Copper twisted pair telephone wiring could someday support terabit speeds over distances of approximately 100 meters, according to Dr. John Cioffi, chairman and CEO of broadband software development company ASSIA. Technologies underlying terabit DSL would include a waveguide approach in combination with vectoring, explained Cioffi in an interview with Telecompetitor, in which he provided details on ASSIA’s research in this area, originally outlined last year.
“We’re in the early stages of looking into an orders of magnitude improvement in copper,” explained Cioffi.
Making terabit DSL a reality would require “different sectors in the industry to work together,” Cioffi said.
Key to supporting dramatically high speeds over copper would be to use higher frequencies for data transmission – such as the millimeter wave frequencies some service providers plan to use for 5G wireless communications, Cioffi said. As with wireless communications, ultra-high frequencies can support dramatically higher data rates in comparison with lower frequencies.
“Once you get into high frequencies, even 1% of bandwidth has a very large data rate – two bits per hertz,” explained Cioffi.
Also as with wireless communications, however, the downside to using higher frequencies is the distance covered.
Cioffi believes the terabit DSL technology – based on waveguides, ultra-high frequency transmission and vectoring over twisted pair – could support speeds of 100 Gbps over 300 meters and 10 Gbps over 500 meters. The technology might support lower speeds over distances of up to 700-1,000 meters, but might not be practical beyond that distance, he said.
Asked whether terabit DSL had advanced beyond the theoretical, Cioffi noted that a scientist at Brown University already has proven the concept in tests over short distances using a single wire.
What Cioffi would like to see happen next is for waveguide experts to measure data rates – a task that requires $10 million-laboratories. Cioffi said he is trying to get the National Science Foundation to fund these measurements. Some service providers also have said they will provide some funding for terabit DSL research, he noted.
Once researchers have enough data to support their calculations, Cioffi anticipates that chip vendors and service providers will be interested in developing the technology.
“Network planners will be working for years to figure out how to offer services and not spend too much on infrastructure,” he observed.
As for commercializing the technology, Cioffi said he doesn’t expect cost to be a huge issue.
He also noted that the terabit DSL technology he envisions should have extremely low latency – lower than for traditional DSL and for some fiber – which could make it well suited for providing backhaul for 5G wireless networks, which are designed to support low-latency applications such as connected cars.