Broadband Communities

MAY-JUN 2016

BROADBAND COMMUNITIES is the leading source of information on digital and broadband technologies for buildings and communities. Our editorial aims to accelerate the deployment of Fiber-To-The-Home and Fiber-To-The-Premises.

Issue link: https://bbcmag.epubxp.com/i/686592

Contents of this Issue

Navigation

Page 89 of 94

MAY/JUNE 2016 | www.broadbandcommunities.com | BROADBAND COMMUNITIES | 77 Our goal was to minimize, if not eliminate, tenants' purchases of third- party Wi-Fi equipment. Te on-site survey tested several metrics: • Adequacy of coverage and data rate in a variety of unit sizes with neighboring units' GigaCenters active • Presence of outside interference sources, particularly radar, which can temporarily cause the GigaCenter to shift trafc away from the DFS channels and impacts the channel plan. • Validation of assumptions regarding construction materials and absorption. RECOMMENDATIONS On-site testing of the channel plan resulted in the following recommendations: • Disable the 802.11b modulation setting on all radios. Tis legacy modulation scheme propagates much farther than 802.11g/n and is unnecessary in consumer deployments. • Consider disabling the 2.4 GHz radios and re-enabling them if a subscriber has devices that do not support 5 GHz. Tis could be done on a trial basis during deployment of the frst building, and call volumes could be assessed to determine the impact. In the long term, monitor use of 2.4 GHz and disable it when possible. • Signal bleed between foors was higher than the model predicted, violating the requirement for absolute isolation between units at 5 GHz. As a result, the recommendation was to use a 20 MHz channel plan instead of a 40 MHz channel plan at 5 GHz. A 20 MHz plan still ofers sufcient throughput to support 100 Mbps data service but with less co-channel interference. In a fully loaded RF environment, contention between neighboring stations is a more serious issue than raw speed. • Generally, the GigaCenters in most units should be confgured at 10 percent power for both 2.4 GHz and 5 GHz bands. • Te channel plan should be designed to minimize co-channel interference between units and provide as much isolation as possible between channels. • Te channel plan should make use of dynamic frequency selection (DFS) channels, which share spectrum with military and weather radar. Should a unit detect a radar signal in the DFS band, it will select another channel for use. On-site testing with a spectrum analyzer confrmed that it was safe to use these channels in the plan. Tough the majority of client devices are capable of using DFS channels, some 5 GHz devices are not. Older non-DFS clients should be allowed to associate to the 2.4 GHz radio in a unit that happens to be assigned a DFS channel at 5 GHz. • Te channel plan was formulated to exclude channels 112–128 as well as 140, 144 and 165, which have been shown to be problematic with some client devices. • Diferent SSID names should be used at 2.4 GHz and 5 GHz so that users will prefer the less-congested 5 GHz band. Tere are two important considerations: – Te 5 GHz SSID should sort before the 2.4 GHz SSID in alphanumeric order so that iOS devices will prefer it if both SSIDs are defned. – Te 5 GHz SSID should have a more attractive name, such as "Unit138_FAST," so that users can readily distinguish SSIDs without having to know the details of 5 GHz versus 2.4 GHz. v Andrew Peterson is a principal technical consultant at Calix Consulting Services. You can reach him at CalixConsulting@ calix.com. Registration Now Open Economic Development Conference Series Radisson Blu Minneapolis Downtown Minneapolis, MN October 17 – 20, 2016 To Exhibit or Sponsor contact: Irene G. Prescott irene@bbcmag.com | 505-867-3299 877-588-1649 www.bbcmag.com

Articles in this issue

Links on this page

Archives of this issue

view archives of Broadband Communities - MAY-JUN 2016