skip to main content

Explore Research Products in the NSF-PAR

Title: PnP-DRL: A Plug-and-Play Deep Reinforcement Learning Approach for Experience-Driven Networking
While Deep Reinforcement Learning has emerged as a de facto approach to many complex experience-driven networking problems, it remains challenging to deploy DRL into real systems. Due to the random exploration or half-trained deep neural networks during the online training process, the DRL agent may make unexpected decisions, which may lead to system performance degradation or even system crash. In this paper, we propose PnP-DRL, an offline-trained, plug and play DRL solution, to leverage the batch reinforcement learning approach to learn the best control policy from pre-collected transition samples without interacting with the system. After being trained without interaction with systems, our Plug and Play DRL agent will start working seamlessly, without additional exploration or possible disruption of the running systems. We implement and evaluate our PnP-DRL solution on a prevalent experience-driven networking problem, Dynamic Adaptive Streaming over HTTP (DASH). Extensive experimental results manifest that 1) The existing batch reinforcement learning method has its limits; 2) Our approach PnP-DRL significantly outperforms classical adaptive bitrate algorithms in average user Quality of Experience (QoE); 3) PnP-DRL, unlike the state-of-the-art online DRL methods, can be off and running without learning gaps, while achieving comparable performances.  more » « less
Award ID(s):
1704662
NSF-PAR ID:
10300535
Author(s) / Creator(s):
; ; ; ; ;
Date Published:
Journal Name:
IEEE journal on selected areas in communications
Volume:
39
Issue:
8
ISSN:
1558-0008
Page Range / eLocation ID:
2476-2486
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ( , North American Power Symposium)
    Mostafa Sahraei-Ardakani ; Mingxi Liu (Ed.)
    This paper explores the application of deep reinforcement learning (DRL) to create a coordinating mechanism between synchronous generators (SGs) and distributed energy resources (DERs) for improved primary frequency regulation. Renewable energy sources, such as wind and solar, may be used to aid in frequency regulation of the grid. Without proper coordination between the sources, however, the participation only results in a delay of SG governor response and frequency deviation. The proposed DRL application uses a deep deterministic policy gradient (DDPG) agent to create a generalized coordinating signal for DERs. The coordinating signal communicates the degree of distributed participation to the SG governor, resolving delayed governor response and reducing system rate of change of frequency (ROCOF). The validity of the coordinating signal is presented with a single-machine finite bus system. The use of DRL for signal creation is explored in an under-frequency event. While further exploration is needed for validation in large systems, the development of this concept shows promising results towards increased power grid stabilization. 
    more » « less
  2. ; ; ( , Proceedings of the 17th International Conference on Emerging Networking EXperiments and Technologies)
    Network slicing allows mobile network operators to virtualize infrastructures and provide customized slices for supporting various use cases with heterogeneous requirements. Online deep reinforcement learning (DRL) has shown promising potential in solving network problems and eliminating the simulation-to-reality discrepancy. Optimizing cross-domain resources with online DRL is, however, challenging, as the random exploration of DRL violates the service level agreement (SLA) of slices and resource constraints of infrastructures. In this paper, we propose OnSlicing, an online end-to-end network slicing system, to achieve minimal resource usage while satisfying slices' SLA. OnSlicing allows individualized learning for each slice and maintains its SLA by using a novel constraint-aware policy update method and proactive baseline switching mechanism. OnSlicing complies with resource constraints of infrastructures by using a unique design of action modification in slices and parameter coordination in infrastructures. OnSlicing further mitigates the poor performance of online learning during the early learning stage by offline imitating a rule-based solution. Besides, we design four new domain managers to enable dynamic resource configuration in radio access, transport, core, and edge networks, respectively, at a timescale of subseconds. We implement OnSlicing on an end-to-end slicing testbed designed based on OpenAirInterface with both 4G LTE and 5G NR, OpenDayLight SDN platform, and OpenAir-CN core network. The experimental results show that OnSlicing achieves 61.3% usage reduction as compared to the rule-based solution and maintains nearly zero violation (0.06%) throughout the online learning phase. As online learning is converged, OnSlicing reduces 12.5% usage without any violations as compared to the state-of-the-art online DRL solution. 
