Get started
Work

Optimale Resource-Allocatie voor Gelijktijdige Draadloze Informatie- en Vermogensoverdracht in MISO-OFDM-Systemen met Meerdere Gebruikers

Year: 2024
Type: dissertation
Abstract: Traditional wireless systems are often constrained by finite energy sources, which simultaneous wireless information and power transfer (SWIPT) promises to alleviate through the integration of wireless communication and power transfer. A key aspect of SWIPT is resource allocation, which is generally a challenging non-convex problem. This dissertation explores optimal resource allocation for multiuser multiple-input single-output SWIPT systems using OFDM in the broadcast channel. Maximizing the weighted sum data rate for multiuser SWIPT systems using OFDM and precoding is intractable due to its non-convex nature and because the computational complexity increases exponentially with the number of subcarriers. Although techniques like zero forcing precoding and OFDMA aim to simplify the problem by mitigating multiuser interference, they are not always optimal. The first contribution of this dissertation demonstrates that Lagrange decomposition, utilizing the time-sharing property of multicarrier systems, achieves a near-zero duality gap thereby reducing the computational complexity to linear with respect to the number of subcarriers. This approach eliminates the need for interference suppression for tractability and enables efficient optimization of precoding vectors and power allocation through a MAC-BC duality. This flexible method enhances the adaptability of precoding and OFDM to diverse wireless environments and power constraints. Time-switching is a prevalent technique in SWIPT receivers that balances the time spent decoding information against harvesting energy. Existing methods for optimizing the time-switching ratio do not easily accommodate complex non-convex problem formulations. This dissertation addresses this issue by analyzing the geometric relationship between the optimal time-switching ratio and the resource allocation, deriving a novel mapping to calculate the optimal ratios. It further generalizes the problem to support resource allocation based on weights and quality of service fairness through alpha-Fairness, rather than specific harvesting constraints. The study provides new insights into the interplay between the time-switching ratio, system weights, and fairness parameters. A nested algorithm is developed to optimize precoding vectors, power allocation, and time-switching ratios using this mapping. Finally, this dissertation incorporates a common non-linear power harvesting model into the relative resource allocation for the system. This integration is challenging due to the non-convex nature of the harvesting model. Alpha-Fairness utility functions are first applied to concavify the power harvesting model, demonstrating that the composite utility becomes strictly concave for fairness parameters above a certain threshold related to the harvester model. This eases the optimization and broadens the applicability of the harvesting model to a wider range of fairness parameters. The dissertation then illustrates the application of this composite utility in the context of the existing relative resource allocation framework.
Source: Lirias
Language: English
Open Access status: green
OpenAlex ID: W6995521382
Cites:
Cited by:
Related to: