Payment system: a liquidity saving mechanism in a distributed ledger environment

This article focuses on the distributed ledger technology and the operation effect of the Stella project. Distributed Ledger Technology (DLT) is a set of tools for recording data, such as asset holdings or financial transactions, allowing a computer network to verify and store updates without a single central management system. In December 2016, the Bank of Japan (BOJ) and the European Central Bank (ECB) announced the launch of a joint research project called “Stella” to evaluate the applicability of DLT and the applicability of the solution in the field of financial market infrastructure . This report is the first result of the cooperation between the two parties. The Stella project contributes to the ongoing extensive discussion on the availability of DLT in financial market infrastructure. This joint research aims to promote safer, faster and more economical Financial transactions. The project has an exploratory nature within the stated limited scope, and currently only focuses on actual testing, while the areas of cost efficiency, market integration and supervision are left for future research. The Institute of Financial Technology, Renmin University of China compiled the core content of the paper.

作者 | European Central Bank & Bank of Japan

The main results of the conjoint analysis

This report details the main results of the joint analysis, which are summarized as follows:

1. DLT-based solutions can meet the performance requirements of real-time full payment systems (RTGS):

The analysis found that in the Eurozone and Japan, DLT applications are comparable to the amount of requests flowing to the RTGS system. Taking into account the average traffic of these two centralized payment systems (about 10 to 70 throughput per second (RPS)) and the average transaction processing time is less than one second. However, when the RPS increased to 250, the analysis confirmed that the trade-off between flow and performance cannot be ignored. More broadly, the test proved the feasibility of implementing standard LSM processing logic (queuing and bilateral offset) in a DLT environment and in a DLT environment.

2. The performance of DLT is affected by the size of the network and the distance between nodes:

The analysis confirmed the well-known trade-off between network size and performance. Increasing the number of nodes leads to an increase in payment execution time. In addition, the impact of the distance between nodes on performance depends on the network configuration: as long as the minimum number of nodes (quorum) required to reach consensus is close enough, the dispersion in other parts of the network has a limited impact on delay. Nevertheless, nodes on the periphery of the network may still be inconsistent with the quorum. If the quorum is sufficiently dispersed, the impact on the delay will be greater.

3. DLT solutions have the potential to strengthen resilience and reliability:

Although the analysis is not exhaustive, it shows that the DLT network has the ability to withstand the test of problems, such as verifying node failures and incorrect data formats. Regarding node failures, it has been observed that as long as the number of nodes required by the consensus algorithm is operational, the availability of the system will not be affected. The test also confirmed that no matter the length of the downtime, the verification node can be recovered.

However, it should also be considered that the selected DLT setup includes a single certification authority. This is a single point of failure that can undermine the advantages of distributed verification. In addition, tests using incorrect data formats show that the system can detect incorrect data formats without affecting overall performance.

Research results on efficiency

1. The impact of network size on efficiency:

We have conducted tests to verify the impact of increased verification nodes on performance, and the performance of simple smart contracts (that is, payment transfers without LSM) and LSM smart contracts (that is, payment transfers with LSM) In case, the test was conducted.

The results based on a simple smart contract are as follows:

Payment system: a liquidity saving mechanism in a distributed ledger environment

figure 1

The results based on the LSM smart contract are as follows:

Payment system: a liquidity saving mechanism in a distributed ledger environment

figure 2

2. The effect of the distance between nodes on efficiency:

We conducted tests to evaluate the performance of the verification nodes in the following situations (that is, causing the communication between them to take longer), and explored two cases, each based on four nodes.

Payment system: a liquidity saving mechanism in a distributed ledger environment

image 3

Payment system: a liquidity saving mechanism in a distributed ledger environment

Figure 4

The results obtained in the centralized situation (see Figures 3 and 4) show that the closer the nodes are, the less the performance is affected; the delay measured between the nodes is equivalent to the baseline scheme without delay. However, in this case, the nodes separated from other nodes showed large delays (112% higher than the baseline scheme) or showed signs of not being able to catch up with other nodes without participating. In the case of decentralization, it showed more High latency. This is due to the long distance between each group of nodes, which increases the delay by 67% compared with the baseline scheme. The results from these two scenarios show that, on average, when nodes that need to interact with each other, the formation of consensus will be faster. When nodes are separated, it takes more time for distant nodes to participate in order to reach a consensus. .

Potential impact on security issues

1. Failed to verify node:

Due to internal failure or network disconnection, one or more verification nodes cannot participate in the formation of consensus, so it is necessary to formulate procedures to allow reconnected nodes to catch up with the status of other verification nodes.

We conducted tests to evaluate the consequences of hypothetical failure of a validator node. Specifically, one of the four nodes in total is shut down for a certain period of time, and then restarted and the time required for the node to catch up with other nodes and the time for other nodes are measured.

2. Certificate authorization failed:

Registering and authenticating participants and transactions is the key to ensuring system security. System security. Fabric ensures this through a certificate authority (CA). Although transaction verification is distributed by design, Fabric introduces a single point of failure to the system. In order to determine how Fabric handles the situation where the CA is unavailable, we stopped the CA and restarted it while verifying the node’s ability to continue sending and processing transactions.

The test results show that as long as the CA is unavailable, the transaction will be rejected, reminding the sender to pay attention to the unavailability of Fabric. Once the CA is available again, transaction processing starts without any other system intervention warnings.

3. Resilience to incorrectly formatted requests:

One of the challenges in ensuring the resilience of the DLT system is to ensure that it can continue to operate when a large number of incorrectly formatted transaction requests are submitted. For example, there are unintentional behaviors of participants in the system, and these incorrectly formatted information triggers the error detection mechanism embedded in the smart contract. Tests have shown that regardless of the proportion of incorrectly formatted information, the system has no difficulty in processing transactions in the correct format.

Summary and conclusion

The European Central Bank and the Bank of Japan, as operators of important market infrastructure services, decided to conduct in-depth experiments to determine whether specific existing functions of their respective payment systems can operate in a DLT environment.

The results of research on efficiency show that in terms of the specific aspects of the RTGS service currently tested, the DLT-based solution can meet the performance requirements of the current large-value payment system. Given the nature of DLT, where the process of verifying transactions and reaching consensus is more complicated than that of a centralized system, this is an encouraging conclusion. The project also confirmed the well-known trade-off between network size and performance: increasing the number of verification nodes will lead to an increase in payment execution time. In addition, the distance between verification nodes also has an impact on performance: the time required to process transactions increases with the distance between the verification node groups.

The test results also show that a series of node configurations and system parameters need to be considered when designing the DLT arrangement. As discussed in this report, the number of nodes and the distance between these nodes have a crucial impact on performance. Similarly, system parameters such as the number of transactions grouped in a block and the minimum interval (timeout) required to create a new block may affect the overall delay. Node configuration and parameters should also be taken into consideration. It depends on the needs of the application.


Posted by:CoinYuppie,Reprinted with attribution to:
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.

Like (0)
Donate Buy me a coffee Buy me a coffee
Previous 2021-09-01 12:28
Next 2021-09-01 12:30

Related articles