Exploring the Potential of Spiking Neural Networks in UWB NLOS Identification
Organizations: School of Instrument Science and Engineering, the China-UK Research Centre on Intelligent Mobility, State Key Laboratory of Comprehensive PNT Network and Equipment Technology, Southeast University, Nanjing 210096, China
Abstract
Ultrawideband (UWB) nonline-of-sight (NLOS) identification is critical for accurate localization, yet prevailing deep learning solutions impose computational demands incompatible with low-cost edge devices. This letter explores the potential of spiking neural networks (SNNs) for this task and develops a fully unsupervised SNN solution. Against supervised deep learning baselines, our approach achieves comparable accuracy (~82%) with only 4.7K synaptic operations (SOPs) per inference, representing a three-order-of-magnitude improvement in efficiency. These results establish that SNNs can serve as a viable neuromorphic alternative for edge UWB applications.
Spiking Neural Networks
Spiking neural networks process information through discrete, time-dependent neural events, aiming for efficient computation and biologically inspired learning.
See all 376 papers