Artificial Networks and Biological Networks

Artificial networks vs biological networks: Two systems, one common architecture

Computer networks and the neural networks of the human brain may seem like very different systems, but they share a similar modular and hierarchical organization.

One is artificially constructed to optimize information transfer, while the other has been selected to process signals and make decisions.

These two networks, computer and neural, exhibit striking similarities in their organization and functioning. The structural and functional correlations between these two complex systems mean that understanding one can shed light on the other.

Physical Layer → Primary Sensory and Motor Areas

The physical layer (layer 1 of the OSI model: Open Systems Interconnection) manages the raw transmission of electrical, optical, or radio signals without interpreting the data. Similarly, the primary sensory and motor areas of the brain process raw signals: sensory areas (such as the somatosensory cortex) receive direct nerve impulses from receptors (skin, eyes, ears), while the motor cortex sends muscle commands in the form of action potentials. These areas act as "low-level" interfaces, just as the physical layer is the interface between the network and the external world (environment).

• Functional correlation: Axons are the Ethernet cables of the brain, synapses are its WiFi – an architecture optimized by millions of years of evolution.

Data Link Layer → Thalamus and Cerebellum

The data link layer (layer 2) ensures error control, MAC addressing, and frame management, guaranteeing reliable transmission between neighboring nodes. In the brain, the thalamus plays a similar role by filtering and routing sensory signals to the correct cortical areas, while the cerebellum optimizes and corrects movements, similar to error correction protocols (CRC) or retransmission (ACK/NACK). Together, they form a precise regulation system, avoiding "collisions" in information processing.

• Functional correlation: The thalamus is the brain's switch, the cerebellum its TCP protocol – together, they ensure an error-free 'neural network'.

Network Layer → Posterior Parietal Cortex and Hippocampus

The network layer (layer 3) manages logical routing (IP addresses) and inter-network transmission. This function is reflected by the posterior parietal cortex, which integrates spatial and sensory information to guide actions, and the hippocampus, essential for navigation and spatial memory (like a cognitive "routing table"). These structures determine "where" and "how" information should flow, just as a router chooses the best path for packets.

• Functional correlation: The posterior parietal cortex is the brain's router, the hippocampus its routing table – together, they optimize our movements in physical space, just as the network layer does for data.

Transport Layer → Limbic System

The transport layer (layer 4, e.g., TCP/UDP) ensures the reliability of communications (flow control, segmentation, reassembly). The limbic system (amygdala, hypothalamus, etc.) fulfills a comparable function by regulating emotional and motivational "connections." For example, the amygdala prioritizes danger signals (like TCP prioritizes ACKs), while the hypothalamus maintains internal balance (homeostasis), analogous to congestion control.

• Functional correlation: The limbic system is the brain's 'TCP/UDP protocol': it ensures the reliable delivery of memories (hippocampus), prioritizes alerts (amygdala), and sends raw vital signals (hypothalamus).

Session Layer → Prefrontal Cortex

The session layer (layer 5) establishes, maintains, and synchronizes dialogues between applications (e.g., authentication). The prefrontal cortex plays this role by managing complex interactions: it initiates and supervises tasks (like opening/closing a session), inhibits distractions (conflict management), and plans sequences of actions (synchronization). It is the "moderator" of cognitive processes, just as the session layer orchestrates network exchanges.

• Functional correlation: The prefrontal cortex is the brain's 'session manager': it authenticates actions, maintains context, and orchestrates internal/external dialogues.

Presentation Layer → Temporal Cortex

The presentation layer (layer 6) translates, encrypts, and formats data so that it is understandable by the application (e.g., JPEG, mp3, SSL). The temporal cortex (especially the associative auditory and visual areas) performs similar work: it interprets sensory stimuli (speech, objects) by giving them meaning (word recognition, face recognition). This layer is the bridge between raw signals and their abstract representation.

• Functional correlation: The temporal cortex is the brain's 'decoding engine': it transforms raw signals into meaningful perceptions, just as the presentation layer converts raw data into usable information.

Application Layer → Multimodal Associative Cortex

Finally, the application layer (layer 7) corresponds to high-level functions (HTTP, FTP, messaging). In the brain, the multimodal associative cortex (such as the parietal-temporal cortex) integrates various information (visual, auditory, mnemonic) to produce complex behaviors (language, reasoning). This is the level where information becomes conscious action or thought, just as an application transforms data into usable services.

• Functional correlation: The associative cortex is the brain's 'web browser': it integrates all information and transforms it into conscious actions and thoughts, just as the application layer transforms data into usable services.

What can we deduce from this?

The brain is the product of millions of years of natural selection to process information efficiently. If our technologies indirectly borrow these principles, we can see a universality in the organizational principles of complex systems.

To manage complexity, any efficient system – biological or artificial – must separate tasks into specialized layers while ensuring their smooth integration, making them universally robust.

Is our technological organization a reflection of our brain?

The striking similarity between the architecture of computer networks (such as the OSI model) and that of the human brain raises a fascinating question: Have we unconsciously copied our own biology to design our technological systems?

This opens up perspectives for better understanding living organisms through computational models, designing more intelligent technologies inspired by the brain, and unifying theories between biology, physics, and computer science.