NeuroWav: Toward Real-Time Waveform Design for VANETs using Neural Networks

Boubin, Jayson; Jones, Aaron M.; Bihl, Trevor J.

doi:10.1109/vnc48660.2019.9062792

Cited by 6 publications

(3 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As previously mentioned, prior work [23,29,30] has shown the potential for neural network waveform design to surpass the ERA and RUWO algorithms in terms of accuracy and latency. That work showed that neural networks could be used as classifiers for indexing a lookup table of pre-computed RUWO waveforms (see Figure 4).…”

Section: Prior Workmentioning

confidence: 97%

“…The high complexity and ensuing long convergence times for these algorithms have prompted researchers to seek alternatives. Prior work has shown that waveform design is one possible area where neural networks could greatly improve inference latency [23]. Low size, weight, and power (low SWaP) neural networks and specialized neuromorphic computing hardware for spiking neural networks have proven to be viable options; however, the goal for real-time latency without a precision drop-off has yet to be achieved.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Lean Neural Networks for Autonomous Radar Waveform Design

Baietto

Boubin

Farr

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

In recent years, neural networks have exploded in popularity, revolutionizing the domains of computer vision, natural language processing, and autonomous systems. This is due to neural networks ability to approximate complex non-linear functions. Despite their effectiveness, they generally require large labeled data sets and considerable processing power for both training and prediction. Some of these bottlenecks have been mitigated by recent increased availability of high-quality data sets, improvements in neural network development software, and greater hardware support. Due to algorithmic bloat, neural network inference times and imprecision make them undesirable for some problems where fast classical algorithm solutions already exist, other classes of algorithms, such as convex optimization, with non-trivial execution times could be reduced using neural solutions. These algorithms could be replaced with light-weight neural networks, benefiting from their high degree of parallelization and high accuracy when properly trained. Previous work has explored how low size, weight, and power (low SWaP) neural networks and neuromorphic computing can be used to improve autonomous radar waveform design techniques that currently rely on convex optimization. Autonomous radar waveform design helps meet the need for interference mitigation caused by an ever-growing number of consumer and commercial technologies which pollute the radio frequency (RF) spectrum. Spectral notching, a radar waveform design technique, augments transmitted radar waveforms to avoid frequencies with excessive interference while maintaining the integrity of the waveform. In this paper, we extend that work, demonstrating that lean neural networks and specialized hardware can improve inference time for waveform design without sacrificing accuracy. Our lean neural solution incorporates problem-specific information into the layer structures and loss functions to decrease network size and improve accuracy. We provide model outcomes implemented on radio frequency system on a chip (RFSoC) hardware that support our simulation results. Our neural network solution decreases inference time on traditional CPU hardware by 1057x and on GPU hardware accelerators by 883x while maintaining 99% cosine similarity.

show abstract

Section: Prior Workmentioning

confidence: 97%

Section: Introductionmentioning

confidence: 99%

Lean Neural Networks for Autonomous Radar Waveform Design

Baietto

Boubin

Farr

et al. 2022

Sensors

Self Cite

View full text Add to dashboard Cite

show abstract

“…With the evolution of intelligent technology, the feature extraction and predictive capabilities of neural networks have offered more solutions for waveform generation techniques [9], [10], [11], [12]. Neural networks can categorize input waveforms into general cases corresponding to output waveforms, and the output waveforms can be pre-calculated as per actual requirements [13], [14], [15]. In the similar domain of waveform design, the approach in [16], utilizing neural networks to generate radar spectrum-notch waveforms, achieves the suppression to jamming.…”

mentioning

confidence: 99%

Jamming Waveform Generation Method Based on Generative Adversarial Network Model

Jiang,

Guan,

Liu

et al. 2024

IEEE Signal Process. Lett.

View full text Add to dashboard Cite

For the issues of unclear jamming effects in traditional intra-pulse jamming waveform generation and the lagging jamming false targets caused by reconnaissance windowing in relay jamming, this paper proposes a jamming waveform generation method based on Generative Adversarial Network (GAN) Models. This model, trained on GAN, generates jamming waveforms driven by jamming effects. Using minimal segments of radar signal headers as input, the jamming waveform generation model reversely predicts and generates complete radar jamming signals based on jamming effects. By reducing the reconnaissance windowing for the jamming side and pre-modulating jamming waveforms based on anticipated jamming effects, this method possesses the capability to achieve transcendental jamming (where the peak value of the jamming false target formed after the pulse compression of jamming waveforms occurs before the peak value of the real target). Experimental results indicate that the radar jamming waveforms generated based on GAN Models can achieve transcendental jamming, increasing the difficulty of radar jamming signal recognition.

show abstract