How to use the USRP-2921 to monitor wideband spectrum

NI USRP-292X software-programmable wireless transceivers are developed specifically for teaching and researching radio communications, as a teaching solution for learning RF and communications at a moderate price. National Instruments combines NI LabVIEW software with hardware capabilities from Ettus Research to develop these affordable software-defined radio (SDR) transceivers, which greatly help communications education, experimentation, research, and rapid source code development.

USRP-2921 Hardware Introduction:

figure 1

Tunable frequency band range: 2.4 GHz - 5 GHz for Wi-Fi, Bluetooth and other ISM applications

Reasonably priced teaching and research solutions

Up to 20 MHz baseband I/Q bandwidth, capable of reading and writing data at 25 MS/s for data processing with NI LabVIEW on the host computer

NI technical support and one-year, extendable warranty

Compatible with Windows 7/Vista/XP

USRP-2921 front panel:

The USRP-2921 has two transmit and receive ports, but they cannot transmit and receive at the same time. The MIMO expansion slot supports 2x2 MIM systems.

figure 2

USRP Software Introduction:

URSP software development uses the standard development process of open, configure, initiate, Read/Write, Abort, and close to write code.

image 3

2. Detection of broadband spectrum

From the introduction of USRP-2921, we can know that USRP-2921 has a baseband I/Q bandwidth of up to 20MHz and can read and write data at a rate of 25MS/S, which is basically enough for general applications, but sometimes developers still want to see a wider bandwidth. Take WiFi signals as an example. WiFi has 11 channels, each with a bandwidth of 5MHz. If you want to see the spectrum of the signal within the entire bandwidth, you have to detect a bandwidth of about 55MHz, which is difficult to achieve for some traditional instruments. However, the USRP-292X software programmable wireless transceiver developed by NI is a software programmable wireless transceiver that can realize spectrum detection through software programming. NI combines the advantages of NI USRP-292x hardware and NI LabVIEW software to provide a powerful and flexible software radio platform for rapid prototyping of wireless communication systems. Based on the intuitive graphical programming language NI LabVIEW to complete the signal processing algorithm and combined with NI USRP hardware to interact with real RF signals in real time, the prototype development of a complete wireless communication system can be realized.

Let's take the detection of the spectrum of a 50MHz bandwidth WiFi signal as an example, where the RBW (Resolution Bandwith) is 100KHz.

There are two things we need to pay attention to when developing a program:

1. USRP-2921 directly multiplies the frequency, which may cause the problem of local oscillator leakage, which is manifested as a DC signal at zero frequency, as shown in Figure 4.

2. According to experiments, it can be observed that when the bandwidth is 25MHz, the power of the 2.5MHz bandwidth on the left and right sides will be attenuated by about 3dB, as shown in Figure 4.

Figure 4

Before using USRP, you must first install the USRP driver. After installing the driver, you can go to Instrument IO->Instrument Driver->NI-USRP Toolkit, as shown in Figure 5.

Figure 5

The toolkit includes Rx, Tx, Synchronization, Utility and other VIs. Here we only need to use the Rx VI, which includes open, configure, initiate, fetch, abort and close VIs, as shown in Figure 6.

Figure 6

Figure 7 shows the code for sampling a 50MHz bandwidth spectrum. The process is similar to the general acquisition code. However, it is necessary to pay attention to the two problems mentioned above. One is the problem of local oscillator leakage, and the other is the problem of edge power reduction. For the problem of edge power reduction, the center frequency can be offset, and 25MHz bandwidth can be sampled, and then the 5M data at the edge can be discarded, and only 20M data can be used; for the problem of local oscillator leakage, the method of discarding numbers can be used, that is, repeatedly sampling a section of data at a certain center frequency, and then discarding the data in the previous section, and only retaining the data in the latter section for operation. Of course, a better method is to offset the carrier frequency to avoid zero frequency.

Figure 7

The method of detecting wideband spectrum is to splice spectrum segments into a wider bandwidth spectrum. As shown in Figure 8, a 50MHz bandwidth spectrum is divided into five 10MHz bandwidth spectrums, and then USRP is used to sample five 25M bandwidth spectrums, but only the 10M data from -2M to 12M on the left is taken, and finally it is spliced ​​into a 50M spectrum. This can effectively avoid the problems of edge power reduction and local oscillator leakage.

Figure 8

First, the carrier frequency of the first 25M bandwidth spectrum needs to be calculated, Fc1=Fc-25M+10.5M, as shown in the mark 1 of Figure 9. After that, the carrier frequency of each segment increases by 10M, that is, Fc5=Fc4+10M; Fc4=Fc3+10M; Fc3=Fc2+10M; Fc2=Fc1+10M; as shown in the mark 2 of Figure 9.

Fig. 9

One thing to note is that each time you change the carrier frequency, you need to stop the last acquisition and restart the acquisition when the new carrier frequency is configured.

In addition, in order to solve the problem of local oscillator leakage, 2.5M data can be collected as shown in Figure 10, and then the front data is discarded, and only the back 250K data is retained, which can effectively reduce the impact of local oscillator leakage.

Fig.10

Since we need to collect WiFi signals, which are emitted in pulses, and USRP is not triggered, we can only compare the segment with the largest mean square value as the collected WiFi signal, as shown in Figure 11.

Figure 11Max Bur.vi

Then perform FFT on the data found by Max Bur, and finally splice the 10M bandwidth spectrum with an array, as shown in Figure 12, until a 50M bandwidth spectrum is spliced ​​out and output, and finally convert the output voltage value into a power value.

Fig.12

Conclusion

According to the above method, it is theoretically possible to obtain a wider bandwidth spectrum, but obtaining a wider bandwidth spectrum means that it takes more time to collect, and the bandwidth of WiFi and Bluetooth is also limited, so detecting a wider bandwidth is not very meaningful. This article only provides a way to obtain a wider bandwidth by splicing spectrum by taking the acquisition of WiFi signals as an example, and using LabVIEW software to develop programs can effectively reduce the development cycle, so that some principle verification or research can be done in a short time.