Answers are grounded in this paper and cite their sources.
Computing first networks supported by elastic optical networks provide flexible spectrum resource allocation for cross-region computation offloading. When multiple task requests arrive concurrently within the same decision time slot, tasks may still be blocked even when candidate computing nodes with matching service types and sufficient resources are available. This occurs because contiguous frequency slot resources along the selected transmission path are insufficient, or because the end-to-end task completion delay exceeds the prescribed threshold. To address this, we consider a static scenario with concurrent multi-task offloading and formulate a joint optimization model that accounts for task completion delay, the number of blocked tasks, and electricity cost. A GA-guided selective repair and local search (GSRL) algorithm is then proposed, which optimizes computing node selection and spectrum allocation while reducing the search space. GSRL is compared against three benchmark algorithms on 30-node and 56-node network topologies. Simulation results show that GSRL effectively reduces average task completion delay and blocking rate while keeping system electricity cost under control.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: