Avoidance of Time-Varying Radio Frequency Interference With Software-Defined Cognitive Radar

Kirk, Benjamin H.; Narayanan, Ram M.; Gallagher, Kyle A.; Martone, Anthony F.; Sherbondy, Kelly D.

doi:10.1109/taes.2018.2886614

Cited by 71 publications

(16 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [24], the authors developed a compressed sensing scheme to reduce the load of spectrum sensing on digital signal processors and used statistical techniques to estimate the behavior of a Primary User (PU) to guide radar transmissions. In contrast to scheduling user transmissions, SAA methods, which adapt the time and frequency locations of pulsed radar transmissions, have been utilized to reactively avoid interference [13]. However, SAA assumes that the interference behavior is stationary between time steps, and can be ineffective in scenarios when the channel's coherence time is very short.…”

Section: Related Workmentioning

confidence: 99%

“…The former focuses on physical layer performance improvement and includes techniques such as transmit waveform optimization [8], cognitive beamforming [9], and resource management [10]. The latter focuses on spectrum sharing capabilities, and has employed game-theoretic approaches [11], spectrally contained waveform design [12], and rapid-reaction Sense-and-avoid (SAA) techniques [13] to avoid or mitigate interference. However, due to the massive amount of received data radars process as well as the ease of utilizing Channel State Information (CSI) due to co-location of the transmitter and receiver makes the radar control problem a natural candidate for decision-making machine learning algorithms [14].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep Reinforcement Learning Control for Radar Detection and Tracking in Congested Spectral Environments

Thornton,

Kozy,

Buehrer

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

In this paper, dynamic non-cooperative coexistence between a cognitive pulsed radar and a nearby communications system is addressed by applying nonlinear value function approximation via deep reinforcement learning (Deep RL) to develop a policy for optimal radar performance. The radar learns to vary the bandwidth and center frequency of its linear frequency modulated (LFM) waveforms to mitigate mutual interference with other systems and improve target detection performance while also maintaining sufficient utilization of the available frequency bands required for a fine range resolution.We demonstrate that our approach, based on the Deep Q-Learning (DQL) algorithm, enhances important radar metrics, including SINR and bandwidth utilization, more effectively than policy iteration or senseand-avoid (SAA) approaches in a variety of realistic coexistence environments. We also extend the DQL-based approach to incorporate Double Q-learning and a recurrent neural network to form a Double Deep Recurrent Q-Network (DDRQN). We demonstrate the DDRQN results in favorable performance and stability compared to DQL and policy iteration. Finally, we demonstrate the practicality of our proposed approach through a discussion of experiments performed on a software defined radar (SDRadar) prototype system. Our experimental results indicate that the proposed Deep RL approach significantly improves radar detection performance in congested spectral environments when compared to policy iteration and SAA.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep Reinforcement Learning Control for Radar Detection and Tracking in Congested Spectral Environments

Thornton,

Kozy,

Buehrer

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…The bandwidth of high-resolution radar usually ranges from 100 MHz to several GHz, which is easily influenced by time-varying radio frequency interference (TRFI) [1], characterized by instantaneous frequency laws changing in time domain. These interferences can be either narrowband or wideband [2][3][4][5][6][7][8][9][10][11], and their center frequencies hardly follow a pattern.…”

Section: Introductionmentioning

confidence: 99%

Time-Varying RFI Suppression of Dechirping Radar Based on l ₁ Norm Filtering and RAM-Based Frequency Estimation

Wan

Tong

2020

IEEE Access

View full text Add to dashboard Cite

This paper presents a novel framework for suppressing time-varying radio frequency interference (TRFI) in wideband dechirping radar. The TRFI challenges the wideband radar because of 1) changing in wide spectra range; 2) dynamic nature and unpredictability; 3) significant noise power in limited band. In this paper, an essential phenomenon of real measured data about TRFI in dechirping process is analyzed, where it is smashed into several segments because of the narrowband filtering. Based on that, a 2-step framework combining the interference filtering and frequency estimation is proposed. First, since the different sparsity of signal of interference, a simple yet very effective regional filtering method based on l1 norm of STFT is applied, which transforms original problem into a sparse signal recovery problem. Second, for 1-D signal, the reweighted atomic norm minimization (RAM) algorithm is applied to estimate the signal from defect filtered data; for 2-D signal block, a weighting strategy is proposed to promote the accuracy of original 2-D atomic norm minimization method. This method can be implemented as an interference suppression stage for wideband radar with dechirping processing. Experiment of real measured signal in TRFI environment and simulation of defect signal block illustrate its effectiveness. INDEX TERMS sparse recovery; reweighted atomic norm minimization (RAM); radio frequency interference (RFI); dechirping.

show abstract

“…One proposed method for adaptive radar waveform selection is Sense-and-Avoid (SAA). SAA identifies occupied portions of the spectrum and directs radar transmissions to the largest contiguous open bandwidth [5]. However, SAA does not learn to recognize patterns, which can be used to avoid Radio Frequency Interference (RFI) in a strategic manner, where some RFI is avoided and interference with other signals may be allowed depending on application-specific preferences.…”

Section: Introductionmentioning

confidence: 99%

Experimental Analysis of Reinforcement Learning Techniques for Spectrum Sharing Radar

Thornton,

Buehrer,

Martone

et al. 2020

Preprint

Self Cite

View full text Add to dashboard Cite

In this work, we first describe a framework for the application of Reinforcement Learning (RL) control to a radar system that operates in a congested spectral setting. We then compare the utility of several RL algorithms through a discussion of experiments performed on Commercial off-the-shelf (COTS) hardware. Each RL technique is evaluated in terms of convergence, radar detection performance achieved in a congested spectral environment, and the ability to share 100MHz spectrum with an uncooperative communications system. We examine policy iteration, which solves an environment posed as a Markov Decision Process (MDP) by directly solving for a stochastic mapping between environmental states and radar waveforms, as well as Deep RL techniques, which utilize a form of Q-Learning to approximate a parameterized function that is used by the radar to select optimal actions. We show that RL techniques are beneficial over a Sense-and-Avoid (SAA) scheme and discuss the conditions under which each approach is most effective.

show abstract

Avoidance of Time-Varying Radio Frequency Interference With Software-Defined Cognitive Radar

Cited by 71 publications

References 31 publications

Deep Reinforcement Learning Control for Radar Detection and Tracking in Congested Spectral Environments

Deep Reinforcement Learning Control for Radar Detection and Tracking in Congested Spectral Environments

Time-Varying RFI Suppression of Dechirping Radar Based on l ₁ Norm Filtering and RAM-Based Frequency Estimation

Experimental Analysis of Reinforcement Learning Techniques for Spectrum Sharing Radar

Contact Info

Product

Resources

About

Avoidance of Time-Varying Radio Frequency Interference With Software-Defined Cognitive Radar

Cited by 71 publications

References 31 publications

Deep Reinforcement Learning Control for Radar Detection and Tracking in Congested Spectral Environments

Deep Reinforcement Learning Control for Radar Detection and Tracking in Congested Spectral Environments

Time-Varying RFI Suppression of Dechirping Radar Based on l 1 Norm Filtering and RAM-Based Frequency Estimation

Experimental Analysis of Reinforcement Learning Techniques for Spectrum Sharing Radar

Contact Info

Product

Resources

About

Time-Varying RFI Suppression of Dechirping Radar Based on l ₁ Norm Filtering and RAM-Based Frequency Estimation