor does not match the operator configuration.

  • NVCV_ERROR_NOT_COMPATIBLE – The handle does not refer to a Capon ROI operator.

  • NVCV_ERROR_DEVICE – Command submission failed.

  • NVCV_ERROR_INTERNAL – An internal operator error occurred.

  • NVCV_SUCCESS – Operation submitted successfully.

NVCVStatus pvaDoaCaponROISubmit(
NVCVOperatorHandle handle,
cupvaStream_t stream,
NVCVTensorHandle const snapshotsROI,
NVCVTensorHandle const steeringVectorsROI,
NVCVTensorHandle peakPower,
NVCVTensorHandle azimuthBinIndices,
NVCVTensorHandle elevationBinIndices,
)#

Submits the Capon ROI DoA operator to a cuPVA stream.

All tensors must match the requirements passed to pvaDoaCaponROICreate.

Parameters:
  • handle – [in] Handle to the operator instance returned by pvaDoaCaponROICreate.

    • Must not be NULL.

  • stream – [in] Handle to a valid cuPVA stream.

  • snapshotsROI – [in] Input tensor handle containing the complex snapshot ROIs.

  • steeringVectorsROI – [in] Input tensor handle containing the complex steering vectors.

  • peakPower – [out] Output tensor handle containing the peak powers.

  • azimuthBinIndices – [out] Output tensor handle containing the azimuth bin indices.

  • elevationBinIndices – [out] Output tensor handle containing the elevation bin indices.

Return values:
  • NVCV_ERROR_INVALID_ARGUMENT – An argument is null or a tensor does not match the operator configuration.

  • NVCV_ERROR_NOT_COMPATIBLE – The handle does not refer to a Capon ROI operator.

  • NVCV_ERROR_DEVICE – Command submission failed.

  • NVCV_ERROR_INTERNAL – An internal operator error occurred.

  • NVCV_SUCCESS – Operation submitted successfully.