Towards Self-supervised Learning for Multi-function Radar Behavior State Detection and Recognition

<div>The analysis of intercepted multi-function radar (MFR) signals has gained considerable attention in the field of cognitive electronic reconnaissance. With the rapid development of MFR, the switch between different work modes is becoming more flexible, increasing the agility of pulse parameters. Most of the existing approaches for recognizing MFR behaviors heavily depend on prior information, which can hardly be obtained in a non-cooperative way. This study develops a novel hierarchical contrastive self-supervise-based method for segmenting and clustering MFR pulse sequences. First, a convolutional neural network (CNN) with a limited receptive field is trained in a contrastive way to distinguish between pulse descriptor words (PDW) in the original order and the samples created by random permutations to detect the boundary between each radar word and perform segmentation. Afterward, the K-means++ algorithm with cosine distances is established to cluster the segmented PDWs according to the output vectors of the CNN’s last layer for radar words extraction. This segmenting and clustering process continues to go in the extracted radar word sequence, radar phase sequence, and so on, finishing the automatic extraction of MFR behavior states in the MFR hierarchical model. Simulation results show that without using any labeled data, the proposed method can effectively mine distinguishable patterns in the sequentially arriving PDWs and recognize the MFR behavior states under corrupted, overlapped pulse parameters.</div>

Towards Self-supervised Learning for Multi-function Radar Behavior State Detection and Recognition

<div>The analysis of intercepted multi-function radar (MFR) signals has gained considerable attention in the field of cognitive electronic reconnaissance. With the rapid development of MFR, the switch between different work modes is becoming more flexible, increasing the agility of pulse parameters. Most of the existing approaches for recognizing MFR behaviors heavily depend on prior information, which can hardly be obtained in a non-cooperative way. This study develops a novel hierarchical contrastive self-supervise-based method for segmenting and clustering MFR pulse sequences. First, a convolutional neural network (CNN) with a limited receptive field is trained in a contrastive way to distinguish between pulse descriptor words (PDW) in the original order and the samples created by random permutations to detect the boundary between each radar word and perform segmentation. Afterward, the K-means++ algorithm with cosine distances is established to cluster the segmented PDWs according to the output vectors of the CNN’s last layer for radar words extraction. This segmenting and clustering process continues to go in the extracted radar word sequence, radar phase sequence, and so on, finishing the automatic extraction of MFR behavior states in the MFR hierarchical model. Simulation results show that without using any labeled data, the proposed method can effectively mine distinguishable patterns in the sequentially arriving PDWs and recognize the MFR behavior states under corrupted, overlapped pulse parameters.</div>

Hybphaser identifies hybrid evolution in Australian Thelypterid ferns

Hybridisation can lead to reproductive isolation and consequently speciation. It has been proposed to play an important role in fern evolution, but has been difficult to investigate. This study explores the utility of target sequence capture and reference guided read phasing to investigate the role of evolutionary reticulation in ferns using Australian Thelypteridaceae as a model. The bioinformatics workflow HybPhaser was used to assess divergence between alleles, phase sequence reads to references to construct accessions resembling parental haplotpes, and include them in phylogenetic and network analyses to detect hybrids and parentage. This approach identified two novel hybrid lineages in Thelypteridaceae, one occurring between two different genera (Abacopteris and Christella), and provided evidence that reticulation is likely to have played an important role in the diversification of Australian thelypterids. In addition, hybrid phasing successfully reduced conflicting data and improved overall resolution in the Thelypteridaceae phylogeny, highlighting the power of this approach for reconstructing evolutionary history in reticulated lineages.

Changing social inequality from first farmers to early states in Southeast Asia

When the first rice farmers expanded into Southeast Asia from the north about 4,000 y ago, they interacted with hunter-gatherer communities with an ancestry in the region of at least 50 millennia. Rigorously dated prehistoric sites in the upper Mun Valley of Northeast Thailand have revealed a 12-phase sequence beginning with the first farmers followed by the adoption of bronze and then iron metallurgy leading on to the rise of early states. On the basis of the burial rituals involving interment with a wide range of mortuary offerings and associated practices, we identify, by computing the values of the Gini coefficient, at least two periods of intensified social inequality. The first occurred during the initial Bronze Age that, we suggest, reflected restricted elite ownership of exotic valuables within an exchange choke point. The second occurred during the later Iron Age when increased aridity stimulated an agricultural revolution that rapidly led to the first state societies in mainland Southeast Asia.

Development of a Digital Model of a Beta Function Based on Mellin’s Integral Transforms

The implementation of a digital model of the beta function for use in computer algorithms is a time-consuming task. This is due to the complexity of the high-precision representation of its integrand functions, which require a large number of intermediate operations, which entails a large load on the computational power. Purpose: Development of the basic theoretical provisions of the integral Mellin transform in relation to the theory of signal processing against the background of noise and research of their discrete representation. Results: It is shown in the paper that the beta function can be considered as a special case of Mellin’s integral transforms. Based on this statement, a mathematical model of the beta function was developed. Using the properties of parametrically periodic oscillations belonging to the class of trigonometric-logarithmic functions, it was possible to create a digital model for representing the beta function. Practical relevance: Based on the established digital model can be realized a high-speed algorithm for calculating the beta function with a given accuracy. Such algorithms can serve as a basis for creating signal processing programs in order to detect wideband phase-shift keyed signals against a background of noise with an unknown phase sequence. An example of using such algorithms is the search for Wi-Fi bugs.

Voltage Control of Multiphase Cage Induction Generators at a Speed Varying over a Wide Range

The subject of this publication is a method of controlling the DC voltage of a PWM rectifier supplied by a multiphase cage induction generator with the number of stator phases greater than three operating in a wide range of driving speeds. Voltage regulation is performed by changing the frequency and amplitude of the stator voltages with simultaneous switching of the phase sequence of these voltages. The step change of the voltage sequence is made in the designated ranges of the generator speed, which enables the stabilization of the output voltage in a wide range from the minimum speed of about 25% of the rated speed. Such sequence switching changes the number of pole pairs produced by the winding for each supply sequence. The difference compared to multi-speed induction machines is that, in the presented solution, there is only one winding, not a few, which enables good use of the machine’s magnetic core in the same dimensions as for the three-phase machine of a similar power. Steady-state characteristics and dynamic operation were obtained using laboratory measurements of a standalone nine-phase induction generator. The automatic control system maintained the output voltage at the set level, regardless of the generator load and driving power.

Secure OFDM with Peak-to-Average Power Ratio Reduction Using the Spectral Phase of Chaotic Signals

In this paper, a new physical layer security technique is proposed for Orthogonal Frequency Division Multiplexing (OFDM) communication systems. The security is achieved by modifying the OFDM symbols using the phase information of chaos in the frequency spectrum. In addition, this scheme reduces the Peak to Average Power Ratio (PAPR), which is one of the major drawbacks of OFDM. The Selected Mapping (SLM) technique for PAPR reduction is employed to exploit the random characteristics of chaotic sequences. The reduction with this algorithm is shown to be similar to that of other SLM schemes, but it has lower computational complexity and side information does not have to be sent to the receiver. The security of this technique stems from the noise like behavior of chaotic sequences and their dependence on the initial conditions of the chaotic generator (which are used as the key). Even a slight difference in the initial conditions will result in a different phase sequence, which prevents an eavesdropper from recovering the transmitted OFDM symbols.