Speaker Aakanksha Chowdhery
Affiliation Stanford University
Host Ranveer Chandra
Date recorded 10 April 2013
The explosion of Internet-enabled devices in home networks, especially video-content on these devices, threatens to choke the current broadband-access speeds (25 Mbps) feeding the Wi-Fi and small-cell wireless connections. The data-rates and service-quality in next-generation copper-access networks, such as digital subscriber lines (DSLs), can be improved by an order-of-magnitude by using an innovative approach called Dynamic Spectrum Management (DSM). DSM algorithms effectively mitigate the interference and intermittent noises on the lines by configuring them dynamically based on the operational network data collected at a Spectrum Management Center.
In this talk, I will provide a brief overview of three increasingly sophisticated levels of DSM algorithms that improve the access speeds and service quality: Level 1 DSM to tackle intermittent noises, Level 2 DSM to manage power and interference, and Level 3 DSM to cancel interference. In particular, the talk focuses on novel Level 2 DSM algorithms for next-generation DSL technology based on "vectored" DSLs that promise to deliver multi-100 Mbps speeds by using sophisticated signal processing to cancel multi-line interference. In many practical deployments, lines employing "vectored" DSLs co-exist with legacy DSL lines, owing to unbundled networks or incremental network upgrades, resulting in undesirable interference. I will present novel DSM algorithms that can maximize the data-rate gains of vectored DSLs even when they co-exist with legacy lines, especially focusing on practical approximations that are easy to implement and standardize. The proposed DSM algorithms promise gradual incremental upgrade of next-generation DSLs to multi-100Mbps/Gbps backhaul for Wi-Fi connections or small wireless cells everywhere.
©2013 Microsoft Corporation. All rights reserved.