Satellite edge computing (SEC) is an important component for future network deployments. This article studies the task offloading in the SEC system with limited buffers (SEC–LB) and data dependencies among tasks. Due to the competition for LBs, a resource circular-waiting pattern may be formed, leading to a deadlock. To well analyze deadlocks, the Petri net model of SEC–LB is constructed for the first time. Then, a Petri-net-based modified Banker's algorithm (MBA) is proposed to determine the safety of a state, and a deadlock avoidance policy (DAP) $\rho_{\text{MBA}}$ with polynomial complexity is developed for SEC–LB. Moreover, we present an improved version of MBA, namely topology-based Banker's algorithm (TBBA), as well as its corresponding DAP $\rho_{\text{TBBA}}$. Compared with $\rho_{\text{MBA}}$, a scheme of simplifying the safety-determining is employed in $\rho_{\text{TBBA}}$. Simulation results demonstrate that $\rho_{\text{TBBA}}$ has higher efficiency and better performance than $\rho_{\text{MBA}}$ and the existing DAPs, especially for large-scale SEC–LB.
Polynomial-Complexity Deadlock Avoidance Policy for Tasks Offloading in Satellite Edge Computing With Data-Dependent Constraints and Limited Buffers
IEEE Transactions on Aerospace and Electronic Systems ; 60 , 5 ; 6822-6838
2024-10-01
3039089 byte
Article (Journal)
Electronic Resource
English
High-Performance, Low-Complexity Deadlock Avoidance for Arbitrary Topologies/Routings
BASE | 2018
|Resource Allocation and Offloading Strategy for UAV-Assisted LEO Satellite Edge Computing
DOAJ | 2023
|