How the Internet Actually Works: A Deep Dive into Networking Infrastructure

Most people treat the internet as a utility — something that works until it doesn’t. The actual infrastructure underneath is a layered engineering marvel involving physics, economics, geopolitics, and decades of accumulated technical debt. Understanding how it works changes how you think about internet outages, censorship, surveillance, and the surprisingly fragile assumptions that keep the global network functioning. Corn and Herman have covered networking infrastructure across a series of deep-dive episodes that go well past the router setup guide. This is the listener’s guide for anyone who wants to understand the plumbing.

The Foundation: Models and Protocols

• Decoding the Internet is where to start if you want to understand the conceptual architecture of networking. The OSI (Open Systems Interconnection) model divides networking into seven layers — physical, data link, network, transport, session, presentation, application — and the episode explained what each layer does, why the model was invented, and why it doesn’t perfectly describe how the modern internet actually works. The TCP/IP model (which does describe how the internet works) has four layers, not seven, and the episode explained the relationship between the two frameworks and why engineers still teach the OSI model despite its imperfections.

• Bandwidth vs. Speed tackled the most common source of confusion in consumer networking. Bandwidth is the capacity of a connection — how much data can flow per second. Latency is the delay — how long it takes for a signal to travel from source to destination. For most internet applications (video calls, gaming, web browsing), latency matters more than bandwidth above a certain threshold. The episode explained why a 1Gbps connection with 200ms latency feels worse than a 50Mbps connection with 10ms latency for real-time applications, and what the underlying physics of different transmission technologies imply for each.

BGP: The Secret Routing System

• BGP: The Secret Glue covered Border Gateway Protocol — the routing protocol that allows the internet’s roughly 80,000 autonomous systems (ISPs, CDNs, large corporations, universities) to exchange routing information and collectively determine how traffic flows between any two points on earth. The episode explained why BGP was designed the way it is (for resilience, not security), why it works so well for its intended purpose, and why its design choices create vulnerabilities that have never been fully addressed.

• The Invisible Roads followed up with BGP hijacking — the attack where a network announces that it is the correct path to an IP address block it doesn’t actually own, causing traffic intended for the target to be routed through the attacker’s infrastructure. The episode covered real incidents (the 2010 China Telecom incident where 15% of global internet traffic was briefly routed through Chinese infrastructure, the 2018 Amazon Route 53 hijacking used to steal cryptocurrency) and explained why RPKI (Resource Public Key Infrastructure), the main technical defense, has been slow to achieve adoption.

Physical Infrastructure

• Fiber vs. Copper compared the physical layer of wired networking — optical fiber versus copper cable. The episode went beyond “fiber is faster” to examine the actual physics (light vs. electrical signal), the engineering trade-offs (fiber is immune to electromagnetic interference but requires more careful handling), the economics of deployment (digging trenches is the same cost regardless of what you put in them, but fiber lasts longer once installed), and the hybrid architectures that most networks use because fiber and copper each have genuine advantages in different contexts.

• Wireless Fiber examined point-to-point microwave links — the technology that cities and telecoms use to carry large amounts of data through the air between specific fixed points without laying cable. The episode explained where this technology is used (backhaul links from cell towers, connections between data centers, last-mile connections in locations where fiber trenching is impractical), the physics of microwave transmission (frequency vs. range vs. bandwidth trade-offs), and why microwave links are often invisible to the public despite carrying significant fractions of urban internet traffic.

• The SFP+ Revolution covered small form-factor pluggable transceivers — the modular optical components that allow network switches to support different fiber standards without replacing the entire switch. The episode explained why enterprise and prosumer networking is moving toward fiber backbones even in local networks, what SFP+ and SFP28 transceivers actually do, and how to think about building a home or small business network that can scale to multi-gigabit speeds without expensive rework.

Mobile and Wireless Networks

• The Hidden Hierarchy explained Mobile Virtual Network Operators — the companies that sell mobile service without owning any network infrastructure. MVNOs (like Mint Mobile, Google Fi, Ting) lease capacity wholesale from the major carriers (AT&T, T-Mobile, Verizon) and resell it to consumers at lower prices. The episode explained what consumers actually give up when they choose an MVNO: typically, they get lower priority during network congestion, which means their connection degrades first when the tower is overloaded.

• The Physics of Plane Wi-Fi covered the engineering behind in-flight internet — one of the more challenging wireless connectivity problems given that the antenna is moving at 900km/h and needs to maintain contact with satellites or ground stations. The episode compared the older geostationary satellite approach (high latency due to the 36,000km orbit distance) with the newer LEO satellite approach (Starlink and similar, with much lower latency but requiring phased-array antennas that can track fast-moving satellites) and explained why the passenger experience varies so dramatically between airlines and routes.

Specialized Network Infrastructure

• The Invisible Hand took networking into unexpected territory: the communications and control infrastructure of modern rail systems. Rail traffic management is a safety-critical application that requires extraordinarily reliable communication between trains, track-side equipment, and central control systems. The episode covered the European Train Control System (ETCS), the spectrum allocations reserved for rail communications, and the surprisingly complex logistics of managing dozens of trains on a shared track using systems that must fail safely rather than simply fail.

---

The internet feels like infrastructure in the same way that water or electricity do — ambient, reliable, and invisible until something goes wrong. The episodes in this guide make the invisible visible, giving listeners the technical vocabulary to understand what’s actually happening when data moves from one place to another, and why the system is both more robust and more fragile than it appears.