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
Due to large connectivity and conductance, synaptic functions can be physically implemented in the neuromorphic circuits using memristors. Biologically, the firing pattern of pre-synaptic and post-synaptic neurons determines the weight of synapse. For efficient functioning of biological systems, several learning rules have been discovered like rate- and timing-dependent synaptic plasticity, connectivity, etc. In memristors, these necessary rules have also been successfully demonstrated. The STDP (spike timing-dependent plasticity) is an important learning rule, which states that the relative timing of pre- and post-synaptic neuron modulates the synaptic weight. The careful designing of input neuronal signals is required for the implementation of STDP with memristors. Jo et al. [59] first carried out this work with the help of CMOS/memristor circuit.
Resistive Memories for Spike-Based Neuromorphic Circuits
Published in Simon Deleonibus, Emerging Devices for Low-Power and High-Performance Nanosystems, 2018
E. Vianello, O. Bichler, B. De Salvo, L. Perniola
In the brain, memory is stored in synapses, which are the connections between neurons. Synapses not only transmit information from one neuron to another but also modify their strength in response to their local experience. This adaptation is known as synaptic plasticity, and it is considered to be the biological substrate for learning and memory of the brain. Recent developments in the neuroscience community evidence that biological synapses exhibit different kinds of plasticity rules. We will focus on the coimplementation of both short-term synaptic adaptation and long-term STDP [36, 37]. Long-term STDP produces stable modifications of the synapses according to the timing of pre- and postsynaptic spike events. Along with long-term STDP, cortical synapses undergo short-term synaptic adaptation, which is the dynamic modulation of the synaptic strength as a function of input stimulations (presynaptic activity) and it shows retrievable dynamics in short time scales [37].
Nanoionic Switches as Post-CMOS Devices for Neuromorphic Electronics
Published in Krzysztof Iniewski, Tomasz Brozek, Krzysztof Iniewski, Micro- and Nanoelectronics, 2017
Spike-timing-dependent plasticity (STDP) [13] and rate-based plasticity are important memorization mechanisms related to synaptic weight in biological circuits. The STDP mechanism requires precise control of the relative timing of the application of signals to the synapse, which corresponds to the pre- and post synaptic potentials in biological systems. Some ReRAM (or memristor) devices operated by ion migration have exhibited STDP characteristics. For example, a TiOx bilayer ReRAM has functioned as a synapse with STDP operation [14]. A single synaptic ReRAM device based on an InGaZnO (IGZO) material has also been demonstrated [15]. The use of CMOS technology has also been considered to emulate biological synapses with STDP signal processing. In this case, STDP was demonstrated using a hybrid nanocrossbar/CMOS circuit including a mixed Ag and Si active layer [16]. A three-terminal device with a STDP mechanism was also fabricated by integrating a RbAg4I5 ionic conductor layer, an ion-doped conjugated polymer, and a CMOS transistor [17]. In addition, a STDP circuit of a one-transistor–one-resistor structure comprising a HfO2-based ReRAM and a MOSFET has been reported [18]. The realization of unique functions, that is, stimulation rate-based plasticity and the STDP mechanism, appear to hold great potential for demonstrating biological synaptic behavior for future artificial neural networks.
Biological function simulation in neuromorphic devices: from synapse and neuron to behavior
Published in Science and Technology of Advanced Materials, 2023
Hui Chen, Huilin Li, Ting Ma, Shuangshuang Han, Qiuping Zhao
Once the STSP is triggered from the STSP, PPF and PPD also become LTP and LTD, respectively. These synaptic functions, displaying intrinsically frequency dependent, are also considered as the basis of learning and memory. Therein, STDP is regarded as one of the most basic protocols of these and significantly affects the long-term synaptic modification, in which the synaptic weight is regulated by the relative time (∆t) between the presynaptic and postsynaptic spikes [55]. For typical STDP, the synaptic weight (∆w) will increase and display the LTP characteristic if presynaptic spike pulls ahead postsynaptic one. In contrast, if presynaptic spike lags behind postsynaptic one, the synaptic weight (∆w) will decrease and display the LTD characteristic. With the development of neuroscience, four STDP forms have been in the biological system containing asymmetric Hebbian STDP, asymmetric anti-Hebbian STDP, symmetric Hebbian STDP and symmetric anti-Hebbian STDP [56]. For these, the weight updates can be expressed as the following equations:
Neurophysiological and molecular approaches to understanding the mechanisms of learning and memory
Published in Journal of the Royal Society of New Zealand, 2021
Shruthi Sateesh, Wickliffe C. Abraham
STDP describes a special form of Hebbian plasticity whereby the direction of the change (LTP versus LTD) depends on the timing of synaptic transmission and postsynaptic action potential (AP) generation (Caporale and Dan 2008). When evoked by repetitively pairing of pre- and postsynaptic activity with a specific timing (typically, 60–100 times), most synapses exhibit LTP when the postsynaptic action potential occurs up to ∼20 ms after synaptic transmission while the reverse ordering of activity induces LTD (Bi and Poo 1998). This Hebbian form of plasticity also depends on NMDAR activation at the synapses, supported by the APs back propagating along the dendrites (Magee and Johnston 1997). It is considered that STDP may be a paradigm that reflects more physiologically the patterning of activity that generates plasticity in the neural circuitry of behaving animals.