Data aggregation and dissemination in emerging communication networks

Al-habob, Ahmed A. (2022) Data aggregation and dissemination in emerging communication networks. Doctoral (PhD) thesis, Memorial University of Newfoundland.

[English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (2MB)

Abstract

The emerging communication networks generate a huge amount of data that need to be aggregated/disseminated, processed, and responded to in a very short time. Major challenges associate with the need to handle such tremendous amount of data, including high energy consumption, larger delay, and constrained computation capabilities. Consequently, more efficient frameworks could be exploited for data aggregation, dissemination, and processing. This work aims to design energy-efficient and age-optimum frameworks for data aggregation/dissemination. Moreover, reliable and low latency offloading frameworks for sequential and parallel mobile edge computing (MEC) offloading are also developed. A device-role assignment framework is designed to optimize the role of each device in the network and enable in-network data processing. More sophisticated scenarios with mobile data aggregator/disseminator are explored as well. Mobile data aggregator(s)/disseminator(s) for terrestrial and underwater scenarios are considered. Different metrics are studied including the overall energy consumption in the data aggregation/ dissemination systems, age-of-information (AoI) in the data aggregation systems, and the latency and offloading error in the MEC systems. A novel metric referred to as the correlation-aware AoI is also proposed to captures both the freshness and diversity in the aggregated data. Computationally efficient solution approaches are developed to find solutions for the proposed frameworks, including genetic algorithms, ant colony optimization, conflict graphs, and deep reinforcement learning agents. To show the effectiveness of the developed solution approaches, their performance is compared to baseline approaches. Extensive simulations show that the proposed solution approaches provide performance close to the optimal solutions, which are obtained through exhaustive search or computationally intensive methods.

Actions (login required)

View Item