Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

When Computing follows Vehicles: Decentralized Mobility-Aware Resource Allocation for Edge-to-Cloud Continuum

Zeinab Nezami, Emmanouil Chaniotakis, Evangelos Pournaras

TL;DR

The paper tackles the challenge of providing ultra-low-latency computation for smart mobility by shifting from centralized cloud solutions to a mobility-aware edge-to-cloud continuum. It introduces MERA, a decentralized, cooperative resource-allocation framework that jointly optimizes QoS, provisioning costs, energy use, and sustainability using an EPOS-based multi-agent algorithm. The authors formalize a mobility-aware service placement problem with system-wide and local objectives, propose an eight-component cost model, and validate MERA on real-world traffic data from Munich, demonstrating substantial improvements in workload balance and SLA adherence while mitigating environmental impact. The results indicate MERA's potential to enhance resilience and efficiency in edge-to-cloud infrastructures for future connected mobility, with avenues for extending to other dynamic edge applications and real testbeds.

Abstract

The transformation of smart mobility is unprecedented--Autonomous, shared and electric connected vehicles, along with the urgent need to meet ambitious net-zero targets by shifting to low-carbon transport modalities result in new traffic patterns and requirements for real-time computation at large-scale, for instance, augmented reality applications. The cloud computing paradigm can neither respond to such low-latency requirements nor adapt resource allocation to such dynamic spatio-temporal service requests. This paper addresses this grand challenge by introducing a novel decentralized optimization framework for mobility-aware edge-to-cloud resource allocation, service offloading, provisioning and load-balancing. In contrast to related work, this framework comes with superior efficiency and cost-effectiveness under evaluation in real-world traffic settings and mobility datasets. This breakthrough capability of 'computing follows vehicles' proves able to reduce utilization variance by more than 40 times, while preventing service deadline violations by 14%-34%.

When Computing follows Vehicles: Decentralized Mobility-Aware Resource Allocation for Edge-to-Cloud Continuum

TL;DR

The paper tackles the challenge of providing ultra-low-latency computation for smart mobility by shifting from centralized cloud solutions to a mobility-aware edge-to-cloud continuum. It introduces MERA, a decentralized, cooperative resource-allocation framework that jointly optimizes QoS, provisioning costs, energy use, and sustainability using an EPOS-based multi-agent algorithm. The authors formalize a mobility-aware service placement problem with system-wide and local objectives, propose an eight-component cost model, and validate MERA on real-world traffic data from Munich, demonstrating substantial improvements in workload balance and SLA adherence while mitigating environmental impact. The results indicate MERA's potential to enhance resilience and efficiency in edge-to-cloud infrastructures for future connected mobility, with avenues for extending to other dynamic edge applications and real testbeds.

Abstract

The transformation of smart mobility is unprecedented--Autonomous, shared and electric connected vehicles, along with the urgent need to meet ambitious net-zero targets by shifting to low-carbon transport modalities result in new traffic patterns and requirements for real-time computation at large-scale, for instance, augmented reality applications. The cloud computing paradigm can neither respond to such low-latency requirements nor adapt resource allocation to such dynamic spatio-temporal service requests. This paper addresses this grand challenge by introducing a novel decentralized optimization framework for mobility-aware edge-to-cloud resource allocation, service offloading, provisioning and load-balancing. In contrast to related work, this framework comes with superior efficiency and cost-effectiveness under evaluation in real-world traffic settings and mobility datasets. This breakthrough capability of 'computing follows vehicles' proves able to reduce utilization variance by more than 40 times, while preventing service deadline violations by 14%-34%.
Paper Structure (32 sections, 25 equations, 11 figures, 4 tables)

This paper contains 32 sections, 25 equations, 11 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Service Placement Framework
  4. Motivational Application
  5. Distributed Architecture
  6. Handling Requests and practicality
  7. Mobility-aware Service Placement Problem Formulation
  8. System-wide Objectives
  9. Workload Balance
  10. Energy Consumption versus Workload Balance
  11. Renewable Energy Incentives
  12. Local Objective
  13. Cost of Processing and Memory Resources
  14. Cost of Service Delay
  15. Cost of Service Deployment
  16. ...and 17 more sections

Figures (11)

  • Figure 1: Layered core components of MERA framework, consisting of: (1) Real-world traffic network, (2) Real-world data layer, (3) Vehicle-fog-cloud connectivity and computation infrastructure, and (4) Cooperative service placement orchestrating resource allocation.
  • Figure 2: Rectangular test area in Munich city center, core routers connected with black lines and LTE access points within the test area are highlighted.
  • Figure 3: Traffic distribution across the edge-to-cloud network: the number of connected vehicles to each access point, along with their presence time, is unbalanced and higher in default routes compared to optimized routes.
  • Figure 4: 48 hours (192 15-minute IoT profiles) of MAWI trace data (April 12-13, 2017) serve as the input workload, delineating the experimental time intervals
  • Figure 5: The influence of traffic patterns on optimization objectives reveals distinct differences in workload distribution among placement approaches, while service provisioning costs are at a similar level.
  • ...and 6 more figures