    more » « less
  3. ; ; ; ; ( , 2022 IEEE Global Communications Conference)
    In this work, we propose an energy-adaptive moni-toring system for a solar sensor-based smart animal farm (e.g., cattle). The proposed smart farm system aims to maintain high-quality monitoring services by solar sensors with limited and fluctuating energy against a full set of cyberattack behaviors including false data injection, message dropping, or protocol non-compliance. We leverage Subjective Logic (SL) as the belief model to consider different types of uncertainties in opinions about sensed data. We develop two Deep Reinforcement Learning (D RL) schemes leveraging the design concept of uncertainty maximization in SL for DRL agents running on gateways to collect high-quality sensed data with low uncertainty and high freshness. We assess the performance of the proposed energy-adaptive smart farm system in terms of accumulated reward, monitoring error, system overload, and battery maintenance level. We compare the performance of the two DRL schemes developed (i.e., multi-agent deep Q-Iearning, MADQN, and multi-agent proximal policy optimization, MAPPO) with greedy and random baseline schemes in choosing the set of sensed data to be updated to collect high-quality sensed data to achieve resilience against attacks. Our experiments demonstrate that MAPPO with the uncertainty maximization technique outperforms its counterparts. 
    more » « less
  4. ; ; ; ( , Artificial Intelligence in Education. AIED 2020)
    null (Ed.)
    In recent years, Reinforcement learning (RL), especially Deep RL (DRL), has shown outstanding performance in video games from Atari, Mario, to StarCraft. However, little evidence has shown that DRL can be successfully applied to real-life human-centric tasks such as education or healthcare. Different from classic game-playing where the RL goal is to make an agent smart, in human-centric tasks the ultimate RL goal is to make the human-agent interactions productive and fruitful. Additionally, in many real-life human-centric tasks, data can be noisy and limited. As a sub-field of RL, batch RL is designed for handling situations where data is limited yet noisy, and building simulations is challenging. In two consecutive classroom studies, we investigated applying batch DRL to the task of pedagogical policy induction for an Intelligent Tutoring System (ITS), and empirically evaluated the effectiveness of induced pedagogical policies. In Fall 2018 (F18), the DRL policy is compared against an expert-designed baseline policy and in Spring 2019 (S19), we examined the impact of explaining the batch DRL-induced policy with student decisions and the expert baseline policy. Our results showed that 1) while no significant difference was found between the batch RL-induced policy and the expert policy in F18, the batch RL-induced policy with simple explanations significantly improved students’ learning performance more than the expert policy alone in S19; and 2) no significant differences were found between the student decision making and the expert policy. Overall, our results suggest that pairing simple explanations with induced RL policies can be an important and effective technique for applying RL to real-life human-centric tasks. 
    more » « less
  5. ; ; ; ; ( , IEEE/ACM IWQoS)
    null (Ed.)
    In this paper, we present design, implementation and evaluation of a control framework, EXTRA (EXperience-driven conTRol frAmework), for scheduling in general-purpose Distributed Stream Data Processing Systems (DSDPSs). Our design is novel due to the following reasons. First, EXTRA enables a DSDPS to dynamically change the number of threads on the fly according to system states and demands. Most existing methods, however, use a fixed number of threads to carry workload (for each processing unit of an application), which is specified by a user in advance and does not change during runtime. So our design introduces a whole new dimension for control in DSDPSs, which has a great potential to significantly improve system flexibility and efficiency, but makes the scheduling problem much harder. Second, EXTRA leverages an experience/data driven model-free approach for dynamic control using the emerging Deep Reinforcement Learning (DRL), which enables a DSDPS to learn the best way to control itself from its own experience just as a human learns a skill (such as driving and swimming) without any accurate and mathematically solvable model. We implemented it based on a widely-used DSDPS, Apache Storm, and evaluated its performance with three representative Stream Data Processing (SDP) applications: continuous queries, word count (stream version) and log stream processing. Particularly, we performed experiments under realistic settings (where multiple application instances are mixed up together), rather than a simplified setting (where experiments are conducted only on a single application instance) used in most related works. Extensive experimental results show: 1) Compared to Storm’s default scheduler and the state-of-the-art model-based method, EXTRA substantially reduces average end-to-end tuple processing time by 39.6% and 21.6% respectively on average. 2) EXTRA does lead to more flexible and efficient stream data processing by enabling the use of a variable number of threads. 3) EXTRA is robust in a highly dynamic environment with significant workload change. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email