Task scheduling in cloud computing systems using honey badger algorithm with improved density factor and foucault pendulum motion.

Cloud computing is an emerging technology in the field of distributed computing with the flexibility to pay per usage based on user needs and requirements. Task scheduling is considered to be a NP-hard problem, which directly affects the operational efficiency of the whole system, load balancing and system energy consumption, so it is challenging to find the best solution. A honey badger algorithm (HBA) based on improved density factor and Foucault pendulum motion is proposed to improve the efficiency of task execution in cloud computing systems. It improves the digging phase of the honey badger's foraging strategy using representations of Foucault pendulum motion in a right-angle coordinate system and a spherical coordinate system, respectively, and improves the density factor by using a variable-order sinusoidal curve. The performance of the proposed improvement scheme is tested by using 23 benchmark functions. Simulation experiments are conducted for the total cost, time cost, load cost and price cost of the system under large-scale and small-scale tasks. Compared to traditional scheduling algorithms such as ACO, PSO, WOA, AOA, RSO, SOA, CDO, RIME and GGO, HBA-Z (10) reduces about 15%, 39% and 12% in total, load and price costs over the next best algorithm in the small-scale task case, and about 25%, 51% and 27% over the worst algorithm. In the case of large-scale tasks, HBA-Z (10) reduces the total cost, load cost and price cost by about 16%, 40% and 14% compared to the next best algorithm, and reduces about 25%, 52% and 26% compared to the worst algorithm. Experimental results show that the proposed HBA-Z (10) has significant advantages in efficiently searching for optimal task scheduling policy. [ABSTRACT FROM AUTHOR]