Blockchains and decentralization
According to Vitalik Buterin, a complete analysis of the (de)centralization for blockchain technology must include political, architectural, and logical statements. In the following rows, we will see how all that can influence the true purpose of blockchains as public shared ledgers. But what is the main purpose of a shared public ledger? In my opinion that will be transparency. This offers anyone the possibility to search a specific transaction that had been placed on the ledger.
Banks are a very good example of political centralization. Being organized in hierarchical structures, led by a consortium, and have a central office. Banks are also regulated structures. This means for every client they have KYC procedures. All client’s data cannot go public due to the regulation present in many countries. Also, transactions between two parties are protected by banks rules. Fortunately, blockchains are structures that, at least in theory, are not politically centralized.
The second approach refers to architectural (de)centralization. The main differences between centralized and decentralized networks are fault tolerance and attack resistance. However, in my opinion, neither of these two can affect transparency. But there is one more important aspect. Often centralized architectures go to political control. And this opens the way for the issues described earlier.
Logical centralization is a characteristic of blockchain technology. This means that a majority of nodes must agree, at one point in time, on the state of the network. This maybe sounds a little bit like centralization. Fortunately, blockchains can use governance to solve network state problems and avoid political centralization. For example, in the event case of a fork (which can be considered somehow a centralized decision), transparency was not affected.
The conclusion is that centralized blockchain can undermine public shared ledgers.
Scaling decentralized networks is not an easy task. On their basis blockchains are not scalable. It is almost impossible to scale architecture with a high level of redundancy. There is a well-known trilemma for blockchain. They should sacrifice security, decentralization, or scalability in their approach.
Decentralization means that at least a majority of nodes must reach a consensus. Scalability means a high level of TPS and future growth. Today many blockchains achieve a high level of security and decentralization. However, scalability remains a tough target. Several different solutions are developed and many are still in process of developing.
The first type of solution is known as on-chain. They require a hard fork of the blockchain. The most known examples are SegWit for Bitcoin and sharding for Ethereum. SegWit is a protocol for Bitcoin which removes signature data for each transaction, making space for more transactions. Sharding breaks blockchain into groups that run parallel processes. Both reduce the time for a block to be produced, increasing in this way scalability for the chain without affecting decentralization.
Nested blockchain represents the first solution from the off-chain package. Build on parent-child architecture the construction consists in fact of many secondary chains linked to the main chain. Child chains do the work and finish it into the main chain. State channels are used to improve transactions capacity and speed. After transactions are complete the final state for them is validated on the blockchain. Sidechains are parallel structures with main chains. They have their own consensus mechanism which is optimized for speed.
All these are Layer-1 and Layer-2 solutions to extend scalability and the same time to maintain security and decentralization. However, further, development is needed, and more tests in a real environment until a silver bullet for scalability problem will be found.