TASK SCHEDULING OPTIMIZATION METHOD IN A PATIENT HEALTH MONITORING INFORMATION SYSTEM

Abstract

Within the conducted research, an method for optimizing task scheduling in real-time health monitoring information systems for patients has been developed, based on the evolutionary Particle Swarm Optimization (PSO) algorithm. The proposed approach ensures adaptive allocation of computational resources, taking into account the high dynamism of data inflow from medical sensors that track vital signs – heart rate, blood oxygen levels, arterial blood pressure, etc. Unlike traditional scheduling methods such as FCFS (First Come, First Served) and EDF (Earliest Deadline First), which are unable to effectively respond to variable task priorities and strict time constraints, the PSO model simulates the collective behavior of agents (particles) in a multidimensional space of possible schedules, minimizing a comprehensive cost function. This function accounts for actual execution delays, system load levels, and the proportion of critical data loss, thereby preventing communication channel overload and ensuring instantaneous processing of alarm signals.

Experimental testing confirmed the high efficiency of the method. Simulation results showed that over 96% of tasks were processed within acceptable time frames, with the average waiting time reduced to 1.9 seconds — a critical indicator for medical systems where every fraction of a second can impact intervention outcomes. In load tests simulating the simultaneous operation of 200 sensors, the system maintained full stability, data processing continuity, and operational responsiveness, demonstrating high scalability and resilience to peak loads. Comparison with baseline algorithms revealed significant reductions in delays, decreased probability of data loss, and optimal resource utilization.

Thus, the proposed PSO-based approach opens new opportunities for creating reliable, flexible, and high-performance telemedicine and remote monitoring systems, contributing to improved quality of medical care under real-time conditions and high dynamics of information flows.