This study aims to analyze the performance of the Post Quantum Cryptography ML-KEM-768 algorithm in the Industrial Internet of Things system which has limited resources and communication stability needs. The approach used is a quantitative experiment with end-to-end latency measurement on a virtual machine-based architecture that represents IIoT system communication. The evaluation parameters included latency distribution, percentile values, and characteristics of steady state and spike conditions. The test results showed that the ML-KEM-768 had a median latency of 51.84 ms with a P95 value of 64.17 ms and a P99 of 76.85 ms, as well as a steady state proportion of above 96 percent. The spike latency condition persists with low frequency and does not affect the overall stability of the system. These results show that ML-KEM-768 can be implemented in IIoT systems with communication performance that remains within operational tolerance limits.
Copyrights © 2026