Date of Award

6-2010

Degree Name

MS in Computer Science

Department

Computer Science

Advisor

Alexander Dekhtyar, Ph.D.

Abstract

Contemporary small office phone systems are specialized computers that connect a variety of phones within the office and to the local phone company. These systems use digital signaling processors (DSPs) to convert signals from analog to digital and vice-versa. Many different types of applications run on the DSPs and different businesses have varying application needs. Given the systems have limited amounts of DSP resources and growing numbers of applications for a phone system, an administrator needs a way to configure the uses of resources based on their individual business needs.

This thesis provides an overview of a system for configuring resources on various types of DSP hardware some of which are removable and have differing tradeoffs between application uses. The system has to be able to change resource allocations while the phone system is running with minimal interruptions to calls. The configuration system needs to be designed to be flexible enough that new applications or DSP hardware could be supported without major changes to code.

This thesis presents a system that uses a database-driven model along with algorithms that optimize configuration of DSP hardware given the administrator’s individual application needs.