A Primer on DAG Technology

“Bitcoin”, “blockchain”, “crypto”, or the more precise “Distributed Ledger Technology” (DLT) are all keywords used by the public and people in the tech industry to refer to the new phenomenon of using cryptographic principles and distributed computing to solve use cases that were previously intractable.  But the fact that so many of these terms float around says a lot about how new the industry is. We don’t use the vague or incorrect terms when talking about other tech businesses; a social network, a productivity app, and an online marketplace are different platforms and are referred to as such – while all crypto projects are usually lumped into the category of “blockchain technology”.

This trend oversimplifies the industry to the extreme.  Every investor or anyone with an interest in the domain needs to know that in a couple of years we won’t just be calling it “blockchain” – crypto technology is branching out into drastically different use-case areas that share only a passing resemblance. As of now, there is one principal divergence in the industry: high throughput vs high precision platforms.

High precision protocols are those like Bitcoin or Monero: systems designed to ensure integrity of a network that allows for very secure, reliable (and in the case of Monero, highly anonymous) transactions.  By paying a transaction fee to a blockchain-based network,  transactions are impossible to fake and the network is prohibitively hard to manipulate. Large financial transactions, voting systems, or property records all require this level of double reliability.

But the transaction costs of this kind of network are prohibitive and unnecessary for another kind of use case: which we can call “high throughput”. These systems aim at adding a layer of crypto integrity and decentralisation to everyday business usecases that usually require data transfers that are orders of magnitude more intensive than the previous examples, either in terms of data volume (measured in GB per day) or data frequency (measured in transactions per second). Think internet of things (IoT) applications, collaborative research platforms, or decentralised cloud computing platforms.  Running these high throughput applications on a high precision platform like the Bitcoin blockchain would not work. It is also unnecessary; because these business cases involve very regular database updates on the blockchain (imagine a sensor tracking your location), in most cases one outlier error in data transferral can be easily identified as inaccurate and then safely ignored (imagine the location sensor listing your whereabouts as 300 miles away for a second, then reverting to normal); those use cases can manage with the odd irregularity since they are constantly being refreshed.

And right now, many of the most exciting projects are working on the high throughput category.  Within this category, a paradigm called Directed Acyclic Graphs (DAGs) or “tangles” is getting all the attention. For platforms that require high throughput, DAGs are a game changer.

The most visible projects using tangles are those related to IoT. IOTA and IoT Chain are closely competing in this regard and are driving the state of research forward for DAG knowledge. Another project called CyberVein is offering an even more advanced capability; it used a leaner and more efficient consensus mechanism known as Proof of Contribution (PoC) to make transfer of even larger datasets on the DAG possible. There is also Byteball which makes DAGs useful for even faster crypto payments worldwide. Aside from these leaders, there are new projects constantly finding use cases for the inherent power of the DAG concept.

Dawn of the DAG

To explain DAGs, it is useful to break the term into its constituent parts.  The Graph simply refers to a network of nodes and edges (the connections between nodes) which have specific attributes. A social network of your friends and family could be expressed as a graph in which the edges are complex and bidirectional (i.e. multiple people can have complex relationships with each other). The “Acyclic” term refers to the nature of how connections (transactions in the case of cryptocurrencies) cannot route back to the originating node no matter what – in this way it is like a tree of nodes rather than a loop of nodes. Related to this is how it is “Directed”, meaning that the edges (read: connections) between nodes go in one direction only. A useful way to get an idea of how this works is with a diagram (from Wikimedia Commons):

The important result of this configuration is that the system has built-in checks to prevent manipulation of the consensus on the network without having to ensure that every node has a full copy of the ledger at all times: different parts of the network can work more independently than a network on a blockchain, while the record of transactions is still extremely difficult to defraud.

Use cases

While the differences between different types of crypto platforms might seem somewhat involved and obscure, thy have fundamental impacts on the types of usecases fit for each platform. As mentioned earlier, there are some distinct areas that stand out in terms of the need for DAGs. Firstly, the requirements placed on a blockchain by IoT systems involving thousands of data points linked to sensors in homes, cars, and almost anywhere else in the real world would rapidly overwhelm any blockchain based system. Secondly there are usecases that require intensive collaboration, like scientific research, that would also become too involved for a block based paradigm. Finally there is also the issue of large data sets; fees per mB transferred on the likes of the Bitcoin blockchain would be so high as to preclude the level of data transfer commonly seen in regular business applications, let alone the amounts involved with Big Data.

A split in the crypto community?

Unlike some of the more contentious debates in the cryptocurrency domain recently, there will be no “hard fork” (if you will) of the crypto community between blockchains and DAGs. They both serve different use cases as outlined above. Furthermore, this poses no problems for the coming paradigm of blockchain-to-blockchain transactions as envisaged with cross-chain bridges and atomic swaps since several projects (such as ARK and Hcash) are working to enable transactions from, to and amongst blockchains and DAGs. In the near future, high throughput and high precision platforms will be able to seamlessly interact.

To Top

Pin It on Pinterest

Share This