CAREER: Developing a Flexible Serverless Multimedia Streaming Cloud Platform

随着多媒体流正在成为许多应用程序不可或缺的一部分，人们希望构建一个特殊用途的平台，使流服务能够快速发展，并以一种健壮且经济高效的方式在边缘到云的连续体中容纳它们。因此，该项目的目标是设计具有高级管理抽象的多媒体流云（MSC）平台，该平台可以获得连续体的好处，同时隐藏使用连续体的复杂性。MSC推广了函数即服务计算范式，并提供了对象即服务（OaaS）抽象作为一级公民来表示各种流源，从而减轻了用户和开发人员在资源和数据管理方面的困难。OaaS支持对象跨连续体的灵活迁移，也加速了流媒体服务的开发。保留对象的状态有助于对象的无缝迁移，因此，实现了对象在边缘到云连续体中的弹性和移动性。MSC通过在异构云资源上运行的智能任务管理和资源分配方法，为每个流对象提供经济高效和健壮的QoS。MSC平台的开源原型通过录音设备、本地边缘计算系统和公共云进行开发和测试，以评估其可用性和稳健性。MSC平台实现了多媒体流媒体的民主化，为基于多媒体流媒体的下一代应用奠定了基础。MSC将为流媒体提供商以解决方案为导向铺平道路，让他们专注于开发先进的流媒体服务，并将其集成到自己的应用程序中。该项目的成果不仅将影响科学发现的前沿，还将通过在各种应用中提供灵活而强大的流媒体服务来造福社会。MSC平台的对象抽象和智能资源分配方法可以适应和应用于其他软实时计算系统。该项目将把应用程序整合到分布式计算课程中，从而影响计算机科学和信息技术教育。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

As multimedia streaming is becoming an integral part of many applications, there is a desire to build a special-purpose platform that enables rapid development of streaming services and accommodates them in a robust and cost-efficient manner across the edge-to-cloud continuum. Accordingly, the goal of this project is to design the Multimedia Streaming Cloud (MSC) platform with high-level management abstractions that can reap the benefits of the continuum while hiding the complexity of working with it. MSC promotes the Function-as-a-Service computing paradigm and offers the Object-as-a-Service (OaaS) abstraction as a first-class citizen to represent various streaming sources, hence, relieving the users and developers from both the resource and data management difficulties. OaaS enables flexible migration of the objects across the continuum and accelerates the development of streaming services too. Preserving the object's state facilitates seamless object migration, hence, elasticity and mobility of objects across the edge-to-cloud continuum are achieved. MSC offers cost-efficient and robust QoS to each streaming object via smart task management and resource allocation methods operating on heterogeneous cloud resources. An open-source prototype of the MSC platform is developed and tested via recording devices, a local edge computing system, and a public cloud to assess its usability and robustness. The MSC platform democratizes multimedia streaming and lays the foundation for the next generation applications operating based on multimedia streaming. MSC will pave the way for stream providers to become solution-oriented and let them focus on developing advanced streaming services and integrate them to their applications. The project results will impact not only the forefront of scientific discovery, but also benefit the society by offering flexible and robust streaming services across various applications. The object abstraction and the intelligent resource allocation methods of the MSC platform can be adapted to and employed in other soft real-time computing systems. The project will impact computer science and IT education by integrating applications into the distributed computing curriculum.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.