Incentive Mechanisms for Big Data Analytics in the Internet of Things (IoT)
Xiaoyuan Fu, Jingyu Wang, Qi Qi and Jianxin Liao
In the Internet of Things (IoT), there are a large number of tasks of big data analytics due to high demands from heterogeneous terminals. Usually, these tasks in IoT could take advantage of stream-processing platforms, such as Storm, to satisfy real-time requirements of IoT applications. For a participated task embedded to IoT networks, it is necessary to increase its physical resources whenever the task meets a steep increase of input data. However, the resource scaling-out of tasks could introduce redundant costs for themselves in IoT networks, which causes inactive participation of resource scaling-out. This may bring a great challenge for the performance of big data analytics. In order to simulate the participatory of resource scaling-out, we consider two incentive mechanisms that are inspired by the economic methods. First, we formulate a weighted-effort total cost model that presents the topological effect for a participated task. Then, the Vickey-Clark-Grove (VCG) based incentive mechanisms are applied to stimulate the participation of tasks. We make simulations in different scenarios, and the results show that the effectiveness of the incentive mechanisms for resource scaling-out of big data analytics in IoT.
Keywords: IoT, big data analytics, stream-processing, incentive mechanism