How does NGK’s algorithmic trading and information work?

3 min readApr 22, 2021

Blockchain is commonly used in the data structure of “block + chain”. In simple terms, the “block” refers to the overall package of transactions by the bookkeeper within a unit time, and the new block is connected to the previous block. Together, they form a blockchain.

Although this chain structure has the advantages of being relatively simple and quick confirmation, it also has inherited stubborn issues. First, it is difficult to improve the throughput, so the efficiency has always been low, and there can only be a single chain in the entire network, which cannot be executed concurrently. If the block is too small, it will inevitably cause a significant delay in the transaction, and if the block is too large, it will cause data expansion, where ordinary node users will overwhelm and suffer from DDos attacks and other problems. In addition, there are also potential problems such as energy consumption and safety issues.


DAG (i.e. Directed Acyclic Graph) was originally a commonly used data structure in the computer field. Due to its good multithreading characteristic, it can upgrade the blockchain from single linear writing to three-dimensional multi-point parallel work. This has now become an increasingly important development direction in the blockchain world.

In addition, NGK also considers the trading and side information algorithm. A typical scenario of the double-spending problem is a sidechain attack. Attackers often send multiple transactions within a short period of time to rapidly grow the sidechain, and after they approve each other, a series of deceptive sidechains may lead to double spending.

To prevent this from happening, the NGK will do a set join between two approved transactions, and the judgment of this right is determined by the side information, whereas the side information is confirmed by time. By using side information to readjust the transaction, the attack effect is attenuated, and the smooth operation of the network is maintained.


When the static graph algorithm updates the node rights, it needs to calculate the weight of each node from the beginning of the initial node. This calculation is very complex. And if we cache the statically calculated information and update the cached information only when a new tip is added, the calculation complexity will be greatly reduced.

At present, the NGK’s solutions have gradually moved from research to implementation. Many new DAG solutions are being proposed in the industry. From a technical point of view, the NGK may have an impact on existing blockchain projects and transform the blockchain design thinking to a certain extent.

NGK has the advantages of fast speed and high throughput. As more and more projects participate in the ecological development of NGK, the DAG is a very promising mechanism in the long run. The NGK blockchain creatively designed a DAG-based throughput cache layer by expanding the multi-layer protocol, and made breakthroughs from theory and reality, and took the lead in the blockchain dispute.