Development of a multi-mode self-adaptive algorithm to create an efficient wireless network on a university campus

The expanding use of ubiquitous computing has created a significant demand on existing network infrastructures. The demands of voice, video, and data on the same medium require a quality of service (QoS) at a level acceptable to users. Many network providers simply scale their networks to increase bandwidth and hardware to deal with the increasing demands. However, a network may still reach its design limits with peak traffic or malicious overuse of resources. In addition, with technology changing at a rapid pace, it is difficult to provide sufficient staffing to monitor and adjust the network settings to avoid issues during periods of network saturation. One of the common method to address these issues involves implementing a traffic shaper. A traffic shaper is a computer network management technique by which data sent across the network is delayed or routed in a way to accommodate a specific level of traffic to reach a desired QoS. There are many existing traffic shaping algorithms, each performing well under specific circumstances improving some QoS measures. The algorithms make use of queuing schemes to sort and send traffic based on the parameters provided to the system. To determine the need for this research, a survey was administered which revealed dissatisfaction with QoS of the wireless network. The purpose of this study focused on the development of a traffic shaping algorithm that would improve the QoS on a local area network on a university campus. The goal of the research was to create a new architecture that would allow a router to dynamically shift between different queuing mechanisms to improve network delay and packet loss without negatively impacting data throughput. The Multi-Mode Self-Adaptive (MMSA) algorithm was proposed to define a mechanism for this architecture. The MMSA was implemented within the code of a Cisco® router in the OPNET Modeler software and tested in a simulated university network environment. The results of the simulation revealed an improvement in end to end delay and packet loss rate with an insignificant change in average transmit rate between the router and the external server. The results of this research can be used as a basis for future research to create a new QoS framework. The new framework could be implemented in a router to allow configurations tailored to the network requirements of a service provider. DEVELOPMENT OF A MULTI-MODE SELF-ADAPTIVE ALGORITHM TO CREATE AN EFFICIENT WIRELESS NETWORK ON A UNIVERSITY CAMPUS A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree Doctor of Technology Approved: ____________________________________________ Dr. Jin Zhu, Chair ___________________________________________ Dr. Paul Gray, Co-Chair ___________________________________________ Dr. Shashi Kaparthi, Committee Member ___________________________________________ Dr. Ali Kashef, Committee Member ___________________________________________ Dr. Hong Nie, Committee Member