China
Africa
Explore chapters and articles related to this topic
Predistortion Algorithms and Applications
Published in Jingchang Nan, Mingming Gao, Nonlinear Modeling Analysis and Predistortion Algorithm Research of Radio Frequency Power Amplifiers, 2021
Self-adaptation means that the adaptive algorithm automatically tunes the weight coefficient ωpu(t) of the neural network predistorter based on the magnitude of the error E2(t) in Figure 10.44, thereby minimizing the mean square error. The most widely used adaptive algorithm is the descent algorithm, which has two implementations, that is, adaptive gradient algorithm and adaptive Gauss-Newton algorithm. The former includes the LMS algorithm and various improved types, and the latter includes the RLS algorithm and various improved algorithms. LMS is simpler in structure and less in computational complexity compared with RLS, so LMS is used as the adaptive algorithm in the improved structure.
M-Sweeps multi-target analysis of new category of adaptive schemes for detecting χ2-fluctuating targets
Published in Journal of Information and Telecommunication, 2020
The detection capability is one of the most significant factors in the radar system. The main concept of this process is to take a decision as to whether the target of interest is present or not. Modern radar systems perform this process automatically through the setting of a fixed threshold, based on the interference power level, in such a way that the false alarm rate is kept constant. This fixed threshold processor can be used if the background is spatially and temporarily stationary. However, the presence of different sources of unwanted clutter and interference, make the noise level to be fluctuating with time and space. In this situation, a variable threshold is a significant proposal to maintain the false alarm rate unchanged in the face of these background variations. This indicates that the threshold level must be varied up and down, according to the clutter fluctuations, for the rate of false alarm to be controllable. This is the basic rule of constant false alarm rate (CFAR) detection. In other words, this type of signal detection has an adaptive mechanism. In this regard, an adaptive algorithm can be defined as an algorithm that varies its behaviour at the time it is run, based on a pre-defined mechanism and available information. In the CFAR situation, such information is related to the environment in which the radar operates (El Mashade, 2008; El Mashade, November 2006; Ivković et al. 2013; Laroussi and Barkat 2006; Wang, 2013).
Data-driven decision making: new opportunities for DSS in data stream contexts
Published in Journal of Decision Systems, 2022
Nuria Mollá, Ciara Heavin, Alejandro Rabasa
Adaptive methods are similar to incremental methods but with a slight difference in terms of the user interference on the system. With incremental algorithms, the user needs to define several parameters that will affect the result of the system, such as a window size or selection criteria, depending on the algorithm. While adaptive algorithms do not use parameters, they are able to estimate the best fitting conditions in each moment with no user interaction. Thus, an adaptive algorithm can run in a system with optimised features for an undefined amount of time, being able to integrate new knowledge and adapt to changes in potentially infinite data.
Goal-Oriented Regional Angular Adaptive Algorithm for the SN Equations
Published in Nuclear Science and Engineering, 2018
Bin Zhang, Liang Zhang, Cong Liu, Yixue Chen
The adaptive algorithm is a trade-off between the reduced number of unknowns needed to produce a given accuracy and the overhead of the adaptive error tests. In some shielding transport problems, one is often interested in acquiring an accurate detector response and does not necessarily require an accurate solution across the whole domain. The flux solution may also vary by orders of magnitude from the source region to the region of interest. The absolute error in regions of low values may be negligible, whereas the relative error in these same regions may still be unacceptably high.