This paper presents a control/scheduling co-design framework that integrates 5G Configured Grant (CG) scheduling with networked control systems (NCS) design. The objective is to minimize the hyperperiod induced by multiple, application-specific sampling periods, which determines the schedule table size and memory footprint at the base station, subject to control quality and wireless resource limits. Hyperperiod minimization under control and resource constraints is nontrivial due to the combinatorial nature of discrete sampling period choices. To address this challenge, we propose a two-stage Hyperperiod-Minimization-oriented Period Assignment (HMPA) method. In the first stage, HMPA performs a feasibility-oriented period search within candidate period sets constructed from restricted primes and exponents, which bound the hyperperiod. In the second stage, a hyperperiod refinement procedure exploits remaining resource slack to further reduce the hyperperiod while preserving feasibility. Experiments demonstrate the efficiency of the proposed framework in terms of finding solutions with significantly reduced hyperperiods.

The fifth-generation (5G) technology standard in telecommunications is expected to support ultra-reliable low latency communication to enable real-time applications such as industrial automation and control. 5G configured grant (CG) scheduling features a pre-allocated periodicity-based scheduling approach, which reduces control signaling time and guarantees service quality. Although this enables 5G to support hard real-time periodic traffics, synthesizing the schedule efficiently and achieving high resource efficiency, while serving multiple communications, are still an open problem. In this work, we study the trade-off between scheduling flexibility and control overhead when performing CG scheduling. To address the CG scheduling problem, we first formulate it using satisfiability modulo theories (SMT) so that an SMT solver can be used to generate optimal solutions. To enhance scalability, we propose two heuristic approaches. The first one as the baseline, Co1, follows the basic idea of the 5G CG scheduling scheme that minimizes the control overhead. The second one, CoU, enables increased scheduling flexibility while considering the involved control overhead. The effectiveness and scalability of the proposed techniques and the superiority of CoU compared to Co1 have been evaluated using a large number of generated benchmarks as well as a realistic case study for industrial automation.

5G is expected to support ultra-reliable low latency communication to enable real-time applications such as industrial automation and control. 5G configured grant (CG) scheduling features a pre-allocated periodicity-based scheduling approach which reduces control signaling time and guarantees service quality. Although this enables 5G to support hard real-time periodic traffics, efficiently synthesizing the schedule and achieving high resource efficiency while serving multiple traffics, is still an open problem. To address this problem, we first formulate it using satisfiability modulo theories (SMT) so that an SMT-solver can be used to generate optimal solutions. For enhancing scalability, two efficient heuristic approaches are proposed. The experiments demonstrate the effectiveness and scalability of the proposed technique.

To support nonsequential writes, persistent cache (PC) is constructed in drive managed SMR (DM-SMR) drive. However, PC cleaning introduces drastic performance degradation and enlarges tail latencies. In this article, we propose to utilize reinforcement learning (RL) to mitigate the long-tail latency of PC cleaning. Our scheme uses the lightweight$Q$-learning method to monitor and learn the idle time of I/O workloads, based on which PC cleaning is intelligently guided, thus maximally exploit idle time between requests and hiding tail latency from normal requests. In addition, a multiagent RL scheme with clustering algorithm is adopted to further mitigate the tail latencies and adapt to variable workloads. We emulate a DM-SMR drive inside a Linux device driver to implement our proposed scheme. According to the experimental results, our scheme can effectively reduce the tail latency by 59.45% at the 99.9th percentile and the average latency by 48.75% compared with a typical shingled magnetic recording (SMR) design.

DM-SMR adopts Persistent Cache (PC) to accommodate non-sequential write operations. However, the PC cleaning process induces severe long-tail latency. In this paper, we propose to mitigate the tail latency of PC cleaning by using Reinforcement Learning (RL). Specifically, a real-time lightweight Q-learning model is built to analyze the idle window of I/O workloads, based on which PC cleaning is judiciously scheduled, thereby maximally utilizing the I/O idle window and effectively hiding the tail latency from regular requests. We implement our technique inside a Linux device driver with an emulated SMR drive. Experimental results show that our technique can reduce the tail latency by 57.65% at 99.9th percentile and the average response time by 46.11% compared to a typical SMR design.

Weight sparsity has been explored to achieve energy efficiency for Resistive Random-access Memory (ReRAM) based DNN accelerators. However, most existing ReRAM-based DNN accelerators are based on an overidealized crossbar architecture and mainly focus on compressing zero weights. In this paper, we propose a novel ReRAM-based accelerator — PattPIM, to achieve space compression and computation reuse by studying DNN weight patterns based on practical ReRAM crossbars. We first thoroughly analyze the weight distribution characteristics of several typical DNN models and observe many non-zero weight pattern repetitions (WPRs). Thus, in PattPIM, we propose a WPR-aware DNN engine and a WPR-to-OU mapping scheme to save both space and computation resources. Furthermore, we adopt an approximate weight pattern transform algorithm to improve the DNN WPRs ratio to enhance the reuse efficiency with negligible inference accuracy loss. Our evaluation with 6 DNN models shows that the proposed PattPIM delivers significant performance improvement, ReRAM resources efficiency and energy saving.

Non-intrusive load monitoring (a.k.a. power disaggregation) refers to identifying and extracting the consumption patterns of individual appliances from the mains which records the whole-house energy consumption. Recently, deep learning has been shown to be a promising method to solve this problem and many approaches based on it have been proposed. In this paper, we propose a sequence-to-subsequence learning method, which makes a trade-off between traditional sequence-to-sequence and sequence-to-point method, to balance the convergence difficulty in deep neural networks and the amount of computation in the inference period. We build our model based on conditional generative adversarial network that helps us avoid designing the loss function manually. In addition, we apply U-Net and Instance Normalization techniques to our model and demonstrate their effectiveness. Evaluations are performed on real-world data sets and we achieve the state-of-the-art performance.