Cost–benefit evaluation of computer network systems with warm standby units and general repair times

Abstract

The rapid growth of artificial intelligence (AI) and internet of things (IoT) across several industries emphasizes on the necessity for reliable and strong network infrastructure. In our increasingly interconnected world, network availability and reliability are critical for the success of diverse sectors such as businesses, academic institutions, communication, and healthcare systems. However, issues like software bugs, hardware malfunctions, and human error pose serious threats to the stability of the network. Addressing these problems is becoming more and more crucial as the need for uninterrupted connectivity keeps rising. To address these challenges this paper compares three retrial computer network systems comprising warm standby units, a single repairman with planned vacations, and general repair time distributions. The steady state probabilities and related availabilities for three retrial systems are systematically derived by employing supplementary variable technique and recursive analytical method. Additionally, comparison of the cost–benefit ratios and availabilities for these systems is presented. The best retrial system is obtained by sorting the efficiency of the three retrial systems under consideration. The adopted technique outperforms typical models by a significant margin, according to numerical results, this enhances efficiency of system availability even in case of unit failures. Based on the calculated numerical outcomes, more strategies for decision making management are also suggested for practical implications.