نشریه پژوهش های نوین در سامانه های دفاع الکترونیکی
سال دوم شماره 2 (تابستان 1402)

PRI modulation recognition and classification is a critical task in ESM and ELINT systems to accurately detect radar threats. However, this practice is challenging in a real environment due to missing pulses and spurious pulses and undesirable antenna scanning effects that lead to noisy PRI modulation sequences. To address this issue in this papper, three methods based on deep convolutional neural network (LeNet5, AlexNet, GooglNet) which have been optimized using extreme learning machine (ELM) have been proposed. In fact, in the first step, a deep convolutional neural network (DCNN) is used as a feature extractor. Then, in the second step extreme learning machine (ELM) is used for real-time recognition and classification of PRI modulation. To evaluate the proposed methods, data corresponding to the real data were designed and simulated, and all the destructive effects on the PRI sequence were considered. The results of simulations on 60,000 images show that the AlexNet-ELM network performs better in most of the evaluation criteria and has achieved a high accuracy of 93%.

The Global Satellite Navigation Systems (GNSS) play a critical role in command and control processes as a tool for positioning and navigation of military forces. The acquisition block is a key element in GNSS systems. This block utilizes correlation to detect and distinguish received signals. Moreover, it provides a rough estimation for code delay and Doppler frequency of each signal. In this paper, we study acquisition methods for weak signals and low-coverage areas. Considering the computational complexity of existing algorithms, we present a novel algorithm for weak signal acquisition. The proposed scheme divides the acquisition integration time into short intervals. By detecting and striping incompetent frequencies in each step, we achieve increased acquisition speed and reduced complexity. Finally, we provide a numerical comparison between the proposed algorithm and the non-coherent integration (NCI) method. The simulation results demonstrate that our algorithm outperforms the NCI method in terms of execution time and computational complexity.

In this article, a hybrid encapsulation scheme based on polar codes is presented, in which the advantages and efficiency of asymmetric and symmetric encryption systems are combined with each other. The proposed hybrid encapsulation scheme includes two parts: a) key encapsulation based on polar codes whose task is to encapsulate, send and de-envelop the secret key required in the data encapsulation scheme based on polar codes, b. ) data encapsulation based on polar codes whose task is to encapsulate, send and deencapsulate data between authorized parties using the secret key exchanged by the key encapsulation scheme. The security analysis of the proposed hybrid encapsulation scheme has also been done by applying the comprehensive key search attack and the information set decoding attack. The results of the security analysis show that the proposed plan is safe against the intended attacks. The efficiency level of the proposed scheme has also been evaluated in terms of decryption failure rate and key length. The results of the analysis also indicate the suitability of the decoding failure rate and the short length of the key.

Object detection is a method in image processing and machine vision that is used to locate samples of objects in images or videos. Object detection algorithms usually use machine learning methods or deep learning methods. The object recognition methods can be single-stage or two-stage. At the same time, these methods are based on the region, based on the transformer and based on pre-training. Also, these methods are based on supervised pre-training or self-monitoring pre-training. Considering that region-based convolutional neural networks are a family of convolutional neural network models that are used to detect objects, and have the characteristic that in the case of classification deficiencies, in addition to extracting feature, more components are also taught in advance, in this article, the method of detecting objects based on the region, according to the mentioned categories, is discussed. Also, the performance criteria used in the detection methods of mentioned objects are introduced.

An efficient algorithm for smoke detection on sequences of color video images obtained from a still camera has been proposed. Our proposed algorithm in this research takes into account the dynamic and static properties of smoke and consists of different basic steps: segmentation of slow moving areas and pixels in the current input frame of the camera based on the adaptive background subtraction algorithm; The combination of slow moving areas with pixels in the form of bubbles and the classification of bubbles from the previous step. The adaptive background subtraction method has been used in the motion detection stage. The classification of bubbles is based on the calculation of optical flow, checking the Weber contrast, and also considers the initial path of smoke diffusion. Real sequences of security cameras have been used for smoke detection using our algorithm. A series of experimental results of video images are presented in this article.Translate

A sensor network is formed by a set of wireless components that are installed in a wide or limited environment and collect environmental data. Typically, these networks are installed in pre-identified locations based on a specific map, but sensors are installed in different locations based on analysis and find each other through radio waves.Considering the vast progress in wireless technologies for the emergence of small and cheap sensors and considering this importance in wireless sensor networks, this paper proposes an optimal solution for locating and tracking unknown moving targets. The development of wireless sensor networks has been developed using military applications including battlefield surveillance.On the other hand, energy consumption and accuracy of positioning and tracking are fundamental issues in sensor networks, so one of the main limitations of sensor networks is energy. Placing batteries with expired batteries is impractical in every sensor node. Therefore, the lifetime of the network depends on the battery life. On the other hand, to increase the accuracy in locating and tracking, scientific and practical solutions are proposed based on the inference of the simulation results. At the end, the proposed model with 3D target location and tracking algorithm will be investigated.