Collaborative Pipeline Using Opportunistic Mobile Resources via D2D for Computation-Intensive Tasks

TL;DR

This paper introduces a mobile pipeline computing system utilizing D2D communication with mmWave technology to perform real-time, computation-intensive tasks on moving devices, addressing limitations of traditional cloud computing.

Contribution

It proposes a novel pipeline computing framework with a pathfinding algorithm and optimization techniques tailored for resource-limited mobile devices in D2D networks.

Findings

Feasibility demonstrated through simulations and experiments.

Pipeline stability affected by mmWave blockage analyzed.

Graph trimming technique improves algorithm efficiency.

Abstract

This paper proposes a mobile pipeline computing concept in a Device-to-Device (D2D) communication setup and studies related issues, where D2D is likely based on millimeter-wave (mmWave) in the 5G mobile communication. The proposed opportunistic system employs a cluster of pipelined resource-limited devices on the move to handle real-time on-site computation-intensive tasks for which current cloud computing technology may not be suitable. The feasibility of such a system can be anticipated as high-speed and low-latency wireless technologies get mature. We present a system model by defining the architecture, basic functions, processes at both system-level and pipeline device level. A pipeline pathfinding algorithm along with a multi-task optimization framework is developed. To minimize the search space since the algorithm may need to be run on resource-limited mobile devices, an adjacency-matrix-power-based graph trimming technique is proposed and validated using simulation. A preliminary feasibility assessment of our proposed techniques is performed using experiments and computer simulation. As part of the feasibility assessment, the impact of mmWave blockage on the pipeline stability is analyzed and examined for both single-pipeline and concurrent-multiple-pipeline scenarios. Our design and analysis results provide certain insight to guide system design and lay a foundation for further work in this line.