CAVE: Crowdsourcing Passing-By Vehicles for Reliable In-Vehicle Edge Computing

TL;DR

This work addresses accelerating ego-vehicle task computing by crowdsourcing passing-by vehicles in a dynamic edge environment. It formulates an end-to-end latency minimization problem with reliability constraints and introduces the CAVE framework, which decomposes the problem into task assignment (solved via adaptive barrier PSO) and resource allocation (solved via KKT) while accounting for time-correlation through in-progress tasks. Empirical results from end-to-end simulations show that CAVE reduces latency and improves reliability and scalability compared with Baseline and FPSO-MR, at the cost of increased redundancy when stricter reliability is required. The approach enables reliable in-vehicle edge computing under limited network controllability, enabling practical deployment in real-world driving scenarios.

Abstract

In-vehicle edge computing is a much anticipated paradigm to serve ever-increasing computation demands originated from the ego vehicle, such as passenger entertainments. In this paper, we explore the unique idea of crowdsourcing passing-by vehicles to augment computing of the ego vehicle. The challenges lie in the high dynamics of passing-by vehicles, time-correlated task computation, and the stringent requirement of computing reliability for individual user tasks. To this end, we formulate an optimization problem to minimize the end-to-end latency by optimizing the task assignment and resource allocation of user tasks. To address the complex problem, we propose a new algorithm (named CAVE) with multiple key designs. We build an end-to-end network and compute simulator and conduct extensive simulation to evaluate the performance of the proposed algorithm.

Figures (7)

• Figure 1: Overview of in-vehicle edge computing.
• Figure 2: CDF of task latency under different methods.
• Figure 3: CDF of task failure probability under different methods.
• Figure 4: Number of redundant vehicles per task under different methods.
• Figure 5: Average task latency under various traffic.