How PIM Over CPRI Makes Field Testing More Efficient

December 7, 2018

The introduction of RF passive intermodulation (PIM) testing using live traffic over the Common Public Radio Interface (CPRI) link – commonly referred to as PIM over CPRI – has made verifying network performance more efficient. Operators save capital expense (CapEx) and operating expense (OpEx) while field technicians do not have to unnecessarily scale an antenna tower.

PIM over CPRI utilizes multiple unique techniques to measure PIM more efficiently. The method allows tests to be made from the ground to reduce the need to deploy tower climbing crews and eliminate the need to take the site down. In this post, we will review some additional benefits, as well as explain procedures to more accurately measure PIM.

PIM over CPRI Advantages

One key benefit is that PIM measurements made over a CPRI link are frequency-independent because the CPRI IQ data is baseband. That means a single measurement solution can conduct measurements on any LTE frequency band.

The biggest advantage of PIM over CPRI, however, is that it uses live traffic to conduct the measurement, allowing the site to stay on while tests are being done. Conventional methods send two continuous wave (CW) tones to try and locate PIM. By using live traffic, PIM over CPRI establishes a real-world PIM scenario, helping to discover PIM that is lost using the conventional method.

Conducting PIM Over CPRI

To perform the measurement, turn to max power on all unused resource blocks (RBs), which forces the radio to maximum Tx power. (Note: network performance is maintained via this approach, as RBs for live traffic are not used.) A worst-case PIM scenario is created so more interference sources will be seen. This approach will work for macro and DAS systems alike.

Figure 1 shows conceptually how PIM over CPRI can locate PIM. The righthand side shows a 2x2 MIMO in the downlink (DL) in which the subcarriers are mixing. This creates potential intermodulation products on the uplink (UL) shown on the left.

Figure 1

The PIM over CPRI algorithm looks at all the DLs and the UL simultaneously to conduct a more accurate measurement. A key element is that the measurement must be made in real time to align the UL and DLs. By doing so, the correlated PIM (shown in figure 2) can be discovered.

Figure 2

PIM Desensitization and Correlated PIM Value

Accurate PIM measurements are accomplished by using a different method then traditional RF PIM tests that measure dBc/dBm using two CW tones. PIM over CPRI uses two unique features:

PIM Desensitization – This capability replaces dBc by establishing a relative value using dB. It uses LTE carrier RBs to quantify how much PIM degrades the UL noise floor. There is a general industry consensus that 10 dB desensitization affects data throughput by 50%, making this analysis critical to networks achieving KPIs.

Figure 3 shows how to establish the PIM desensitization advanced settings, which can be done on one screen of the BTS Master MT8220T, when the PIM over CPRI option is installed. Users simply input the site configuration – either SISO, 2x2 MIMO or 2x4 – and the PIM desensitization Pass/Fail setting. Fields for the DL/UL configuration and UL under test settings must also be filled for the measurement to be conducted.

Figure 3

Correlated PIM – This is determined in dBFS/dBm. Base station OEMs use dBFS (dB Full Scale) as a digital level that shows dB relative to the maximum value that can be represented on the CPRI link. To convert dBFS to dBm, the user needs to know the specific radio thermal noise floor specification and noise figure values. Both can be provided by the respective equipment OEM.

Figure 4 shows the effectiveness of PIM over CPRI. A nearby water tower reflected PIM back to the antenna, causing interference on the 4^th UL. PIM desensitization value was 3.34 dB, which affected the UL service quality. There was also a correlated PIM value of -63.7 dBFS, which was higher than the thermal noise floor of the radio. The result was a 3.3 dB rise in the noise floor, degrading signal transmission.

Figure 4

Conversely, figure 5 is an example of how a display may look ominous but network performance is not affected. The purple trace shows the typical shark fin that generally causes issues but, in this case, the correlated PIM value is -122 dBm. The total UL power is only -102 dBm, which is too small to impact radio performance. PIM desensitization value is only 0.04 dB, so it has virtually no impact on the UL noise floor.

Figure 5