China
Africa
Explore chapters and articles related to this topic
Phase-Change Devices and Their Applications
Published in Khurshed Ahmad Shah, Farooq Ahmad Khanday, Nanoscale Electronic Devices and Their Applications, 2020
Khurshed Ahmad Shah, Farooq Ahmad Khanday
With the realization that resistive switching cells can be used as building blocks for non-volatile memory, the research interest in RRAM has increased exponentially in recent years. The pinched hysteresis loop in their current–voltage relationships qualifies them as memristive systems [49], which further extends their use beyond just non-volatile storage elements. Studies in recent years have been conducted to explore the possibility of using RRAM behaviors for neuromorphic and stochastic computing domains. From a system designer’s perspective, in order to explore these possibilities, it is important to be able to model the behavior of RRAM cells. Computing models allow the systems to be simulated in a cheap and efficient manner. It allows newer systems to be tested without the need for actual hardware. Accurate models are important in understanding device behaviors and optimize designs. RRAM technology is still in its early stages; a proper and accurate model will prove to be an efficient tool for studying designs and standardize implementations. A considerable research effort in recent times has been focused towards RRAM modeling. Several models have been given for both bipolar and unipolar RRAMs. The models for bipolar RRAM include Chua’s model [48], Simon’s tunnel barrier model [105], Yakopcic’s model [106], threshold adaptive memristor model (TEAM) [107], voltage threshold adaptive memristor model (VTEAM) [108], Stanford/ASU model [109], physical electro-thermal model [110], Huang’s physical model [111], Bocquet’s bipolar model [112], the Berco–Tseng model [113], and the Gonzalez-Cordero et al. bipolar model [114]. The models for unipolar RRAM include random circuit breaker model [115], filament dissolution model [116], and Window function models [117].
Intelligent predictive stochastic computing for nonlinear differential delay computer virus model
Published in Waves in Random and Complex Media, 2022
Nabeela Anwar, Iftikhar Ahmad, Adiqa Kausar Kiani, Shafaq Naz, Muhammad Shoaib, Muhammad Asif Zahoor Raja
ANNs are one of the most important elements of soft computing since they are used to replicate the functionality of the human nervous system and process data. The ability of ANNs to self-learn allows them to calculate accurate solution to problems that are difficult to handle through traditional analytical methods. It can understand, explore, and train without having to be rebooted, grasp missing information, be easy to retain, have high accuracy, be executed in parallel hardware, as well as respond to non-linear complex models without setting any constraints or implications onto input data. Recently, algorithms based on neural networks and stochastic approaches have received a lot of attention of the research community by virtue of their reliability and effectiveness, in the areas of computer science, engineering and artificial intelligence. such as nonlinear nervous stomach model [28], control autoregressive nonlinear systems [29], model of computer virus spread with countermeasures [30], nanofluid flow models [31–33], dengue fever models [33–35], fluid dynamics [36,37], nonlinear singular fractional system [38], smoking model [39], nonlinear corneal model [40,41], cantilever piezoelectric mechanical nonlinear system [42], plant virus model [43], magnetohydrodynamics [44–48], wire coating [49,50], Falkner-Skan systems [51], coronavirus models [52,53], Van-der Pol Mathieu’s nonlinear oscillatory model [54], These are stimulating factors for utilizing ANNs-LMA to analyze the dynamic behavior of a nonlinear DCV model. This investigation is a novel research trend in the stochastic computing paradigm for determining the approximate solution of a nonlinear DCV model. The following are the noteworthy characteristics of the presented investigations: The dynamics of the nonlinear differential delay computer virus model are investigated by means of the novel intelligent predictive stochastic computing via artificial neural networks Levenberg-Marquardt approach.The formation of the reference dataset for design ANNs-LMA is accomplished by applying the Adams approach to probe the dynamics of the nonlinear DCV model by variation in the recruiting and detaching rate for both old as well as new PCs, bilinear transmission rate amongst healthy versus latently infected PCs, rate of latently infected PCs that break out, the rate for which breaking out PCs obtain antiviral ability, the rate for which antimalware PCs defeat all kind of viruses, delay with respect to time.The merit function is viably composed using the MSE based approximation for the analysis of computed outcomes of designed ANNs-LMA by considering the reference solutions based on four classes of nonlinear delayed computer virus model.Negligible magnitudes of absolute errors, mean square errors, and near to optimal regression metrics endorsed the strength, consistency, stability and reliability of the design ANNs-LMA to solve the nonlinear DCV model.