Affiliation:
1. University of Michigan Dearborn, Electrical and Computer Engineering,
USA
2. University of Michigan Dearborn, USA
Abstract
<div>This article presents an optimization scheme for LoRaWAN-based electric vehicle
batteries monitoring system located in warehouses by utilizing techniques to
optimize packet delivery and power settings. Utilizing simulations, we identify
that system optimization largely depends on network traffic, influenced by
active users and the adoption of the pure ALOHA protocol. We define a reward
metric based on the packet delivery rate and power efficiency, aiming for
settings that yield the maximum reward. Our approach includes duty cycle
management to minimize network traffic and maximize throughput, especially
critical when handling urgent data from batteries. Traffic management based on
the number of critical batteries in the warehouse also plays a crucial role.
Predictive modeling of future traffic further refines power settings for optimal
performance. The proposed system, tested through simulations, shows an average
of 31% higher reward compared to traditional methods without duty cycle
management.</div>