- For the components, biological neurons utilize proteins, which are an array of nucleotides configured in the proper order and folded into a functional formation to perform a specific component function. Artificial neurons leverage the device physical properties for building components. - For the devices, biological neurons construct things like signal receptors, transmitters, and storage devices like chemical ligand receptors with biomolecules and protein configurations. Artificial neurons rely significantly on several transistor types for generating devices. • For biological neurons, the layout is based on a typical neuron cell layout and configuration. There is no standardized approach for artificial neurons other than they will each be assembled as an integrated circuit chip with flexible capability within the package. • For receive and transmit connections, a biological neuron utilizes dendrites for receive connections, axons, and the synaptic transmit elements for transmit connections. They take advantage of 3-D space to interconnect. Artificial neurons use crossbar interconnect devices for the receive and transmit connections. • For transmit signal protocols, biological neurons use synaptic signaling transmitters, and for receivers, they use dendrite signal receptors. Artificial neurons use a digital signaling protocol, like TCP/IP, for a transmit and receive protocol. • For memory, biological neurons enable memory by mechanisms in the neuron cell body. Artificial neurons require collocating embedded local memory processing with neural circuits. • For the CPU, the biological neuron has signal conditioning and processing within the nerve body and axons. As signals pass through, the characteristics are modified. Artificial neurons require a microcontroller or a simple processing unit in each neuron module. • For form factor, the biological neuron is contained in a single nerve cell that is modular and designed to work together with many other nerve cells. Artificial neurons place all the various neural circuitry into a neuromorphic chip that contains a discrete number of neurons. • For energy, biological neurons use external chemical energy only when the neuron is activated to participate in a computational function. Artificial neurons utilize electrical energy, often supplied via a continuous power stream to all the neurons in a neuromorphic chip. a. biological neuron realization When looking at the neuron from various points of view, one can see the complementary elements that must work together. There are a variety of biological elements that must be in place to generate a functional biological neuron. With many neuron classifications, a different emphasis is seen from one neuron type to another. b. electronics neuron realization Leveraging what has been done in chip electronics, various electronic neuron elements must be generated to have a functional artificial neuron. Computer architectures have adapted and changed as new ways to improve performance have been discovered. There are several critical artificial neuron differences. With a unique architecture, there are places where it needs to be clarified how to make systems more biologically similar without completely changing the way architectures are devised. 4. Neuron molecular biological composition This section will explore the composition of biological neurons and the primary biomolecules utilized. Unfortunately, much research is still required to clarify the molecular biological composition of neurons further. Active research is exploring this composition. Exploring the molecular architecture of neurons provides insight into the physics of materials and phenomenology utilized to implement neuron capabilities. The physical layer of biological neurons differs entirely from artificial neurons implemented in chip electronics. Examining from an engineering perspective how biological neurons are constructed at the lowest level may suggest new approaches for electronic biomimicry. DNA holds the building instructions for biological components like proteins. Nucleotides are the basic building blocks that DNA is made from. These nucleotides are organic molecules consisting of a nucleoside (a nucleobase and five-carbon sugar) and a phosphate group that forms the structural parts (nucleic acid polymers) of DNA and ribonucleic acid (RNA). The four nucleotides utilized in DNA are adenine (A), cytosine (C), thymine (T), and guanine (G). They are the fundamental building block molecules used in biological life. RNA uses uracil (U) rather than thymine (T). Segments of the DNA code for an organism are transcribed to RNA strands and transferred to building centers (ribosomes) to manufacture proteins. Nucleotides form triplets called codons, and specific codons correspond to amino acids as laid out in the standard codon table. Each of these biochemical layers is part of the physical layer that impacts the physical characteristics of the biological component. This process forms neuron componentry and other biological microsystems and systems. Neuron components are composed of specific proteins optimized for their neuron function. Unfortunately, at the time of this paper’s development, a complete protein characterization of any specific neuron type is incomplete. According to the human protein atlas, the following proteins are characterized parts of the human neuron, as listed below. It is a limited list, and the complete molecular biological layout of the human brain, or portions of the brain, is still being determined. However, there are genes from the various parts of the neuron taken from human and animal scans in representative brain areas (Sjöstedt 2020). • Neuronal dendrite – CAMK2B, ARHGEF33 • Neuronal soma (cell body) – RBFOX3, ELAVL3 • Neuronal nucleus – ZNF3, CBFA2T2 • Neuronal axon – SLC6A4, SPTBN4 • Neuronal synapse – SYNJ2BP, SYP Neuromorphic computing approaches have sought to mimic brain function at higher levels of abstraction. JOHANSEN Human brain function and the creation model 2023 ICC 303
RkJQdWJsaXNoZXIy MTM4ODY=