Abstract
<jats:p>This paper presents a comprehensive analysis of architectural methods for optimizing computational resources of CubeSat standard nanosatellites within the transition to the NewSpace paradigm. The fundamental problem of imbalance between exponentially growing onboard computing requirements and severe hardware constraints of 1U-3U platform formats is examined, characterized by critical power deficit, high radiation vulnerability of components, and limited memory capacity. A detailed comparative efficiency analysis of leading real-time operating systems FreeRTOS and Zephyr OS was conducted based on context switching latency, energy efficiency, and reliability criteria, demonstrating FreeRTOS advantages for power-constrained missions due to minimal kernel overhead at 223 machine cycles. A mathematical model for the onboard computer's power consumption has been formalized, accounting for dynamic and static power components relative to the space thermal environment and the impact of radiation on leakage currents in semiconductor structures. The prospects of implementing a hybrid Edge-to-Cloud architecture are investigated, involving intelligent offloading of resource-intensive computational tasks to ground-based cloud environments via GSaaS (Ground Station as a Service), with detailed analysis of energy and communication feasibility criteria for offloading. The necessity of using software-hardware methods for ensuring radiation tolerance is substantiated, including EDAC memory scrubbing, and multi-level watchdog timer systems consuming up to 20% of processor resources. The feasibility of lightweight WebAssembly-based containerization technologies is demonstrated by recent experimental studies, confirming their applicability to software-defined satellite architectures with secure in-orbit updates and fault isolation.</jats:p>