Search

To model sequentially observed multivariate nonstationary count data, we propose a switching Poisson-gamma dynamical systems (SPGDS), a dynamic probabilistic network with switching mechanism. Different from previous models, SPGDS assigns its latent variables into mixture of gamma distributed parameters to model complex sequences and describe the nonlinear dynamics, meanwhile, capture various temporal dependencies. Moreover, SPGDS can model all discrete and nonnegative real data by linking them to latent counts. To take advantage of Bayesian nonparametrics in handling the unknown number of mixture components, we integrate Dirichlet process (DP) mixture into SPGDS and develop an infinite switching Poisson-gamma dynamical systems (iSPGDS). For efficient and nonparametric inference, we develop a infinite switching recurrent variational inference network, combined with a scalable hybrid stochastic gradient-MCMC and variational inference method, which is scalable to large scale sequences and fast in out-of-sample prediction. Besides, to handle the time-series categorization task, we further propose an supervised attention iSPGDS (attn-iSPGDS), which combines the representation power of iSPGDS, discriminative power of deep neural networks, and selection power of the attention mechanism under a principled probabilistic framework. Experiments on both unsupervised and supervised tasks demonstrate that the proposed model not only has excellent fitting and prediction performance on complex sequences, but also separates different dynamical patterns within them. [ABSTRACT FROM AUTHOR]

Electronic attack devices can interrupt a radar system by emitting jamming signals. This function typically depends on the support from electronic support measure (ESM) devices, which can capture and identify radar signals in the air. A radar system can avoid being interrupted by confusing adversarial ESM devices. In this article, at the background of colocated multiple-input multiple-output radar, we present an active antijamming method that aims at confusing adversarial ESM devices by masking radar signals with accompanying signals. Masking signals can be transmitted in either the pulsed mode or the periodic continuous-waveform mode. They can be either constant-modulus or with amplitude modulation. The masking signal will cover the radar signal in both time and frequency domains, in order to make adversarial ESM devices find it hard to determine whether an intercepted signal source is a spoofing device or an operating radar. We suppress the impact of masking signals on radar-matched filters, measured by the integrated sidelobe level and peak sidelobe level of their cross-correlation sidelobes, through optimization of masking signals. Masking signals whose chip duration is shorter than that of radar signals are considered for a lower cross-correlation sidelobe level. It is termed as the sequence spread spectrum method, and this method is found critical to obtain good cross-correlation sidelobes, as indicated in numerical results with several kinds of well-designed radar signals. In concerned simulation scenarios, a good masking signal tends to focus its power on frequency bands, where the radar signal power is low. [ABSTRACT FROM AUTHOR]