Deep Reinforcement Learning for Trustworthy and Time-Varying Connection Scheduling in a Coupled UAV-Based Femtocaching Architecture
Bibliographic record
Abstract
The paper is motivated by the urgent need, imposed by the COVID-19 pandemic, for trustworthy access to secure communication systems with the highest achievable availability and minimum latency. In this regard, we focus on an ultra-dense wireless network consisting of Femto Access Points (FAPs) and Unmanned Aerial Vehicles (UAVs), known as caching nodes, where there are more than one possible caching node to handle user's request. To efficiently cope with the dynamic topology of wireless networks and time-varying behavior of ground users, our focus is to develop an efficient connection scheduling framework, where ground users are autonomously trained to determine the optimal caching node, i.e., UAV or FAP. Our aim is to minimize users' access delay by maintaining a trade-off between the energy consumption of UAVs and the occurrence of handovers. To achieve these objectives, we formulate a multi-objective optimization problem and propose the Convolutional Neural Network (CNN) and Q-Network-based Connection Scheduling (CQN-CS) framework. More specifically, to solve the constructed multi-objective connection scheduling problem, a deep Q-Network model is developed as an efficient Reinforcement Learning (RL) approach to train ground users to handle their requests in an optimal and trustworthy fashion within the coupled UAV-based femtocaching network. The effectiveness of the proposed CQN-CS framework is evaluated in terms of the cache-hit ratio, user's access delay, energy consumption of UAVs, handover, lifetime of the network, and cumulative rewards. Based on the simulation results, the proposed CQN-CS framework illustrates significant performance improvements in companion to Q-learning and Deep Q-Network (DQN) schemes across all the aforementioned aspects.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".