Document Type

Theses, Ph.D


This item is available under a Creative Commons License for non-commercial use only


Communication engineering and systems

Publication Details

Successfully submitted for the award of Doctor of Philosophy (Ph.D) to the Dublin Institute of Technology, 2013.


Packet aggregation algorithms are used to improve the throughput performance by combining a number of packets into a single transmission unit in order to reduce the overhead associated with each transmission within a packet-based communications network. However, the throughput improvement is also accompanied by a delay increase. The biggest drawback of a significant number of the proposed packet aggregation algorithms is that they tend to only optimize a single metric, i.e. either to maximize throughput or to minimize delay. They do not permit an optimal trade-off between maximizing throughput and minimizing delay. Therefore, these algorithms cannot achieve the optimal network performance for mixed traffic loads containing a number of different types of applications which may have very different network performance requirements. In this thesis an adaptive packet aggregation algorithm called the Adaptive Aggregation Mechanism (AAM) is proposed which achieves an aggregation trade-off in terms of realizing the largest average throughput with the smallest average delay compared to a number of other popular aggregation algorithms under saturation conditions in wireless networks. The AAM algorithm is the first packet aggregation algorithm that employs an adaptive selection window mechanism where the selection window size is adaptively adjusted in order to respond to the varying nature of both the packet size and packet rate. This algorithm is essentially a feedback control system incorporating a hybrid selection strategy for selecting the packets. Simulation results demonstrate that the proposed algorithm can (a) achieve a large number of sub-packets per aggregate packet for a given delay and (b) significantly improve the performance in terms of the aggregation trade-off for different traffic loads. Furthermore, the AAM algorithm is a robust algorithm as it can significantly improve the performance in terms of the average throughput in error-prone wireless networks.