DOCSIS 3.1 promises to dramatically increase data rates of the HFC infrastructure through the introduction of downstream Orthogonal Frequency Division Multiplexing (OFDM) and upstream Orthogonal Frequency Division Multiple Access (OFDMA). However, in order to realize the full potential of these powerful new modulation methods, DOCSIS 3.1 channel parameters must be dynamically adjusted to optimize for the greatest possible throughput given the time varying error performance of the downstream signal path. This paper describes an experimental software-defined profile optimizer for DOCSIS 3.1 networks. In it, we describe the theory of operation driving automated data collection, optimization analysis, and programmatic control of the DOCSIS 3.1 physical layer. To illustrate our approach and learnings to date, we'll examine the current state of the art of D3.1 systems to support profile optimization capabilities; which form the basis of an early implementation of SDN architectural principles in cable broadband networks.
展开▼
