Alam, Mohammad Kowshik and Shuvo, Md Sabbir Hossen and Fahad, Md Lutfur Rahman (2023) Evaluating Autonomous Payment Infrastructure in AI Systems: Measuring Throughput, Latency, and Execution Consistency in Machine-to-Machine Finance. American Journal of Economics and Business Management, 6 (6). pp. 219-246. ISSN 2576-5973

Text AJEBM_Evaluating+Autonomous+Payment (1).pdf - Published Version Download (1MB)

Abstract

With the rise of Artificial Intelligence and the autonomous digital ecosystem, the M2M financial systems that can make transactions without human intervention have been developed rapidly. AASI helps intelligent agents, IoT devices, and AI applications to facilitate efficient digital payments exchange in real-time. In the decentralized finance sector, smart commerce, cloud computing, and automated services, these systems are being more and more adopted. But keeping a high transaction efficiency, low transaction execution delay, and ensuring transaction consistency and reliability across different workloads is still difficult. This study analyzes the operational performance of the autonomous payment infrastructure in AI systems by calculating transaction throughput, transaction execution latency, and transaction execution consistency in machine to machine finance systems. The study employs the PaySim financial transaction dataset, in conjunction with automated rule-based financial transaction software agents, to emulate autonomous financial transactions. Fixed periods of machine-to-machine transactions are initiated between the intelligent agents, forming a credible autonomous financial system. Throughput of the research is the number of completed transactions per time unit and execution latency is the time elapsed between the initiation and completion of a transaction. The ratio of successful transaction executions in comparison with failed transaction executions is used to evaluate execution consistency. The experimental analysis is done for low, medium and high transaction loads to study the behavior of the infrastructure and its scalability under different loads. The experimental results confirm the efficiency of the autonomous payment infrastructure with moderate workloads, in particular, high throughput and low latency. If the volume of transactions increases dramatically, however, it impacts on the performance of the system in a negative way, by adding to processing delay and also decreasing execution consistency. As the workload increases, the number of transactions increases, the likelihood of error rises, and transaction congestion and computational bottlenecks occur. The results of the research enable us to practically define the assessment criteria of autonomous financial infrastructures based on measurable criteria. This research helps propel the growth of AI systems to enable the design of scalable, reliable and efficient autonomous payment systems for future machine-to-machine economies.

Item Type:	Article
Subjects:	A General Works > AI Indexes (General)
Depositing User:	admin eprints
Date Deposited:	08 Jun 2026 17:16
Last Modified:	13 Jun 2026 07:04
URI:	http://eprints.umsida.ac.id/id/eprint/16549

Actions (login required)

View Item