Profile Management, and Resource Scheduling for Low Latency Communications in DOCSIS 3.1 Networks

使用有线电视基础设施的有线互联网接入提供从互联网服务提供商到最终用户的网络边缘连接。为了满足增长的数据速率需求，多个系统运营商的标准组织，即，CableLabs发布了DOCSIS 3.1的更新。随着最终用户应用的质量期望值的飞跃式增长，需要开发创新的解决方案，以适应DOCSIS 3.1网络中的各种未来应用。特别地，对于未来的应用，延迟应该在不牺牲数据通信的速度的情况下大幅减少，并且上游服务将需要与下游服务对称的速度。在这个与Vecima Networks合作的研究项目中，我们将研究和开发技术，通过利用DOCSIS 3.1的新功能，降低实现复杂性，显著提高DOCSIS网络的延迟感知服务质量。本着这种精神，我们将开发：（i）上行和下行配置文件选择方法，以提高跨层性能感知频谱效率，（ii）资源调度，使全双工传输方案能够提高上行容量，以及（iii）创新机制，以预测用户的上行带宽需求，使低延迟授权分配过程能够实现。所开发的优化方法和资源调度任务将是不断发展的下一代有线网络的关键推动因素，并将大大提高未来有线行业宽带数据通信的整体性能。从这项研究的预期结果将导致提高频谱利用率，减少延迟，并提高成本效益的DOCSIS网络。

Cable Internet access that uses the cable television infrastructure provides network edge connectivity from the Internet service provider to an end user. In order to meet the growing data rate demand, the standard organization of multiple system operators, i.e., CableLabs, has released an update of DOCSIS referred to as DOCSIS 3.1. As the quality-of-expectation of the end users' applications is increasing by leaps and bounds, innovative solutions need to be developed in order to accommodate various future applications in the DOCSIS 3.1 networks. Particularly, for the future applications, latency should be substantially reduced without sacrificing the speed of data communication, and the upstream services will require symmetric speed as the downstream services. In this collaborative research project with Vecima Networks, we will research and develop technologies that can improve the latency-aware quality-of-service of DOCSIS networks significantly by exploiting the new features of DOCSIS 3.1 with reduced implementation complexities. With this spirit, we will develop: (i) upstream and downstream profile selection methods to improve the cross-layer performance aware spectral efficiency, (ii) resource scheduling enabling full duplex transmission schemes to enhance the upstream capacity, and (iii) innovative mechanisms to predict the users' upstream bandwidth demand enabling low-latency grant allocation procedure. The developed optimization methods and resource scheduling tasks will be the key enablers for the evolving future generation of cable networks, and will substantially improve the overall performance of the broadband data communications of cable industry in the future. The results anticipated from this research will lead to the improved spectrum utilization, reduced latency, and enhanced cost-efficiency in the DOCSIS networks.