Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Computation Pre-Offloading for MEC-Enabled Vehicular Networks via Trajectory Prediction

Ting Zhang, Bo Yang, Zhiwen Yu, Xuelin Cao, George C. Alexandropoulos, Yan Zhang, Chau Yuen

TL;DR

This work tackles the challenge of efficient task offloading in highly dynamic MEC-enabled vehicular networks by predicting future vehicle trajectories and performing proactive offloading decisions. It introduces the TPPD framework, which first predicts next-position coordinates with a two-layer LSTM and then uses a DDQN-based policy to allocate tasks to the most suitable MEC servers while respecting resource constraints. The approach demonstrably reduces task processing delay and improves resource utilization compared with traditional real-time offloading strategies, as shown by trajectory prediction accuracy metrics and system-level comparisons. The framework’s combination of trajectory forecasting, MEC server filtering, and priority-aware DDQN offloading offers a practical mechanism for reducing signaling overhead and latency in mobile edge computing for connected vehicles.

Abstract

Task offloading is of paramount importance to efficiently orchestrate vehicular wireless networks, necessitating the availability of information regarding the current network status and computational resources. However, due to the mobility of the vehicles and the limited computational resources for performing task offloading in near-real-time, such schemes may require high latency, thus, become even infeasible. To address this issue, in this paper, we present a Trajectory Prediction-based Pre-offloading Decision (TPPD) algorithm for analyzing the historical trajectories of vehicles to predict their future coordinates, thereby allowing for computational resource allocation in advance. We first utilize the Long Short-Term Memory (LSTM) network model to predict each vehicle's movement trajectory. Then, based on the task requirements and the predicted trajectories, we devise a dynamic resource allocation algorithm using a Double Deep Q-Network (DDQN) that enables the edge server to minimize task processing delay, while ensuring effective utilization of the available computational resources. Our simulation results verify the effectiveness of the proposed approach, showcasing that, as compared with traditional real-time task offloading strategies, the proposed TPPD algorithm significantly reduces task processing delay while improving resource utilization.

Computation Pre-Offloading for MEC-Enabled Vehicular Networks via Trajectory Prediction

TL;DR

This work tackles the challenge of efficient task offloading in highly dynamic MEC-enabled vehicular networks by predicting future vehicle trajectories and performing proactive offloading decisions. It introduces the TPPD framework, which first predicts next-position coordinates with a two-layer LSTM and then uses a DDQN-based policy to allocate tasks to the most suitable MEC servers while respecting resource constraints. The approach demonstrably reduces task processing delay and improves resource utilization compared with traditional real-time offloading strategies, as shown by trajectory prediction accuracy metrics and system-level comparisons. The framework’s combination of trajectory forecasting, MEC server filtering, and priority-aware DDQN offloading offers a practical mechanism for reducing signaling overhead and latency in mobile edge computing for connected vehicles.

Abstract

Task offloading is of paramount importance to efficiently orchestrate vehicular wireless networks, necessitating the availability of information regarding the current network status and computational resources. However, due to the mobility of the vehicles and the limited computational resources for performing task offloading in near-real-time, such schemes may require high latency, thus, become even infeasible. To address this issue, in this paper, we present a Trajectory Prediction-based Pre-offloading Decision (TPPD) algorithm for analyzing the historical trajectories of vehicles to predict their future coordinates, thereby allowing for computational resource allocation in advance. We first utilize the Long Short-Term Memory (LSTM) network model to predict each vehicle's movement trajectory. Then, based on the task requirements and the predicted trajectories, we devise a dynamic resource allocation algorithm using a Double Deep Q-Network (DDQN) that enables the edge server to minimize task processing delay, while ensuring effective utilization of the available computational resources. Our simulation results verify the effectiveness of the proposed approach, showcasing that, as compared with traditional real-time task offloading strategies, the proposed TPPD algorithm significantly reduces task processing delay while improving resource utilization.
Paper Structure (15 sections, 19 equations, 6 figures, 1 table)

This paper contains 15 sections, 19 equations, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. System Model
  3. Network Scenario
  4. Local Computation Model
  5. Communications Model
  6. Computation Offloading Model
  7. Problem Formulation
  8. Proposed PRE-TASK OFFLOADING Approach
  9. Vehicle Trajectory Prediction
  10. MEC Server Filtering
  11. Pre-Offloading Strategy
  12. Experimental and Discussion
  13. Trajectory Prediction
  14. Performance of the TPPD Algorithm
  15. Conclusion

Figures (6)

  • Figure 1: The considered scenario of multiple network-connected vehicles uploading environmental information to the network and performing task offloading in nearby MEC servers.
  • Figure 2: The proposed TPPD algorithm architecture.
  • Figure 3: Prediction results of the two-layer LSTM network model.
  • Figure 4: Loss of TPPD algorithm
  • Figure 5: Comparison of overall time consumption of various algorithms
  • ...and 1 more figures