<![CDATA[The rapid growth of the Internet of Things (IoT) has increased the demand for lightweight cryptographic algorithms capable of providing adequate security while maintaining low computational overhead on resource-constrained devices. This study aims to analyze and compare the ASCON, PRESENT, and SPECK algorithms in MQTT-based IoT environments using the MQTT-IoT-IDS2020 dataset as a source of communication payloads. The evaluation was conducted using a Python-based framework with security and randomness metrics, including Avalanche Effect, Strict Avalanche Criterion (SAC), Bit Independence Criterion (BIC), Shannon Entropy, Approximate Entropy, Hamming Distance, Monobit Test, and Runs Test. Performance was evaluated using execution time, throughput, memory usage, and CPU usage. The validity of the results was strengthened through statistical analysis using 95% confidence intervals, One-Way ANOVA, and Tukey HSD. The results indicate that all three algorithms exhibit strong diffusion and randomness characteristics, with Avalanche Effect values close to the ideal value of 50% and ciphertext randomness metrics that satisfy the applied statistical tests. SPECK achieved the best performance with an execution time of 0.000238 seconds and a throughput of 2,199,891 bits/s, while PRESENT demonstrated the lowest memory consumption at 2.85 KB. Based on the calculated Security Score and Efficiency Score, PRESENT achieved the highest Security Score of 0.692, while SPECK achieved the highest Efficiency Score of 0.993, respectively. Statistical analysis revealed that significant differences among the algorithms primarily occurred in diffusion-related and computational performance metrics. Therefore, within the scope of the evaluated diffusion, randomness, and performance metrics, SPECK provides the most favorable balance between security-related characteristics and computational efficiency among the evaluated algorithms for MQTT-based IoT environments.]]>
