Evaluating the Performance of the AES70/AES 67-based Network Architectures for Audio Streaming

Abstract

The consumption of multimedia services via streaming has become a core aspect of modern life. Such a role brings enormous challenges to ensure general satisfaction with the quality of those services, which naturally tends to accelerate their development, i.e., it reflects in a tendency for successive improvements. Their quality, for an user, is measured quite elementally: the service must not present any discernible latency and must deliver the highest possible fidelity. These parameters are precisely the ones this project intends to evaluate in the performance of the AES70 protocol and the AES67 standard, thus constituting the project's main objective. This evaluation is contextualized within the available options for audio transmission over an Ethernet local area network, especially when applied to embedded microcontrollers. As such, the work that comprises this dissertation was initially concerned with mastering the testing environment, i.e., the controller provider for that role, in this case, an STM board with an ARM Cortex-M7 microcontroller. During this process, the analysis of the controller's idiosyncrasies was the main object of focus, contributing to the uncovering of important shortcomings which further challenged not only the evaluation process of these protocols, but particularly, their implementation in the embedded system. However, an implementation of an audio streaming over Ethernet application was accomplished and it is expected to be useful for future work. This is, in fact, the most valuable factor of this dissertation: it incorporates a precise explanation for understanding all key aspects of the STM controller, namely those most relevant to audio and Ethernet applications, rendering it ready for test scenarios and, subsequently, for the comparison of results with similar audio streaming approaches.