VOICE TRANSMISSION USING NI USRP AND LABVIEW

[ashwini11]

Thank you so much for your suggestion. Now my system is working for real time voice transmission. 

Instead of using Queues we can also use the concept of local variables here. (just for an information)

[wiebe@CARYA]

---

@ashwini11 wrote:

Thank you so much for your suggestion. Now my system is working for real time voice transmission. 

Instead of using Queues we can also use the concept of local variables here. (just for an information)

---

I'm not saying you can't use local variables, but they are conceptually quite difference from queues.

Local variables have a value, queue's have a stack of values. If you don't want to lose values, a queue is better then a local variable. Also, each local variable forces a memory copy of the data. That could be very inefficient compared to a queue.

Value properties are a third option (there are more). They work like local variables, but don't require memory copies. Downside is reading\writing them is synchronized with the screen updates.

[ashwini11]

Hello,

Here i have attached my VI that is OFDM system for text transmission. Bit loss is there at the output side. (ex: if i say hello sir, i am getting just 'hello')  My doubt is here for different symbol mapping technique the number of bits that i am getting at the output side is varying. but when mapping and de-mapping is done separately i didn't get any bit loss. why it is so? I am not able to figure out where exactly the bit is getting lost. could you please help me??

[ashwini11]

Hello,

Here i have attached my VI that is OFDM system for text transmission. Bit loss is there at the output side. (ex: if i say hello sir, i am getting just 'hello')  My doubt is here for different symbol mapping technique the number of bits that i am getting at the output side is varying. but when mapping and de-mapping is done separately i didn't get any bit loss. why it is so? I am not able to figure out where exactly the bit is getting lost. could you please help me??

[SamuRei]

Another suggestion for the dataflow in your VI: Because you have moved to using queues (a good call, by the way), your two loops are running in parallel. These are typically placed vertically, one above the other, in producer-consumer architecture. Just a note for the readability of your code.

Now for clarification, I'm not entirely sure what the difference between your mapping and demapping questions go. It sounds like this is the differentiation between it working and not. Can you elaborate further on what's going on? Maybe a VI with a working and non-working case would be helpful. 

Additionally, I would probably start troubleshooting this by probing the data at various points to see where data is being lost. I install these probes on any machine that I intend to to SDR / RF development on: LabVIEW Probes for Easy RF Waveform Viewing

Take a look at these resources and let me know how things work out for you!

-Sam

[ashwini11]

Hello,

I have attached my VI implementing  SSB modulation technique. Here i am getting an error saying that 'dt values' of 2 waveforms are different, when tried to add 2 signals of different phase. I tried using align and resample VI, provided dt value extracted from original signal (0 degree phase). But this is showing a pop up saying memory is full. could you please help me in this regard. 

[psisterhen]

When you are re-creating the signals wire after the Hilbert transform, it doesn't have any dt (sample period) data, so the types don't match when you add that back with the original signals.

The easiest thing is to not use the Simulate Signal Express VIs.  Just use Sine Waveform.vi from the Waveform Generation palette to create your signals and use the Waveform Data Types.