Design Issues and Performance Analysis for Opportunitistic Scheduling Algorithms in Wireless Networks
MetadataVis full innførsel
This doctoral thesis is a collection of six papers preceeded by an introduction. All the papers are related to design issues and performance analysis for opportunistic scheduling algorithms in cellular networks. Opportunistic scheduling algorithms can provide higher throughput and increased quality-of-service (QoS) in wireless networks by giving priority to the users with favorable channel conditions. Such algorithms are already implemented in equipment based on wireless LAN standards, the HDR standard, the HSDPA standard, and the Mobile WiMAX standard, but are often not a part of the standard itself. The implemented algorithms are often based on intuition rather than theoretical investigations, and consequently, it is a need for a better understanding of the theoretical limits for how well such algorithms can perform and how such algorithms can be implemented in the most efficient way. The design issues handled in this thesis are related to feedback algorithms and scheduling algorithms for increased throughput guarantees. Channel quality estimation and feedback is the basis for opportunistic scheduling, and two novel feedback algorithms are proposed to reduce the overhead from channel quality feedback. The results show that the feedback can be reduced to only obtaining feedback from the user that the system wants to schedule. An adaptive scheduling algorithm for obtaining increased throughput guarantees is also developed. Results from simulations show that this algorithm can double the throughput guarantees in modern cellular networks compared to other well-known scheduling algorithms. The performance of opportunistic scheduling algorithms is analyzed through analytical expressions and simulations of feedback delay, fairness and throughput guarantees. It is shown how delayed feedback can lead to reduced throughput or increased bit error rate in a system with opportunistic scheduling. Closed-form expressions are also found for two types of fairness, and throughput guarantees of different well-known scheduling algorithms.