Microprocessor performance for swarm UAV application in distributed environmental monitoring systems

Abstract

This study investigates the microprocessor performance of unmanned aerial vehicle (UAV) swarms used in distributed environmental monitoring and precision agriculture. A parametric analytical model is applied to describe the effective swarm throughput as a function of the number of UAVs, processor speed, internal processor overhead, interprocessor overhead, and the computational cost of message processing. The model accounts for the simultaneous increase in aggregate computing capacity and coordination costs as the swarm size grows. Performance evaluation and sensitivity analysis are conducted for a representative agricultural UAV swarm. The results demonstrate a non-monotonic dependence of microprocessor performance on swarm size. For the selected model parameters, the maximum throughput of approximately 120 messages/s is achieved with five to six UAVs. The sensitivity analysis shows that, for a five-UAV swarm, an increase in the interprocessor overhead ratio directly decreases system throughput. The proposed approach can support early-stage engineering decisions concerning UAV platform selection, rational swarm sizing, system scaling, and redundancy planning. The findings confirm that efficient swarm operation depends not only on the nominal computing power of individual UAVs but also on the organization of interprocessor interaction.