Danny Van Bruyssel

Pilot-based crosstalk channel estimation for vector-enabled VDSL systems

Recently, the International Telecommunication Union has consented the ITU-T G.993.5 recommendation on vectoring in VDSL2 technology, which specifies the required functionality for interoperability between chipsets of different vendors to achieve vectoring. This paper will briefly review the principles of vectoring and the underlying precoding and channel estimation techniques. This framework is then used to introduce mechanisms for crosstalk channel estimation that exploit channel smoothness with the aid of frequency-dependent pilot sequences, as well as mechanisms that allow fast and simultaneous initialization by targeted assignment of pilot sequences. A pilot-based technique is described that exploits the robustness of the synchronization symbols to estimate crosstalk emanating from “legacy” VDSL2 lines within a vectoring group. This allows cancellation of crosstalk from legacy lines without replacement of the installed VDSL2 customer premises equipment. Some of these techniques are compatible with the current version of ITU-T G.993.5 recommendation, while others require changes that are under consideration for possible inclusion in future amendments.

Transmitter-controlled adaptive modulation: Concepts and results

The continued growth of broadband traffic pushes wireline access network operators to continuously offer better service. Due to the large investments needed to upgrade to an all-fiber network, operators seek to boost performance over their existing copper-based infrastructure, while gradually evolving to a fiber network. As a result, Digital Subscriber Line (DSL) technology, which offers broadband over copper, has been drastically improving. The next generation of DSL will enable the continued cost-effective deployment of fiber in the access network by providing aggregate data rates up to 1 Gb/s. To keep operational costs down, this new DSL generation should enable self-installation. Therefore, it should be robust and adaptive to a wide range of time-varying channels. In this paper, we propose a novel technique based on hierarchical modulation to achieve this robustness. It adapts quickly to noise and interference dynamics in frequency and in time and allows the transmitter to select its preferred configuration without the need for an explicit communication of this configuration to the receiver. We present an overview of hierarchical modulation and how we aim to use these concepts in DSL. We further show with simulations that the receiver is able to blindly track the changes introduced by the transmitter and discuss how this protocol can be easily introduced in state-of-the-art systems.