Algorithms and embedded architectures with scheduling and task synchronization for multiple source realtime data streaming

Abstract

This thesis entitled ALGORITHMS AND EMBEDDED ARCHITECTURES WITH SCHEDULING AND TASK SYNCHRONIZATION FOR MULTIPLE SOURCE REALTIME DATA STREAMING is concerned with embedded architectures for realtime where tasks need to be executed by giving priority to those tasks that are under critical condition Low level system API calls are implemented using mutual exclusiveness among tasks to accommodate critical and noncritical scenarios The work is developed in RTOS keil environment and executes concurrent task handling methods for ADC suspend resume with reduced task execution time to suit realtime using task synchronization This work also discusses central queue based earliest deadline first scheduler that provides data security and error detection techniques along with mutual exclusiveness schemes This enables effective resource sharing in applications where sampling rate is low ex speech processing but hardware speed is high EDF algorithm introduced in this work is compared with rate monotonic algorithm for relevant metrics algorithm wise analysis is done specific results are simulated and tabulated Video streaming with integrated services is implemented with queuing implementation and supports with Linux based realtime interval timers profile and virtual timers along with run time values for alarm signals newline