Advanced Radio Frequency Interference Mitigation Strategies for the National Weather Radar Testbed

Advanced Radio Frequency Interference Mitigation Strategies for the National Weather Radar Testbed

 

J. Lake¹, C. Curtis², and M. Yeary¹

¹Advanced Radar Research Center, Univ. of Okla., Norman, OK

²National Severe Storms Lab, NOAA, Norman, OK

 

The Multi-function Phased Array Radar program's ability to perform all of its missions, including meteorological surveillance and radar target detection and identification, is vital to its cost-effectiveness and success. Unfortunately, the presidentially mandated sale of a vital part of the electromagnetic spectrum means that more austere bandwidth requirements may be needed (PCAST 2012). Even without these new, more severe bandwidth requirements, intermittent radio frequency interference (RFI) has already presented itself at the National Weather Radar Testbed (NWRT) phased array radar. Mitigation strategies for RFI will be needed.

	Fig. 1: RFI can already be observed at the NWRT, here seen as nonmeteorological blips. RFI could become a widespread problem with stricter bandwidth requirements.

 

Several strategies to cope with RFI have been developed . For example, the Vaisala algorithms described in the Vaisala User's Manual are simple, operating on sets of three consecutive pulses. The powers (in decibels) of these three pulses at a particular range are compared according to user-defined constants and the particular algorithm selected. When RFI is detected, the RFI-corrupted data is simply replaced with data from the preceding pulse (Vaisala Oyj 2013).

 

By contrast, the Interference Spike Detection Algorithm, ISDA, is able to vary what data is used in its detection algorithm, gathering data to be considered into a set of “neighbor cells.” Then, ISDA compares the power (in watts, not decibels) of the neighbor cells to the power at the current cell – if this ratio exceeds a user-defined threshold, the data is flagged as RFI. When RFI is detected, the RFI corrupted data is replaced with interpolated data from the neighboring cells. Past tests performed on Gaussian noise with artificial RFI injected have shown that ISDA outperforms the Vaisala algorithms under these conditions [8]. Further evaluation of ISDA is needed.

 

An analysis and evaluation of the characteristics of pulsed interference in the spectrum used by weather radar will be performed using RFI corrupted data available from the NWRT, allowing RFI to be more accurately simulated and its effects on weather radar better understood.  The performance of ISDA will be improved by varying what information is used to comprise the group of neighbor cells and studying the best ways to compute the power level of the neighboring cells. This improved performance will subsequently be compared to the performance of the Vaisala algorithms. As noted above, current approaches to recover the content of RFI-corrupted data use either simplistic averaging methods or just copy data from other pulses. The proposed effort will study new interpolation techniques and their impacts on other algorithms, such as SZ-2. The Sachidananda-Zrnic phase coding technique (SZ2) is a highly effective algorithm that utilizes phase information to mitigate range and velocity ambiguities (Frush et al. 2002). During the research period, other challenges will be tackled as discovered (for example, distinguishing RFI from ground clutter in specific cases as shown in Lake et al. (2014)).

 

References: 

 

Lake, J., M. Yeary, and C. Curtis, 2014: Adaptive radio frequency interference mitigation techniques at the National Weather Radar Testbed:  first results. IEEE Radar Conference,840-845.

 

President's Council of Advisors on Science and Technology (PCAST) Report, 2012: Realizing the Full Potential of Government-Held Spectrum to Spur Economic Growth. 1-192.

 

Vaisala Oyj, 2013: Digital IF Receiver / Doppler Signal Processor User's Manual. Helsinki, Finland, pp. 212-215.