Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Multi-Layered Distributed Computing Framework for Enhanced Edge Computing

Ke Ma, Junfei Xie

TL;DR

A new distributed computing framework is introduced that extends resource sharing beyond one-hop neighborhoods through exploring layered network structures and proposes two exact methods that find optimal solutions and three heuristic strategies to improve efficiency and scalability.

Abstract

The rise of the Internet of Things and edge computing has shifted computing resources closer to end-users, benefiting numerous delay-sensitive, computation-intensive applications. To speed up computation, distributed computing is a promising technique that allows parallel execution of tasks across multiple compute nodes. However, current research predominantly revolves around the master-worker paradigm, limiting resource sharing within one-hop neighborhoods. This limitation can render distributed computing ineffective in scenarios with limited nearby resources or constrained/dynamic connectivity. In this paper, we address this limitation by introducing a new distributed computing framework that extends resource sharing beyond one-hop neighborhoods through exploring layered network structures. Our framework involves transforming the network graph into a sink tree and formulating a joint optimization problem based on the layered tree structure for task allocation and scheduling. To solve this problem, we propose two exact methods that find optimal solutions and three heuristic strategies to improve efficiency and scalability. The performances of these methods are analyzed and evaluated through theoretical analyses and comprehensive simulation studies. The results demonstrate their promising performances over the traditional distributed computing and computation offloading strategies.

A Multi-Layered Distributed Computing Framework for Enhanced Edge Computing

TL;DR

A new distributed computing framework is introduced that extends resource sharing beyond one-hop neighborhoods through exploring layered network structures and proposes two exact methods that find optimal solutions and three heuristic strategies to improve efficiency and scalability.

Abstract

The rise of the Internet of Things and edge computing has shifted computing resources closer to end-users, benefiting numerous delay-sensitive, computation-intensive applications. To speed up computation, distributed computing is a promising technique that allows parallel execution of tasks across multiple compute nodes. However, current research predominantly revolves around the master-worker paradigm, limiting resource sharing within one-hop neighborhoods. This limitation can render distributed computing ineffective in scenarios with limited nearby resources or constrained/dynamic connectivity. In this paper, we address this limitation by introducing a new distributed computing framework that extends resource sharing beyond one-hop neighborhoods through exploring layered network structures. Our framework involves transforming the network graph into a sink tree and formulating a joint optimization problem based on the layered tree structure for task allocation and scheduling. To solve this problem, we propose two exact methods that find optimal solutions and three heuristic strategies to improve efficiency and scalability. The performances of these methods are analyzed and evaluated through theoretical analyses and comprehensive simulation studies. The results demonstrate their promising performances over the traditional distributed computing and computation offloading strategies.
Paper Structure (43 sections, 2 theorems, 21 equations, 12 figures, 1 table, 2 algorithms)

This paper contains 43 sections, 2 theorems, 21 equations, 12 figures, 1 table, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. Distributed Computing
  4. Computation Offloading
  5. System Model and Problem Description
  6. System Model
  7. Network Model
  8. Computing model
  9. Communication Model
  10. Energy Consumption Model
  11. Problem Description and Analysis
  12. Multi-Layered Distributed Computing Framework
  13. Transforming Graph into a Sink Tree
  14. Mixed Integer Programming Model
  15. Decision Variables
  16. ...and 28 more sections

Key Result

Lemma 1

Given $\mathcal{T}$ and $Y$, for an arbitrary offloading order $\boldsymbol{o}_k$, suppose $\{{\boldsymbol{y}^*}(\boldsymbol{o}_k), {z}^*(\boldsymbol{o}_k)\}$ is an optimal solution to problem $\mathcal{P}_1^{(a)}$. Then, $\{\frac{Y'}{Y}{\boldsymbol{y}^*}(\boldsymbol{o}_k), \frac{Y'}{Y}{z}^*(\boldsy

Figures (12)

  • Figure 1: Network scenario.
  • Figure 2: An example network represented by a layered tree structure. Servers' indices are highlighted in red.
  • Figure 3: Illustration of how a network tree can be abstracted as a one-layer tree.
  • Figure 4: Network topologies evaluated in simulation studies.
  • Figure 5: Total cost $J$ of different methods when considering (a) only time consumption; and (b) both time and energy consumption.
  • ...and 7 more figures

Theorems & Definitions (4)

  • Lemma 1
  • proof
  • Theorem 2
  • proof