I-Corps: Intelligent Multi-layer Network Optimization

I-Corps项目的更广泛影响/商业潜力是改善远距离数据传输。全球消费者互联网流量每年都在增长，需要不断提高互联网主干的效率。许多行业都受到这种改进的影响。一些应用，如智慧城市，需要在基础设施网络效率方面有相当大的提高。例如，为了实现物联网（IoT）异构应用的智慧城市，提高基础设施网络效率是必要的。此外，作为下一代移动系统的第五代（5G）无线网络也依赖于弹性高速电信骨干网。在广泛的范围内，这些对网络的改进将提高数据中心和计算平台的效率，并可以提高这些系统的能源效率。动态多层网络优化使网络能够适应流量变化和潜在的故障事件，从而消除系统资源的过度配置的成本和浪费。I-Corps项目探索电信网络优化技术的商业化道路。该方法从更高层监控流量需求，并在光协议层和互联网协议层进行多层优化。在保证网络稳定性的同时，提高了数据传输速度和网络资源利用率。这些技术不需要事先了解网络设备规范，可扩展，提供高精度，并且能够调整和了解网络上的变化。实施的可行性已经成功地在小规模原型中得到了证明，大规模的性能改进已经通过数值评估进行了广泛的研究，证明传输速度提高了2倍，设备使用量减少了10%。

The broader impact/commercial potential of this I-Corps project is to improve long distance data transmission. The global consumer internet traffic experiences annual growth requiring continuous improvement in internet backbone efficiency. A wide range of industries are impacted by such improvements. Some applications, such as smart cities, require considerable improvements in infrastructure network efficiency. For example, in order to accomplish smart cities with the heterogeneous applications on Internet of Things (IoT), improving infrastructure network efficiency is necessary. Furthermore, the fifth generation (5G) wireless network as the next generation mobile system is also reliant on a resilient high speed telecommunication backbone network. On a broad scale, these improvements to networks will increase efficiency of data centers and computing platforms and can improve the energy efficiency of such systems. Dynamic multi-layer network optimization will make networks adaptable to traffic variation and potential failure events, and consequently eliminate the costly and wasteful over provisioning of system resources.This I-Corps project explores the commercialization path for telecommunication networks optimization technology. This method monitors traffic requirements from higher layers and performs multi-layer optimization on the optical and internet protocol layers. It achieves improvements in data transmission speed and network resource usage while guaranteeing network stability. These techniques do not require prior knowledge of the network equipment specifications, are scalable, provide high accuracy and are able to adjust and learn changes on the network. The implementation feasibility has been successfully demonstrated in small-scale prototypes and the performance improvements at scale has been extensively studies by numerical evaluations proving up to 2x improvement in transmission speed and 10 percent reduction in equipment usage.