Network Resource Constrained Traffic Allocation for Delay Sensitive Mobile Crowdsourcing

Yanyan Han

doi:10.1109/TCOMM.2018.2796614

Network Resource Constrained Traffic Allocation for Delay Sensitive Mobile Crowdsourcing

2018-08-01

2

OA

AI

DOI：10.1109/TCOMM.2018.2796614

原文链接

查看原文

摘要

En 中文

In mobile opportunistic networks, a common assumption is that nodes are able to complete as much data exchange as needed during a communication opportunity, which is, however, not the case in practice due to limited wireless link bandwidth. Beyond that, the storage capacity at a node is also limited, further impacting the communication efficiency. In this paper, we explore delay sensitive and network resource constrained traffic allocation for mobile device-to-device crowdsourcing. With the requirement of restricted node storage and link bandwidth, we first formulate a non-linear traffic allocation optimization problem that would be at least as hard as NP-hard. In order to practically solve it, based on the submodular property, we propose an approximation algorithm and a distributed heuristic in the same design principle. We implement the latter on Dell Streak tablets and deploy an experiment with 21 nodes for a period of three weeks. Moreover, we extract the implementation codes from the prototype and run simulations using the Haggle trace to study its performance trend. The experiment and simulation outcomes verify that the proposed mechanisms achieve the close-to-optimal performance with affordable computation complexity, which can he easily implanted in practical network environment.

Keyword:

Delay-constrained

crowdsourcing

traffic

optimization

prototype

experiment

Network Resource Constrained Traffic Allocation for Delay Sensitive Mobile Crowdsourcing

摘要

AI总结

期刊

机构