IWR1443: FFT accelerator performance in speed ?

Hi Naoki,

FFT is performed by hardware accelerator in IWR1443. HWA runs on clock or 200MHz. As to bench mark value, you can refer to hwa user's guide www.ti.com/.../swru526.pdf, especially page 31.

BPM removal was discussed in hwa user's guide part 2. www.ti.com/.../swru527a.pdf
Could you give more details on beam forming case, we can reach to the expert to help.

Thanks and Regards,
Michelle

Hi Michelle,

Thanks for your reply. Yes, i found the numbers in the document. Thanks.

Michelle Liu said:

Could you give more details on beam forming case, we can reach to the expert to help.

I`m not specialist in RF, but it seems like they are considering to use the following configuration in RF part. I`m wondering if this might not be adapted to BOOST EVM directly, but might be feasible in their custom board. What do you think ?

Best Regards,
NK 

Hi Michelle,

I found the following information from IWR1443 data sheet:

 

The three transmit channels can be operated up to a maximum of two at a time (simultaneously) for transmit beamforming purpose as required; whereas the four receive channels can all be operated simultaneously.

 

Hence, I did not understand why you said IWR1443 is not support beamforming. Could you please explain this in details? 

We have IWR1443 and IWR1642 EVMs. We want to implement our own beamforming and detection and classification methods for data processing. Could you please suggest which EVM is more suitable for our goal? And which tools are needed to implement beamforming?

Thank you.

Regards,
Jiadi

Jiadi,

Sorry for the confusion. The expert I talked to on beam forming and the datasheet writer may have slight different understanding on the definition of beam forming. The former thinks beam forming should support any value of phase shift, therefore, neither 14xx nor 16xx supports that. The datasheet claims the capability of beamforming which is actually meant for binary phase shift (both 14xx and 16xx can support binary phase shift) and the phase shift is only one value -- 180 degree. Let me know what is desirable in your application.

Thanks and Regards,
Michelle

Hi Michelle,

I am just confused with these explanations. We do not need antenna beam forming in our application.

Can you help me to check the feasibility of using beam forming algorithm (such as MUSIC algorithm) to replace the role of 3D FFT ( for angle information)?

Many thanks.

Regards,
Jiadi

Jiadi,

Regardless how we call it, xwr14xx and xwr16xx can only support binary phase shift, not arbitrary phase shift. If MUSIC algorithm requires phase rotation other than 180 degree, our radar devices can not support it.

Thanks and Regards,
Michelle

Hi,

Another question. As I described in the picture in the previous post, my customer wants to use a waveguide (rather than patch antenna) with IWR1443. Is it possible ?

Best Regards,
NK 

Hi

Yes a circulator can work. The leakage power from Tx -> antenna -> Rx is a concern, typically that has to be at least -35db.
The antenna to receiver has a very low loss, typically 1-2db. The Tx to antenna also should be low loss.

The amp is not usually in the place shown.

IWR1443 -> tx -> Amp -> Circulator -> attenuator -> External LNA -> Rx
<->TDD Ant

Thanks and Regards,
Michelle