Crowdsourcing Driven Truthful Data Collection Methods for Wireless Nodes in Enterprise Networks

Driven by cloud-managed resource allocation solutions, enterprise wireless networks are undergoing profound changes. Moreover, for the sixth generation (6G) of wireless networks, machine learning- (ML) -driven algorithms have shown to bring intelligence to the networks by enabling a shift from reactive-/incident-driven network operations to proactive/data-driven network operations. The potential benefits of leveraging ML techniques for resource allocation in future enterprise wireless networks comes with a set of challenges. For an ML technique, network key performance indicator (KPI) data works as a fuel that powers the ML algorithm. ML models learn the characteristics of a system using the network data. Collecting accurate and reliable wireless resource utilization KPI data, such as wireless channel utilization (CU), through crowdsourcing from numerous independent entities operating in an enterprise wireless network is a challenging task.

In this thesis, to address the challenge of reliable and accurate data collection from strategic wireless agents deployed in enterprise networks, truthful reporting methods and game-theoretic mechanisms have been introduced to incentivize agents in crowd-sourced networks. Typically, enterprise networks deployed in universities, hospitals, offices and residential buildings exhibit recurring patterns which can be exploited by agents (wireless access points) to devise non-truthful reporting strategies.

Our work focuses on challenging scenarios where independent access points (APs) can utilize distribution-aware strategies to obtain higher reward payments while minimizing their cost of measurements. We design truthful reporting methods that utilize logarithmic and quadratic scoring rules for reward payments to the APs. We show that when measurement computation costs are considered, then under certain scenarios, these scoring rules no longer ensure incentive compatibility. To address this, we present a novel reward function which incorporates a distribution-aware penalty cost that charges APs for distorting reports based on recurring patterns. To address the challenge of reliable data collection against selfish deviations, we also propose a game-theoretic crowdsourcing-based wireless data collection mechanism that can be used to collect the reliable data when the APs are strategic. Unlike other game-theoretic works, it does not makes an assumption that the actions of the crowdsourcing agents are known to the crowdsourcing entity. We also propose a metric that can capture the value of the data collected from APs. In the thesis, along with synthetic data, we also use real wireless data sets which were crowd sourced using multiple independent measuring/reporting devices deployed by us in the University of Oulu.