Tolerance-Oriented Wi-Fi Advertisement Scheduling: A Near Optimal Study on Accumulative User Interests
Xu, Wanru,Fan, Xiaochen,Wu, Tao,Tao, Wei,Xi, Yue,Yang, Panlong,Tian, Chang
DOI: https://doi.org/10.1007/s11036-021-01849-8
2021-01-01
Mobile Networks and Applications
Abstract:While public Wi-Fi hotspots provide ubiquitous Internet access to mobile users, they can also generate revenues for venue owners and mobile advertisers by disseminating advertisements. There have been some research efforts in promoting Wi-Fi monetization from different perspectives, but most of them ignore users' tolerance towards watching advertisements. Indeed, tolerance plays a fundamental role in Wi-Fi advertising, as it directly influences the effectiveness of advertisements and the long-term profits of advertisers. To address this issue, we study Wi-Fi advertisement scheduling by considering both the tolerance of users in viewing advertisements and resource constraints of Wi-Fi hotspots for broadcasting advertisements. With the goal of maximizing the overall accumulative interests of users, we formulate a mixed integer programming problem for Wi-Fi advertisement scheduling. We prove that the objective function of this problem is monotone and submodular, subject to a knapsack constraint and a partition matroid constraint. On this basis, we solve it by proposing a novel greedy swap algorithm, which has a 1-e(-2)/2-approximation performance guarantee. Through extensive synthetic and trace-driven evaluations, we show that the proposed algorithm effectively improves users' interests towards different Wi-Fi advertisements and achieves robust results near to the optimal solution.