Abstract
Edge computing networks need to dynamically distribute their limited and heterogeneous computation resources to handle changing tasks efficiently.
Research Summary
Abstract, contributions, and implementation logic
The report keeps the paper’s own language wherever practical, then adds structure around the algorithm, experiment design, and reported outcomes.
Reported contributions
- Loading contributions from the shared research dataset.
Auction procedure
- Loading methodology steps.
Performance headline
- Auction-based allocation reports 98.2% resource utilization compared with 89.4% for static allocation.
- Task assignment rises from 91.7% to 100%.
- Residual fragmentation falls from 10.6 units to 1.8 units.
- Fairness index improves from 6.4 to 3.2, where lower is better.
Tables
Experiment setup and reported allocations
The first table is loaded from the shared JSON extracted from the local PDF. The allocation and performance tables are reproduced directly as web-readable summaries.
Simulation environment
| Platform | Ubuntu 22.04 on Intel Core i7-12700 CPU with 32 GB RAM |
|---|---|
| Trials | 20 randomized runs per scenario |
| Nodes | 5 heterogeneous edge nodes |
| Tasks | 12 computational tasks |
| Demand classes | Small (2-5), Medium (6-10), Large (11-15) |
Task resource demands
| Task | Demand | Class |
|---|---|---|
| Loading task data. | ||
Task allocation mapping
| Node | Static first-fit | Auction-based |
|---|---|---|
| Node A | T1, T2, T4 | T2, T7, T10 |
| Node B | T3, T5, T6 | T1, T4, T12 |
| Node C | T7, T8 | T3, T8, T11 |
| Node D | T9 | T5, T9 |
| Node E | T10, T12 | T6 |
| Unallocated | T11 | — |
Average performance metrics
| Metric | Static | Auction |
|---|---|---|
| Resource utilization rate | 89.4% | 98.2% |
| Task assignment rate | 91.7% | 100% |
| Residual fragmentation | 10.6 units | 1.8 units |
| Fairness index | 6.4 | 3.2 |
| Tasks successfully placed | 11/12 | 12/12 |
Data integrity notes
- Loading provenance notes.
Figures
Original paper visuals plus rebuilt web-native charts
The extracted figures preserve the paper’s own presentation while the Plotly charts below re-express the underlying variables for easier exploration on the web.
Average metric comparison
Tables III and IVUtilization trend with growing task volume
Figure 6 approximation
Architecture
The original figure shows the system architecture, optimization framing, and the auction steps from announcement to resource updates.
Task allocation comparison
Representative run showing how task distribution differs between static first-fit and auction-based assignment.
Resource utilization by node
The auction strategy spreads work more evenly across nodes in the representative figure.
Remaining resources
Residual capacity becomes less fragmented under the auction-based strategy, which is why large tasks are less likely to be rejected.
Citation
Reusable reference block
The BibTeX entry below is included so the repository can be cited directly from the report page without opening the PDF first.
Loading citation.