TempoNet: Slack-Quantized Transformer-Guided Reinforcement Scheduler for Adaptive Deadline-Centric Real-Time Dispatchs

TL;DR

TempoNet introduces a Transformer-based reinforcement learning scheduler that efficiently manages real-time task dispatching under strict deadlines and compute constraints, demonstrating superior performance and stability.

Contribution

The paper presents TempoNet, a novel Transformer-guided reinforcement scheduler with slack quantization and scalable attention mechanisms for real-time dispatching.

Findings

Achieves higher deadline fulfillment than analytic and neural baselines.

Demonstrates near-linear scaling and sub-millisecond inference in large multiprocessor settings.

Shows improved stability and sample efficiency through behavioral cloning pretraining.

Abstract

Real-time schedulers must reason about tight deadlines under strict compute budgets. We present TempoNet, a reinforcement learning scheduler that pairs a permutation-invariant Transformer with a deep Q-approximation. An Urgency Tokenizer discretizes temporal slack into learnable embeddings, stabilizing value learning and capturing deadline proximity. A latency-aware sparse attention stack with blockwise top-k selection and locality-sensitive chunking enables global reasoning over unordered task sets with near-linear scaling and sub-millisecond inference. A multicore mapping layer converts contextualized Q-scores into processor assignments through masked-greedy selection or differentiable matching. Extensive evaluations on industrial mixed-criticality traces and large multiprocessor settings show consistent gains in deadline fulfillment over analytic schedulers and neural baselines, together with improved optimization stability. Diagnostics include sensitivity analyses for slack quantization, attention-driven policy interpretation, hardware-in-the-loop and kernel micro-benchmarks, and robustness under stress with simple runtime mitigations; we also report sample-efficiency benefits from behavioral-cloning pretraining and compatibility with an actor-critic variant without altering the inference pipeline. These results establish a practical framework for Transformer-based decision making in high-throughput real-time scheduling.