Connecting multiple and heterogeneous hardware devices to solve problems raises some challenges, especially in terms of interoperability and communications management. A distributed solution may offer many advantages, like easy use of dispersed resources in a network and potential increase in processing power and data transfer speed. However, integrating devices from different architectures might not be an easy task. This work deals with the synchronization of heterogeneous and distributed hardware devices. For this purpose, a loosely coupled computing platform named Virtual Bus is presented as main contribution of this work. In order to provide interoperability with legacy systems, the IEEE 1516 standard (denoted HLA - High Level Architecture) is used. As proof of concept, Virtual Bus was used to integrate three different computing architectures, a multi-core CPU, a GPU and a board with an Altera FPGA and an ARM processor, which execute a remote image processing application that requires a communication between the devices. All components are managed by Virtual Bus. This proposal simplify the coding efforts to integrate heterogeneous distributed devices and results demonstrated the successful data exchanging and synchronization among all devices, proving its feasibility.

中文翻译：

---

分布式异构计算系统集成环境

连接多个异构硬件设备以解决问题带来了一些挑战，特别是在互操作性和通信管理方面。分布式解决方案可能具有许多优势，例如易于使用网络中分散的资源以及潜在地提高处理能力和数据传输速度。但是，集成来自不同体系结构的设备可能并非易事。这项工作涉及异构和分布式硬件设备的同步。为此，提出了一个名为Virtual Bus的松耦合计算平台，作为这项工作的主要贡献。为了提供与旧系统的互操作性，使用了IEEE 1516标准（表示为HLA-高级体系结构）。作为概念验证，Virtual Bus用于集成三种不同的计算架构，多核CPU，GPU和带有Altera FPGA和ARM处理器的电路板，它们执行需要设备之间通信的远程图像处理应用程序。所有组件均由Virtual Bus管理。该提议简化了集成异构分布式设备的编码工作，结果证明了所有设备之间成功的数据交换和同步，证明了其可行性。