Due to the time characteristics of audio and video data, the processing of multimedia applications has to be done using real time mechanisms. Scheduling algorithms used in these systems should be chosen to meet system requirements and have the knowledge ofthe number of tasks and their criticality. In addition it is mandatory to ensure that embedded controllers never run out of the allocated stack space, in other words a safety against stack overflow. This paper deals with the analysis performed on an embedded system to understand the Central Processing Unit (CPU) utilisation and optimisation of the system with respect to memory and time. Stack consumption ofindividual tasks are also monitored to get the dynamic behaviour of the system under different operating load conditions. The data obtained from our analysis will be extrapolated to work on solutions to improve the existing system response. The profiling of task execution times will help in identifying the areas of optimisation. The analysed data will also help us to gauge the performance of the Real Time Operating System (RTOS).