1. Function call structure

- main()

	- SamsungDiamondVoiceTxEngineInit()

		- DiamondVoiceEngineTxECNSINIT(): Initialize config parameters and frame status here

	- SamsungDiamondVoiceTxEngineModeConfig(): Set turn on/off mode

		- SamsungDiamondVoiceTxEngineMode()

			- SamsungTxECNSModeConfig() 

	- SamsungDiamondVoiceTxEngineECNSConfig(): Not used (will be used to input tuning parameter later)

		- SamsungTxECNSReserved1Config()

		- SamsungTxECNSReserved2Config()

		- SamsungTxECNSReserved3Config()

	- SamsungDiamondVoiceTxEngineExe()

		- SamsungDiamondVoiceTxEngineInOutConfig(): Set input/output buffer

		- SamsungTxECNSState(): Call mode (turn on/off)

		- SamsungTxECNSExe()

			- SamsungTxECNS_NS(): NS main module function



2. Input argument

<Input_tx_main_mic.pcm> <Input_tx_aux_mic.pcm> <Input_rx.pcm> <output_tx.pcm>

<Input_rx.pcm> is used for Acoustic Echo Cancellation and it is not implemented yet.

So, you can use any pcm file with same length of <Input_tx_main_mic.pcm>. (Even it is okay to use <Input_tx_main_mic.pcm>)



3. Instruction for SRC-B

To Add config parameters (not modified with respect to frame index),

 - Definitions are added in struct DiamondVoiceTxEngine_ECNS_Cfg_t (samsungDiamondVoiceECNSTx.h)

 - Initialization are done in DiamondVoiceEngineTxECNSINIT() (SamsungTxECNS.c)

 

To Add frame status (such as buffers and information of current and previous frames),

 - Definitions are added in struct ECNSStatus (samsungDiamondVoiceECNSTx.h)

 - Initialization are done in DiamondVoiceEngineTxECNSINIT() (SamsungTxECNS.c)

 

To Add DNN NS regression function,

 - Add DNN NS at Fx_DNN_regression() (SamsungTxECNSRegression.c)

 - Call Fx_DNN_regression() at Fx_SpeechRegression() (SamsungTxECNSAlg.c)

 - To add/modify prototypes, refer to samsungTxECNSInt.h