Distributed Ledger Technology
The definitions, or rather better explanations for distributed ledger technology are certainly numerous. In order to facilitate a common introduction, let´s start with the definition of the Gabler Business Dictionary here:
“Distributed Ledger Technology (DLT) is a special form of electronic data processing and storage. A distributed ledger or "Distributed Account Book" is a decentralized database that allows participants in a network to share read and write authorization. In contrast to a centrally managed database, there is no need for a central instance in this network to make new entries in the database. New records can be added at any time by the participants themselves. A subsequent update process ensures that all participants always have the latest version of the database. A special characteristic of the DLT is the blockchain.”
An attempt at an introductory statement
When one speaks of ledgers, this basically means so-called "books" or databases in general. Basically one can distinguish three types of such databases or ledgers. Centralized, i.e. centralized databases, Decentralized, i.e. decentralized, and Distributed Ledger. As the names already reveal, the distinction is all about the question of how databases are structured. Are they centralized and controlled by one instance or authority, or are they distributed over different parts of a network? So the question is always who can make transactions, control, manage and verify them. A recurring point in this context is the question of trust and how to deal with so-called "trust management". One of the central ideas behind decentralized or distributed networks or databases is to divert control or power away from centralized entities, which then have to be trusted by everyone else.
What distinguishes the three variants, why DLT is not the same as blockchain and what other examples of distributed ledger technology there are, we would like to discuss in more detail below.
1. Centralized vs. Decentralized vs. Distributed Ledger
As we have already read in the introduction, the various types of networks are fundamentally about trust. As far as the network or trust architecture is concerned, there are the three approaches already mentioned.
We have already touched on the structural differences between the approaches. How do they differ in their application? For their publication "A Survey on Trust Computation in the Internet of Things", the scientists Nguyen Binh Truong, Upul Jayasinghe, Tai-Won Um, and Gyu Myoung Lee have summarized some characteristics.
Failure points: In centralized systems there is one point of failure due to the one central control instance, in decentralized systems there is a so-called "finite number of points of failure", in distributed or distributed systems these are infinite.
Maintenance: Centralized systems are easy to maintain because there is only a single point of failure. Decentralized have more, but still limited. Distributed systems are the most difficult to maintain.
Fault tolerance / stability: Centralized systems are considered very unstable. This point can be described with a symbol:
Kill the leader (in a centralized system) and there will be chaos. Kill the leader within a decentralized system and you will have many decentralized systems. Distributed systems are very stable and a single failure does not do much harm.
Scalability / Maximum Population: Centralized systems have low scalability, decentralized systems have moderate scalability, distributed systems have infinite scalability.
Ease of development / creation: Centralized systems can be created very quickly, you take a framework and apply it everywhere. So they have a low complexity. For decentralized systems the complexity is considered moderate, in case of distributed systems more details are needed.
Evolution / Diversity: Since centralized systems follow a single framework, they have little diversity and develop slowly. For decentralized and distributed systems, the evolution is enormous once the basic infrastructure is in place and is therefore considered high.
2. Tangle (IOTA)
The "Tangle" is an alternative distributed ledger technology developed and applied by IOTA. More precisely, the Tangle is a further development of the blockchain technology and is based, in contrast to the classical chain, on so-called Directed Acyclic Graphs (DAG). A central feature of this further development is that it completely dispenses with blocks. On the one hand, this improves scalability, but on the other hand it also greatly limits the possibilities for smart contracts.
If a new transaction is executed, two preceding transactions are always validated and the transaction is then inserted at the appropriate position in the tangle. Since this is a continuous validation, no so-called "mining" takes place. Since this technology, as already indicated, does not use blocks, transactions can be executed immediately. IOTAs Tangle is not the only application of the Directed Acyclic Graph (DAG). Hashgraph also uses DAG, as we will see soon. But with different system architecture.
In fact, the biggest disadvantage of this technology is the aforementioned abandonment of discrete blocks and the resulting limitation of smart contracts. With blockchain, these are executed during the calculation of the blocks. This fact is also the biggest advantage of the technology. Because the immediate execution of transactions and the abandonment of more complex consensus mechanisms naturally increases transaction throughput and scalability enormously.
2.1 Hashgraph
Hashgraph is another special Distributed Ledger and a further development of the blockchain technology based on Directed Acyclic Graphs (DAG). This technology is also characterized by the complete omission of blocks.
Unlike Tangle (IOTA), however, Hashgraph uses a new form of consensus mechanism: The so-called "Gossip Protocol".
With the "Gossip Protocol", in contrast to the blockchain, not every transaction is communicated between all nodes involved in the network at the same time. Instead, each node randomly searches for other nodes in the network and informs them of its state of knowledge. Since this is communicated randomly between different nodes in the network, a common consensus on the transaction history should develop over time. On the medium blog Coinmonks it is summarized as follows:
Using a "Gossip Protocol", the nodes exchange data with other nodes in the community efficiently and quickly. This automatically creates a hash graph data structure with the new "Gossip over Gossip" protocol. This data structure is cryptographically secure and contains the history of communication in a community. If you use this as input, the nodes execute the same consensus algorithm for virtual voting as other nodes. The community reaches a consensus on the order and timestamp without further communication over the Internet. Each event is digitally signed by its creator.
3. Difference between DLT and Blockchain
Blockchain and distributed ledger technology (DLT) are often equated. However, this is not correct. Rather, the blockchain is only one type of DLT. It is certainly the most well known, since it is the technology behind Bitcoin, Enthereum and other crypto currencies.
As the name suggests, the blockchain is characterized by the fact that data and transactions are grouped together in (discrete) blocks. These are then validated within the network and attached to an already validated chain of blocks.
Cryptographic signatures or hash functions are used for this "chaining" of blocks. Because of the use of such cryptographic signatures, it is correct to refer to a blockchain as a distributed ledger. But as we have seen, there are other distributed ledger technologies that work differently than a classic blockchain.
In summary, it can be said that blockchain is a kind of distributed ledger technology. However, not every distributed database is automatically a blockchain. An indication that such a database is not a blockchain is when neither Proof of Work (PoW) nor Proof of Stake (PoS) is used as a consensus mechanism.
4. Conclusion
The fact that the blockchain, as probably the most prominent variant of a distributed ledger technology, is enjoying steadily growing attention and popularity is now for well-known reasons. As we have seen, a central characteristic of DLT and blockchain is the lack of a central authority or instance as an intermediary between different parties. At the same time, this is probably one of the most frequently seen potentials of these technologies.
Furthermore, the increasing popularity of the technology in traditional industries is based not least on the potential to reduce transaction costs by eliminating intermediaries. In addition, transparency and security continue to be potentials that are not only seen in the blockchain but also in DLT in general.
So, as we have seen, these technologies, which are an alternative to centralized systems, were born out of a question of trust. Distributed ledger technologies are an attempt to find an answer to this very fundamental problem of an increasingly technological society.
