Broadband Communities

MAR-APR 2014

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/281446

Contents of this Issue

Navigation

Page 95 of 104

MARCH/APRIL 2014 | www.broadbandcommunities.com | BROADBAND COMMUNITIES | 89 Understanding Broadband Performance Factors All megabits are not created equal. Speed is only one of many criteria in selecting the right broadband technology for your application. By Andrew Aferbach and Tom Asp / CTC Technology & Energy C onsumers usually compare the performance of data connections by evaluating network speed, which is measured in bits per second and is typically discussed in units of Mbps (approximately 1,000,000 bits per second). However, this measurement can be quite deceptive because there are other important factors in specifying broadband services. For example, a 30 Mbps cable modem connection may cost a residential consumer $50 per month, while a business-grade Metro Ethernet service that delivers 10 Mbps Internet capacity can exceed $500 per month – yet the Internet delivered through the Metro Ethernet service provides better value for many types of applications. Why would a service with one-third the speed cost 10 times as much as the "faster" alternative? Te answer is that all Mbps are not created equal. Factors such as latency, availability of the connection speed and the network's Internet oversubscription rate afect the connection's overall performance. In the example above, the 10 Mbps Metro Ethernet service's total set of performance attributes provides a more robust and secure connection than a 30 Mbps cable modem. Key attributes that impact performance include • Symmetry: Cable modem and DSL services are typically asymmetrical, meaning that their upload (from user to network) and download (from network to user) speeds are diferent. Te download speed is generally greater than the upload speed by a factor of 10. Metro Ethernet services, on the other hand, are typically symmetrical, meaning that upload and download speeds are the same. For businesses that transfer large data or video fles, asymmetrical services often present bottlenecks to both internal users and external customers. A user on a typical cable modem service can download a 5 gigabyte (GB) fle in less than 10 minutes, but uploading the same fle would take more than 90 minutes – which would not be acceptable to a business creating and distributing large fles. • Oversubscription to the Internet: Because Internet service providers (ISPs) recognize that users in a given area do not all access the Internet at the same time, ISPs subscribe to only a portion of their networks' total potential demand. For example, an ISP that has 1,000 subscribers with 10 Mbps service might contract for a 100 Mbps connection rather than the maximum 10,000 Mbps Internet connection its users might require. Te ratio of a network's maximum potential demand to its contracted rates is its oversubscription ratio. In this example, the oversubscription ratio is 100:1. Cable modem and DSL providers often have a 100:1 or greater oversubscription TECHNOLOGY BBC_Mar14.indd 89 3/14/14 3:22 PM

Articles in this issue

Links on this page

Archives of this issue

view archives of Broadband Communities - MAR-APR 2014