No abstract

No abstract

for generating and classification tasks. Only packet size and inter-packet time sequences are used as flow features to unify the inputs for the two tasks. The source feature space is scaled and clustered with K-Means to form discrete sequences as model inputs. The model can be trained in two modes: (i) autoregressively, for network traffic generating, where the first token of training sequence represents a flow class, (ii) as a network flow classifier. The evaluation of generated traffic by means of Kolmogorov-Smirnov statistic demonstrated that its quality is on par with the first-order Markov chain, which was trained on each traffic class independently. The metric measured distances between source and generated empirical cumulative distributions of such parameters as packet size, inter-arrival time, throughput and number of packets per flow in directions to and from traffic origin. It was shown that enriching the dataset with external traffic from different domain improves quality of the generated traffic on target classes. The experiment results showed positive influence of generative pre-training on quality of the traffic classification task. In case of using the pre-trained model as a feature extractor for a linear algorithm, the quality was close to Random Forest trained on the raw sequences. When all model parameters are trained, the classifier outperforms the ensemble on average by 4% according to the F1-macro metric.

No abstract

In this paper, we propose a new pre-processing technique for efficient multidimensional wideband parameter estimation. One application is provided by an orthogonal frequency division multiplexing-(OFDM) based joint radar and communication system, which uses SIMO architecture. In this paper, the estimated parameters are given by the range (time delay), the relative velocity, and the direction of arrival (DoA) pairs of the dominant radar targets. Due to the wideband assumption, the received signals on different subcarriers are incoherent and, therefore, cannot fully exploit the frequency diversity of the OFDM waveform. To estimate the parameters jointly and coherently on different subcarriers, we propose an interpolation-based coherent multidimensional parameter estimation framework, where the wideband measurements are transformed into an equivalent narrowband system. Then, narrowband multidimensional parameter estimation algorithms can be applied. In particular, a wideband R-D periodogram is introduced as a benchmark algorithm, and we develop the R-D Wideband Unitary Tensor-ESPRIT algorithm. The simulations show that the proposed coherent parameter estimation method significantly outperforms the direct application of narrowband parameter estimation algorithms to the wideband measurements. If the fractional bandwidth is significant and the SNR is not too low, the estimates provided by the narrowband estimation algorithms can become inconsistent. Moreover, the interpolation order should be chosen according to the SNR regime. In the low SNR regime, interpolation with a lower-order (i.e., linear interpolation) is recommended. For higher SNRs, we propose an interpolation with higher-order polynomials, e.g., fourthorder (cubic splines) or even higher. ESPRIT, interpolation, joint radar and communication, periodogram, wideband OFDM. INDEX TERMS