ERAP Optimization via Enhanced Constraints and Boundary Detection in GMRA
Bibliographic record
Abstract
Edge computing allows edge devices to offload computational tasks to edge servers, utilizing various hardware resources for efficient computation. Unlike cloud facilities, edge servers have limited resources. A long-term challenge is to quickly evaluate all the edge server resources and select the suitable server for the task, with high requirements for both processing time and allocation effect. The Edge Resource Allocation Problem (ERAP) represents a typical agent evaluation in collaborative work and falls within the realm of the Group Multi-Role Assignment (GMRA) problem. Based on the GMRA model, we formalize the ERAP as an optimization problem with an improved Edge- GMRA model. Additionally, we investigate the feasibility of an enhanced constraint scheme in the improved model. By boundary detection scheme, we implement quickly eliminated the infeasible solutions within the search range for ERAP. Experimental results demonstrate that the enhanced constraint scheme improves the allocation of high-priority tasks with superior acceleration as the number of agents increases, and the boundary detection scheme performs effectively in scenarios with insufficient server resources. The combination of these two schemes significantly accelerates the solution process, achieving an acceleration ratio exceeding 50%. The proposed dynamic adaptation mechanism with asynchronous agent monitoring and sliding-window threshold adjustment maintains the stability of the system under fluctuation of 15% resources, while our task-type recognition system demonstrates 92. 4% classification accuracy in six workload categories. Extensive evaluation shows the framework sustains sub-100ms decision latency during 80% resource contention scenarios, achieving 23% higher throughput than conventional methods while reducing service-level agreement violations by 41% in dynamic edge environments.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.005 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".