Search

The paper investigates statistical distribution testing-based detection methods in an intermittent signal detection scenario. The relevance of the research is driven by 5G networks based on packet transmission, incorporating the concept of cognitive radio and adapting spectrum detection methods from Long Term Evolution (LTE) Licensed-Assisted Access (LAA). The conducted study refers to the recently proposed methods based on testing goodness-of-fit (GoF) of statistical distributions, which are compared with a conventional energy detector. The authors examine the applicability of well-known GoF methods in intermittent transmission, as they require reconsideration in 5G communication systems, and investigate the behavior of the innovative energy-based GoF. The experiments are carried out for different transmitter activity factors, i.e., channel occupancy and signal-to-noise ratio (SNR), demonstrating the superiority of the GoF-based methods in general and particularly the invented GoF test over other energy-based detectors for discontinuous signals detection.

Detection of unknown and weak signals has become a study of particular importance as a consequence of the ongoing development in digital sensing technologies. Among well-known methods, the one most commonly used is energy detection. In this paper the efficiency of the conventional energy detection is compared with the efficiency of the recently proposed, so-called stable field detection. The comparison concerns pseudorandom, additive, white, Gaussian noise (AWGN) produced in the simulation experiment and real radiofrequency (RF) background noise measured in industrial, scientific and medical (ISM) band transmission. The background noise samples were acquired from a laboratory room using a software-defined radio platform NI USRP-2900. The reasonability of the proposed detection approach is confirmed by the results achieved in the modeled system.

This paper presents a novel algorithm for preprocessing of pulses coming from GNSS and signals produced locally by a rubidium oscillator to reduce synchronization time without changing the main algorithm of the control loop. Moving average and Kalman filtration were used as the preprocessing methods. The experimental results for different rubidium oscillators are presented.

Today, cryptographic security depends primarily on having strong keys and keeping them secret. The keys should be produced by a reliable and robust to external manipulations generators of random numbers. To hamper different attacks, the generators should be implemented in the same chip as a cryptographic system using random numbers. It forces a designer to create a random number generator purely digitally. Unfortunately, the obtained sequences are biased and do not pass many statistical tests. Therefore an output of the random number generator has to be subjected to a transformation called post-processing. In this paper the hash function SHA-256 as post-processing of bits produced by a combined random bit generator using jitter observed in ring oscillators (ROs) is proposed. All components – the random number generator and the SHA-256, are implemented in a single Field Programmable Gate Array (FPGA). We expect that the proposed solution, implemented in the same FPGA together with a cryptographic system, is more attack-resistant owing to many sources of randomness with significantly different nominal frequencies.

One of the sources of randomness for a random bit generator (RBG) is jitter present in rectangular signals produced by ring oscillators (ROs). This paper presents a novel approach for the design of delays used in these oscillators. We suggest using delay elements made on carry4 primitives instead of series of inverters or latches considered in the literature. It enables the construction of many high frequency ring oscillators with different nominal frequencies in the same field programmable gate array (FPGA). To assess the unpredictability of bits produced by RO-based RBG, the restarts mechanism, proposed in earlier papers, was used. The output sequences pass all NIST 800-22 statistical tests for smaller number of ring oscillators than the constructions described in the literature. Due to the number of ROs with different nominal frequencies and the method of construction of carry4 primitives, it is expected that the proposed RBG is more robust to cryptographic attacks than RBGs using inverters or latches as delay element.

This paper presents a novel approach for the design of ring oscillators in a combined true random number generator (TRNG). We propose using delay elements made on carry4 primitives instead of series of inverters or latches in ring oscillators, which enables the construction of many high frequency ring oscillators with different nominal frequencies in the same field programmable gate array (FPGA). The output sequences produced by this type of TRNG pass all NIST 800-22 statistical tests for smaller numbers of ring oscillators than the constructions described in the literature. To assess the unpre-dictability of the output bits, the restarts mechanism, proposed in earlier papers, was used. Due to the number of ROs with different nominal frequencies and the method of construction of carry4 primitives, it is expected that the proposed TRNG is less sensitive to frequency injection attacks.

In this paper, we propose a new secure pseudorandom number generator that can be integrated with an arbitrary cryptographic system in the same field programmable gate array (FPGA). The described design uses the Improved Multiplicative Congruential Generators (IMCGs) defined recently. The specific properties of FPGAs and the IMCGs enable the generation of random numbers in only three cycles of the system clock. The output bit rate for eight IMCGs is 2.604 Gbit/s. It can be significantly increased using more IMCGs. The proposed generator can also produce secure bits as a specialized chip, which further increases the output bit rate.

This paper deals with the problem of generation of 1-pps timing signal at the output of computer interface controlled by Network Time Protocol. The methods of synchronization of internal computer clock and transmission of the timing signal to the serial interface are described. The results of the measurement of obtained signals are presented.

In this paper, the measuring system SP-4000, which allows for the analysis of timing signals in telecommunication networks, has been described briefly. The solutions that make possible the cooperation between time error meters of this system and the real-time assessment of the parameters have been described. Because of the limitations of computer technology, the user can choose real-time computation for only a single channel. The current version does not allow simultaneous traditional computations and real-time computations. We expect that in the future such computations will be possible for channels that are not chosen for real-time assessment of ADEV, TDEV, and MTIE. We also plan to increase the number of channels capable of real-time computations.

Summarizing all the series of 24-hours and weekly tests, the authors observed some common tendencies. Neither the server stability nor the speed of generating responses were significantly affected by the fact whether a given test day was a working day or not. It was not confirmed that the server with the shortest time for generating responses is at the same time the most stable one. The influence of the stratum level of the server cannot be noticed, and sometimes a server with better stratum shows worse results. Statistically, the response times of servers are within the range of 100μs.

This paper presents a new method for improving the statistical properties of sequences generated by the sawtooth chaotic map realized in a finite-state machine in the modular arithmetic. It is shown that sequences obtained with the use of the proposed method pass all tests from the standard NIST 800-22 statistical test suite v.1.8.

This article introduces a new online platform for testing binary random number generators. The growing share of low-complex devices in IoT networks increases the demand for basic authorization and authentication tools, the critical block of which is a secure random number generator. Communication devices, therefore, require designers to carry out time-consuming tests and acquire specialist knowledge of statistical testing in evaluation of their results. To meet the current requirements, we have created a test platform to assess the quality of random strings produced by the generator. The presented solution, based on the proprietary evaluation metric, provides feedback on the properties of the uploaded random sequences. Clear interface provides ease of use and by machine learning in the platform’s backend, along with the increase of processed data, the improved quality of the interpretation delivered by the system is ensured. The operation of the platform has been confirmed experimentally, based on the analysis of hardware generators producing random strings with known properties.