Blockchain - Decentralization & Scalability

In his paper, “Bitcoin: A Peer-to-Peer Electronic Cash System”, Satoshi Nakamoto provides an alternative to cash in the form of digital currency while solving the problem of double spending and eradicating the need for a trusted third party such as a bank. His solution is built on a peer-to-peer network that timestamps transactions while forming a chain secured by hash-based proof of work that is nearly impossible to alter – The Blockchain.

Information on the blockchain is digitized and eliminates manual documentation. Each block in the same transaction is recorded and linked chronologically to provide tractability, cryptographically secured to prevent tampering of existing records on the network. Distributed Ledgers enable identical copies of all information or blocks to be shared across all participants on the network. In the event of failure of one or more nodes, the functioning participants continue to perform validation without the need of a central authority. Furthermore, a transaction can only be executed if all or most of the nodes on the network agree to its validity, a consensus-based model that is both complex and requires an immense amount of computing power to achieve.

In the instance of Bitcoin which is the largest decentralized or public network, it is a truly democratized system where anyone with the required computing power can access the shared ledger and participate in executing the consensus algorithm – proof of work. This is in line with the technology’s original purpose where massive amounts of resources are required to maliciously attack the system making it economically nonviable and operationally impractical. Also, the system maintains a true peer-to-peer status with low transaction fees and does not require the need for a central, trusted intermediary.

Centralized networks on the other hand can be less secure and transparent. For instance, IBM’s Hyperledger Fabric is a permissioned enterprise blockchain network that allows for customization by the deploying organization that may own and control a small number of participants making the network susceptible to malicious attacks. Since the information is not publicly available, outside parties are unable to verify the validity of the data. Most importantly, these networks perform transaction execution functions for organizations that maintain their status as a trusted intermediary, thereby undermining the purpose of the technology.

Although decentralization is in line with blockchain’s original purpose, it comes at a cost. To keep the Bitcoin engine churning, an enormous amount of computing power is required, which in turn produces massive energy expenditures. Some estimate the power consumption to be as much as that of a small country. The other major issue is scalability. In an open system with no barriers to entry to join the network, transaction validation time increases as participants are added. For Bitcoin, the block size and creation times are also limiting factors. Bitcoin can only process 7 transactions per second as opposed to credit cards that are clearing approximately 2000 per second.

There are several concepts in play to tackle the scalability problem. Proof of stake is an alternate to proof of work where only a select few participants are chosen for the validation and block creation activity instead of all the nodes in the network as in the proof of work mechanism. These participants are selected according to the size of their collateral and are paid a transaction fee. They are also incentivized to remain honest and keep the network secure as their “stake” in the system is large. Ethereum is now moving to this method of consensus validation. Other mechanisms such as a layer above the main blockchain can also be developed. This layer would perform the heavy lifting calculations outside the blockchain and thereby increase transaction speed.

There are several novel concepts being tested but the bottom line is simple. For blockchain applications, especially cryptocurrencies to be effective, the problem of scalability must be solved.