Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fully Decentralized Task Offloading in Multi-Access Edge Computing Systems

Shubham Aggarwal, Muhammad Aneeq uz Zaman, Melih Bastopcu, Sennur Ulukus, Tamer Başar

TL;DR

It is verified that a higher load at the ES may lead devices to push the tasks to the ES less often, and techniques from stochastic hybrid systems (SHS) theory are invoked to study the tradeoffs between increasing information freshness and reducing power consumption.

Abstract

We consider the problem of task offloading in multi-access edge computing (MEC) systems constituting $N$ devices assisted by an edge server (ES), where the devices can split task execution between a local processor and the ES. Since the local task execution and communication with the ES both consume power, each device must judiciously choose between the two. We model the problem as a large population non-cooperative game among the $N$ devices. Since computation of an equilibrium in this scenario is difficult due to the presence of a large number of devices, we employ the mean-field game framework to reduce the finite-agent game problem to a generic user's multi-objective optimization problem, with a coupled consistency condition. By leveraging the novel age of information (AoI) metric, we invoke techniques from stochastic hybrid systems (SHS) theory and study the tradeoffs between increasing information freshness and reducing power consumption. In numerical simulations, we validate that a higher load at the ES may lead devices to upload their task to the ES less often.

Fully Decentralized Task Offloading in Multi-Access Edge Computing Systems

TL;DR

It is verified that a higher load at the ES may lead devices to push the tasks to the ES less often, and techniques from stochastic hybrid systems (SHS) theory are invoked to study the tradeoffs between increasing information freshness and reducing power consumption.

Abstract

We consider the problem of task offloading in multi-access edge computing (MEC) systems constituting devices assisted by an edge server (ES), where the devices can split task execution between a local processor and the ES. Since the local task execution and communication with the ES both consume power, each device must judiciously choose between the two. We model the problem as a large population non-cooperative game among the devices. Since computation of an equilibrium in this scenario is difficult due to the presence of a large number of devices, we employ the mean-field game framework to reduce the finite-agent game problem to a generic user's multi-objective optimization problem, with a coupled consistency condition. By leveraging the novel age of information (AoI) metric, we invoke techniques from stochastic hybrid systems (SHS) theory and study the tradeoffs between increasing information freshness and reducing power consumption. In numerical simulations, we validate that a higher load at the ES may lead devices to upload their task to the ES less often.
Paper Structure (8 sections, 2 theorems, 7 equations, 6 figures, 2 tables, 1 algorithm)

This paper contains 8 sections, 2 theorems, 7 equations, 6 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. $N$-User MEC Game Problem
  3. Age of Information Calculation
  4. Stochastic Hybrid Systems (SHS)
  5. Average AoI for the $i$th Device
  6. Mean-Field Game
  7. Numerical Results
  8. Discussion and Conclusion

Key Result

Theorem 1

yates2018age Suppose $\Bar{\pi}$ is the state distribution of the FS-MC and there exists a stationary solution $\Bar{v} := [\Bar{v}_1,\cdots, \Bar{v}_m]$ of the conditional distribution $v_{(\cdot)}(t)$ satisfying, Then, the average AoI is given by $\Delta:= \sum_{s \in S}\bar{v}_{s0}$.

Figures (6)

  • Figure 1: A MEC system model consisting of an edge server (ES) and various applications (medical, vehicular and home surveillance examples are shown in the figure) that utilize the ES for timely computation simultaneously.
  • Figure 2: Information flow schematic in a MEC system, For device $D_i$, $L_i$ and $T_i$ denote its local processor, and the transmitter, respectively.
  • Figure 3: Evolution of the AoI at the receiver.
  • Figure 4: Task flow for device $i$: $D_i$, $L_i$, $T_i$ denote the $i$th device itself, its local processor, and its transmitter, respectively.
  • Figure 5: The optimal probability $\hat{p}$ as a function of the MF term $\rho$.
  • ...and 1 more figures

Theorems & Definitions (2)

  • Theorem 1
  • Theorem 2