Showing papers on "K-distribution published in 2021"

Generalized extended Weibull power series family of distributions

Abstract: In this study, we introduce a new family of models for lifetime data called generalized extended Weibull power series family of distributions by compounding generalized extended Weibull distributions and power series distributions. The compounding procedure follows the same setup carried out by Adamidis (1998). The proposed family contains all types of combinations between truncated discrete with generalized and non-generalized Weibull distributions. Some existing power series and subclasses of mixed lifetime distributions become special cases of the proposed family, such as the compound class of extended Weibull power series distributions proposed by Silva et al. (2013) and the generalized exponential power series distributions introduced by Mahmoudi and Jafari (2012). Some mathematical properties of the new class are studied, including the cumulative distribution function, density function, survival function, and hazard rate function. The method of maximum likelihood is used for obtaining a general setup for estimating the parameters of any distribution in this class. An expectation-maximization algorithm is introduced for estimating maximum likelihood estimates. Special subclasses and applications for some models in a real dataset are introduced to demonstrate the flexibility and the benefit of this new family.

Experimental Analysis of Statistical Property of Low Frequency Reverberation Envelope in Shallow Water

Abstract: Reverberation is the main background interference of active detection and identification, which is produced with active signal emission. It is traditionally assumed as a combination signal of a large amount of random scatters at the receiving hydrophone which leads to a Gaussian-distributed time-domain signal and a Rayleigh-distributed envelope following the central limit theorem. However, with the advent of high-resolution active sonar systems, the target-like scatter signals namely clutter arising from fluctuation or variety in seafloor or hydrographic would produce the non-Rayleigh reverberation with longer tail and it is a main factor restricting the active sonar long-range detection. Aiming at this phenomenon, the Rayleigh distribution model and two typical non-Rayleigh statistical models-the Weibull distribution and the K distribution model are studied using the low-frequency reverberation test data (signal center frequency at 420Hz) obtained from a typical shallow-water environment in the northern South China Sea. Based on the measured multi-sample reverberation data, the probability density distributions of different signal forms reverberation are calculated respectively. The parameters of different models are estimated by the method of moments (MOM) to perform the theory simulation and curve fitting, and then the correlation between data processing and model simulation results are calculated. Comparative analysis based on the experiments shows that the pulse length of transmitting signal has an important effect on the statistical characteristics of the reverberation in shallow water, which is reflected in the fact that the envelope distribution of short pulse-width reverberation presents stronger non-Rayleigh properties compared with long pulse-width reverberation. In addition, the long-range reverberation present stronger non-Rayleigh properties fitting with the non-Rayleigh models well relative to the short-range reverberation. The effect of frequency form on the reverberation envelope statistical characteristic is relatively insignificant.

On Generating Distributions with the Memoryless Property

Abstract: The exponential and geometric distribution are well-known continuous and discrete family of distributions with the memoryless property, respectively. The memoryless property is emphasized in introd...