To DAG or not to DAG — HGTP : The new, hot DLT on the block for anyone with large data (Part 1)
Brief summary of DAGs
Directed Acyclic Graphs, better known as DAGs, are a part of the mathematics family, in particular graph theory. It is taught to students who are taking any part of the computer sciences.
DAGs simply consist of two parts : nodes (aka vertices) and edges (aka arcs). Nodes can represent anything from statistical data to mathematical functions and edges represent the correlation between these nodes.
A classical DAG usually has no closed loops, but varying forms of DAGs implement closed loops for specific functionality.
How can DAGs be used as data structures?
A data structure in computer science is simply a format of data organisation, management, storage which enables efficient access and be open to changes.
A DAG, just as it was explained briefly, is a graph that can represent numerous things using its main parts, nodes and vertices. So let us see how a DAG would look like as data structure for a simple IoT application within a personal vertical farm.
Let us say that we want to analyse a correlation between the temperature and the state of the plant using a temperature sensor and an RGB camera using a DAG of some sort.
With a DAG each function can be set as a node, and each of the sub-function can be set as an edge.
So in summary of the figure above, we separate the type of sensor data from our personal vertical farm, RGB images and Temperature, and analyse them each respectively at different rates. Once that is done, a summary of the results from sensors should be given to the final function of retrieving the correlation.
Brief summary of the Constellation Network
Birthed in 2017, the Constellation Network was established as software company which specialises in creating infrastructure tools that gets rid of financial and organisational cost of bad data and insecure data pipelines. It was also one of the first, and few, to become a partner at SpaceISAC.
The computing industry is constantly moving at an increasing manner to the edge & the Constellation Network provides the only scalable and secure solution for a more connected world with data integrity. The main objective from Constellation’s integration with Splunk products (such as the drag and drop dashboard) is to provide a product with unmatched level of security within the market.
Brief summary on how the HGTP works
Now to the awaited part, what is the HGTP? & how does it work?
If you have heard of Bitcoin, then you should have heard of Proof of Work (PoW). Which is the consensus protocol/mechanism that is used within Bitcoin’s network of blockchains in order to incentivise the computing of guessing a transaction’s hash.
But since the mechanism is limited in terms of how many transactions can be within a block this has resulted in major ecological problems as the necessary computing power needed to be the first in completing the guess-work of the hash is high in electrical consumption.
This is only one of the 5 Blockchain’s dilemma, as Seth V describes it in his own article “The Unique Value Proposition and Bull Case for Constellation Network”.
P.R.O
Enter the HGTP (HyperGraph Transfer Protocol) : Based on a Directed Acyclic Hypergraphs, Constellation Network’s consensus protocol (aka Proof of Reputable Observation) defines a new outlook on how the integrity of large data can be preserved.
HGTP is a development built on top of the protocol introduced by K.Cong, known as CHECO, which consists of three different sub-protocols, all tied to an extended version of a data chain known as TrustChain.
TrustChain allows every node within the network to directly interact independently with other nodes using their own hash chain, whilst Extended TrustChain introduces a new type of block, defined as the CheckPoint (CP) block. The fundamental function of the CP block is to represent a node at a specific state in the form of a hash pointer. An assortment of these CH blocks creates the state of the whole network.
The CP blocks, similar to the blocks found in both the TrustChain, also have hash pointers (solid arrows), but differ with the lack of references (dotted lines). The sequence counter for CP blocks is shared with the sequence of TX blocks.
Transaction and Validation protocols
The assortment of the TX blocks simply represents the TX protocol, which is a simple request and response protocol. Where at each round of messages exchanged, the nodes create a new set of TX blocks with their respective hash chains.
The detection mechanism of malicious behaviour is not built into the transaction and consensus protocol themselves. Thus the validation consensus is introduced in order to protect the integrity of the overall system. In order to validate a node’s transaction, the node requests the requester for the agreed fragment of the transaction, which is a sub-part of the requester’s Extended TrustChain that starts and ends with the CP blocks which in addition consists of the TX block belonging to the transactions hash pointer.
Part 2 : The HyloChain and where it differs from the ExtendedTrustChain (coming soon…)
As usual, we give the reader a well deserved break, and in the second part of this article we will dive deeper to discuss the actual blockchain developed by the constellation network which has made them almost untouchable in the world of diluted (and somewhat funny) blockchain-based startups.
