Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Hybrid Proactive And Predictive Framework For Edge Cloud Resource Management

Hrikshesh Kumar, Anika Garg, Anshul Gupta, Yashika Agarwal

TL;DR

The paper tackles the challenge of unpredictable edge workloads by moving beyond reactive resource management to a proactive framework that combines CNN-LSTM time-series forecasting with a multi-agent DRL orchestrator. By embedding forecast information into the DRL agent's lookahead state, the system learns proactive policies that balance latency, energy, and cost while avoiding SLA violations. The authors implement a CNN-LSTM predictor and a CTDE-enabled MADRL orchestrator, evaluate against a DDQN baseline in a simulated iFogSimEnv with synthetic workloads, and demonstrate substantial improvements in total reward, cost, and energy—though with more variable makespan due to long-horizon planning. They also outline promising future directions, including privacy-preserving Federated Learning, uncertainty-aware DRL, and sim-to-real deployment strategies.

Abstract

Old cloud edge workload resource management is too reactive. The problem with relying on static thresholds is that we are either overspending for more resources than needed or have reduced performance because of their lack. This is why we work on proactive solutions. A framework developed for it stops reacting to the problems but starts expecting them. We design a hybrid architecture, combining two powerful tools: the CNN LSTM model for time series forecasting and an orchestrator based on multi agent Deep Reinforcement Learning In fact the novelty is in how we combine them as we embed the predictive forecast from the CNN LSTM directly into the DRL agent state space. That is what makes the AI manager smarter it sees the future, which allows it to make better decisions about a long term plan for where to run tasks That means finding that sweet spot between how much money is saved while keeping the system healthy and apps fast for users That is we have given it eyes in order to see down the road so that it does not have to lurch from one problem to another it finds a smooth path forward Our tests show our system easily beats the old methods It is great at solving tough problems like making complex decisions and juggling multiple goals at once like being cheap fast and reliable

A Hybrid Proactive And Predictive Framework For Edge Cloud Resource Management

TL;DR

The paper tackles the challenge of unpredictable edge workloads by moving beyond reactive resource management to a proactive framework that combines CNN-LSTM time-series forecasting with a multi-agent DRL orchestrator. By embedding forecast information into the DRL agent's lookahead state, the system learns proactive policies that balance latency, energy, and cost while avoiding SLA violations. The authors implement a CNN-LSTM predictor and a CTDE-enabled MADRL orchestrator, evaluate against a DDQN baseline in a simulated iFogSimEnv with synthetic workloads, and demonstrate substantial improvements in total reward, cost, and energy—though with more variable makespan due to long-horizon planning. They also outline promising future directions, including privacy-preserving Federated Learning, uncertainty-aware DRL, and sim-to-real deployment strategies.

Abstract

Old cloud edge workload resource management is too reactive. The problem with relying on static thresholds is that we are either overspending for more resources than needed or have reduced performance because of their lack. This is why we work on proactive solutions. A framework developed for it stops reacting to the problems but starts expecting them. We design a hybrid architecture, combining two powerful tools: the CNN LSTM model for time series forecasting and an orchestrator based on multi agent Deep Reinforcement Learning In fact the novelty is in how we combine them as we embed the predictive forecast from the CNN LSTM directly into the DRL agent state space. That is what makes the AI manager smarter it sees the future, which allows it to make better decisions about a long term plan for where to run tasks That means finding that sweet spot between how much money is saved while keeping the system healthy and apps fast for users That is we have given it eyes in order to see down the road so that it does not have to lurch from one problem to another it finds a smooth path forward Our tests show our system easily beats the old methods It is great at solving tough problems like making complex decisions and juggling multiple goals at once like being cheap fast and reliable

Paper Structure

This paper contains 16 sections, 2 equations, 2 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Key Contributions
  3. System Model and Problem Formulation
  4. Defining the Proactive Resource Management Problem
  5. Proposed System Architecture
  6. Predictive Module (CNN-LSTM)
  7. DRL Orchestrator (DQN Agent)
  8. Base Model (DDQN)
  9. Methodology and Implementation Details
  10. Simulation Environment and Workload
  11. Multi-Objective Reward Function
  12. Proposed Algorithm
  13. Experimental Setup and Results
  14. Experimental Setup
  15. Results Discussion
  16. ...and 1 more sections

Figures (2)

  • Figure 1: Fig. 1: System Architecture Diagram
  • Figure 2: Experimental Results: (a) Baseline DDQN performance metrics. (b) Hybrid model's makespan convergence.