Analysis of Application-Level Load Balancing Algorithms

The present article examines the application-level load balancing strategies employed in the microservice architecture of IoT servers, which use wireless communication protocols like Wi-Fi to exchange radio signals among IoT devices. Load balancing at the application level is a crucial task, add to address it, a hypothesis is proposed. To evaluate the effectiveness of various load balancing algorithms, mathematical models based on queuing theory are developed for four algorithms: Round Robin, Weighted Round Robin, Least Connection, and search of the optimal vector. A numerical analysis is conducted to compare the performance of these algorithms. The results confirm the aforementioned hypothesis, and highlight the need to develop a load balancing algorithm that can account for error coefficients arising from distinct services.