Have You Heard of Internet Decentralization?

Imagine building a network, connecting customers and keeping it running for years. Then ask a question that rarely appears in the business plan: who controls the record that says these Internet addresses belong with your network?

That question is at the heart of Lu Heng’s argument for a decentralized Internet. The cables, routers and networks are spread across the world. The authority over their names and numbers is much more concentrated. Changing that imbalance is what this argument is about.

Start with the Internet’s address book

An IP address is a number that networks use to send data to a destination. Different networks need different address ranges. A registry records which organization holds which range, helping everyone avoid conflicting claims. In the current system, IANA and five Regional Internet Registries coordinate number resources through an allocation hierarchy.

Think of a registry as an address book. Keeping that book accurate is useful. But keeping a record is different from owning the networks in it, governing their customers or speaking for everyone in the countries it covers.

Domain names raise a related question about control over identifiers, but names and IP addresses work differently. Here, the focus is the number registry: the record underneath the networks themselves.

Why should an ordinary Internet user care?

Picture a local Internet provider that has already paid for equipment, hired engineers and connected homes. If its registry becomes caught in a legal dispute, institutional failure or political conflict, the provider still has to serve those homes. It needs its address records to remain trustworthy and usable while the institution’s problems are resolved.

A registry entry is not an on/off switch for the whole Internet. Routers and network operators decide how traffic moves. But registration and the systems built around it help establish which network may announce an address range. For example, RPKI supplies signed authorizations that operators can use in routing checks. Unstable authority at this layer is therefore a real infrastructure concern, even when cables and routers are working.

The cost does not stop with the operator. Uncertainty over essential resources makes it harder to plan, invest and keep services dependable. Users inherit the consequences of decisions made far from them.

What would decentralization actually change?

Lu Heng proposes taking the registry’s essential job—keeping a consistent record of who holds each number resource—and separating it from the power to control everyone who depends on that record.

In a distributed registry, independent participants would hold and check the same records. A holder would use cryptographic proofs to demonstrate control. Shared rules would prevent conflicting assignments. The aim is that no single administrator could rewrite or erase the accepted record on its own.

For the operator, the practical test is simple: can you change the organization that provides registration services without losing your resources or starting your network again? If leaving is impossible, the administrator still holds the power, however many consultation meetings it runs.

Why a new administrator would not solve it

Moving the same authority to a different company, government or international body leaves the same point of dependence. Lu Heng’s Note 8 explains why he sees a decentralized registry as the only viable path: an institution should be able to fail without putting everyone’s recognition of number resources at its mercy.

This is a proposal for how the Internet’s foundations should evolve. Making it work requires operators to accept the records, a way to resolve conflicting claims and a transition that keeps existing networks connected. A blockchain label alone supplies none of those things.

Why start now?

The longer a network depends on an institution it cannot leave, the more costly that dependence becomes to unwind. Waiting for a crisis means trying to repair the rules while people are relying on them.

The goal is an Internet whose participants can keep cooperating even when an administrator fails. To follow the argument from its starting point, read Lu Heng’s Note 5 on decentralising Internet governance.