Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Improved Methods of Task Assignment and Resource Allocation with Preemption in Edge Computing Systems

Caroline Rublein, Fidan Mehmeti, Mark Mahon, Thomas F. La Porta

TL;DR

This work tackles online, distributed edge-cloud resource allocation under limited information by introducing a preemption-enabled two-round auction framework. It formalizes an optimal centralized problem and then proposes a scalable Knapsack Greedy heuristic, augmented by a Double Knapsack with Preemption for comparison, to balance solution quality and online speed. Across pipeline, batch, and real-trace evaluations, the greedy approach achieves near-optimal performance with considerably faster auctions, and preemption enables higher-value tasks to complete under tight deadlines. The findings demonstrate meaningful improvements in utility and practical feasibility for edge-enabled task assignment and resource management.

Abstract

Edge computing has become a very popular service that enables mobile devices to run complex tasks with the help of network-based computing resources. However, edge clouds are often resource-constrained, which makes resource allocation a challenging issue. In addition, edge cloud servers must make allocation decisions with only limited information available, since the arrival of future client tasks might be impossible to predict, and the states and behavior of neighboring servers might be obscured. We focus on a distributed resource allocation method in which servers operate independently and do not communicate with each other, but interact with clients (tasks) to make allocation decisions. We follow a two-round bidding approach to assign tasks to edge cloud servers, and servers are allowed to preempt previous tasks to allocate more useful ones. We evaluate the performance of our system using realistic simulations and real-world trace data from a high-performance computing cluster. Results show that our heuristic improves system-wide performance by $20-25\%$ over previous work when accounting for the time taken by each approach. In this way, an ideal trade-off between performance and speed is achieved.

Improved Methods of Task Assignment and Resource Allocation with Preemption in Edge Computing Systems

TL;DR

This work tackles online, distributed edge-cloud resource allocation under limited information by introducing a preemption-enabled two-round auction framework. It formalizes an optimal centralized problem and then proposes a scalable Knapsack Greedy heuristic, augmented by a Double Knapsack with Preemption for comparison, to balance solution quality and online speed. Across pipeline, batch, and real-trace evaluations, the greedy approach achieves near-optimal performance with considerably faster auctions, and preemption enables higher-value tasks to complete under tight deadlines. The findings demonstrate meaningful improvements in utility and practical feasibility for edge-enabled task assignment and resource management.

Abstract

Edge computing has become a very popular service that enables mobile devices to run complex tasks with the help of network-based computing resources. However, edge clouds are often resource-constrained, which makes resource allocation a challenging issue. In addition, edge cloud servers must make allocation decisions with only limited information available, since the arrival of future client tasks might be impossible to predict, and the states and behavior of neighboring servers might be obscured. We focus on a distributed resource allocation method in which servers operate independently and do not communicate with each other, but interact with clients (tasks) to make allocation decisions. We follow a two-round bidding approach to assign tasks to edge cloud servers, and servers are allowed to preempt previous tasks to allocate more useful ones. We evaluate the performance of our system using realistic simulations and real-world trace data from a high-performance computing cluster. Results show that our heuristic improves system-wide performance by over previous work when accounting for the time taken by each approach. In this way, an ideal trade-off between performance and speed is achieved.
Paper Structure (21 sections, 2 equations, 20 figures, 2 tables, 2 algorithms)

This paper contains 21 sections, 2 equations, 20 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. System Overview
  4. Optimization Formulation
  5. Heuristic Methods
  6. Knapsack Greedy Heuristic
  7. Round 1 Procedure
  8. Preemptive Round 2 Procedure
  9. Complexity Analysis
  10. Double Knapsack with Preemption
  11. Performance Evaluation
  12. Pipeline Paradigm
  13. Optimal
  14. Simulation Setup
  15. Results
  16. ...and 6 more sections

Figures (20)

  • Figure 1: The arrival of jobs, along with the bidding and processing procedures.
  • Figure 2: Relationship described by constraints (\ref{['eq:results_tau']})-(\ref{['eq:down_limit']}), demonstrated using a real-workload job's progression under our KnapsackGreedy heuristic
  • Figure 3: Discounts given by Server $5$ to incoming jobs in Round 1 of timestep 43 during a run of the KnapsackGreedy heuristic (pipeline paradigm) using a normally-distributed workload.
  • Figure 4: Discounts received by job number 532 in Round 1 of timestep 43 during a run of the KnapsackGreedy heuristic (pipeline paradigm) using a normally-distributed workload.
  • Figure 5: Discounts received by job number 540 in Round 1 of timestep 43 during a run of the KnapsackGreedy heuristic (pipeline paradigm) using a normally-distributed workload.
  • ...and 15 more figures