Terabit DSLs and Wireless-Dimensionality in the Terahertz Band

Testing of waveguide mode of copper infrastructure at Brown University.

John Cioffi, ASSIA CEO, was the opening keynote speaker at PMF2019 “The First International Workshop on Polymer Microwave Fiber (PMF) Technology” hosted by KU Leuven in BELGIUM on March 4th and 5th. The workshop brought together important academia and industry players in the fields of PMF and waveguides. Polymer Microwave Fiber is a communication concept that combines mm-wave chips, metal couplers, and cheap plastic fibers.

The Promise of Waveguide Mode

In his keynote, “Terabit DSLs and Wireless-Dimensionality in the Terahertz Band”, Cioffi, gave an update on the progress of the Terabit DSL research, some measurements for which are being conducted at Brown University under the direction of Professor Daniel Mittleman under sponsorship from the US National Science Foundation and ASSIA.

Cioffi originally introduced Terabit DSL at the Paris G.fast Summit conference in May 2017. In that initial presentation, Cioffi asserted that fiber-like speeds of 10 Gbps to 1000 Gbps (e.g., 1 Tbps) could be possible by using the previously unexploited waveguide modes of current copper infrastructure cables that contain twisted-pair phone lines. Waveguide-mode use is similar to the use of millimeter-wave transmissions in advanced wireless and 5G. While current 5G wireless often runs at 28 GHz and 39 GHz, commercial microwave gear can run at 70 GHz and 90 GHz. Waveguides can open the terahertz gap and enable use of frequencies above 100 GHz for significantly faster speeds. Cioffi notes that the wavelengths at these frequencies can “fit” between the wires and that multiple-antenna-like (“MIMO”) processing can be used to transmit and receive well the signals at these frequencies.

Waveguide mode for TDSL

Waveguides for TDSL

Terabit DSL Research Update

In the detailed technical talk in Belgium, Cioffi discussed how this approach combines two known methods into a Digital Subscriber Waveguide that can deliver 1 terabyte per second:

  • Plasmon polariton and other waveguide modes used for sub millimeter wave transmission
  • Vectoring or massive MIMO used by G.fast vectoring, Wi-Fi, and LTE

Dr. Cioffi covered these additional topics in discussing the testing environment, research results, and proposals for real-world deployments.

  • Development of a MU-MIMO model of the waveguide channel based on experimental results
  • Example architecture of transmitters and receivers
  • Ethernet results, exploring the possibility of terabit “TBASE-T” Ethernet
  • Signal Processing – conversion devices and processing capabilities

Early Successes

So far, the research results indicate successful transmission modes, even around bending pairs of wireless in the 200 GHz range. While continued experiments are needed, these particular results continue to motivate further work because of the potential for home customer DSL speeds of:

  • 10 Gbps at 500-meter lengths
  • 100 Gbps at 300-meter lengths
  • 1 Tbps at 100-meter lengths

Evolutionary Adoption

Cioffi acknowledges that not everyone will need a terabit per second of DSL service anytime soon. He suggested that the industry begin by providing a slower speed at longer lengths, such as 10 Gbps up to 500 meters. Then, the infrastructure can be upgraded for 5G wireless in the future when faster speeds can be accomplished at longer lengths, such as 1 terabit per second at 100 meters for 5G wireless small cells “back/front” haul.

More generally, the industry can start taking advantage of using the waveguide modes of copper infrastructure before full deployment to homes might be seriously contemplated. For instance, data centers particularly could more easily and more immediately benefit from the flexibility the approach offers. Fiber cables can be replaced with copper cables up to 100 meters and data centers will no longer have to endure the costs of guessing how much fiber and copper is needed. Further, 5G backhaul is very costly if fiber must be deployed to many new smaller cell sites, which could be far less expensively deployed instead on existing copper to such an enlarged set of 5G cell sites.

Significant Potential Cost Benefits of TDSL

If these performance speeds are eventually verified and realized in production, TDSL could fundamentally change the entire telecommunications industry. Currently there are enormous costs associated to increase access speeds. For example, upgrading each home with fiber to the home can cost $3000 to $4000 per home. Cioffi cited that at the recent Mobile World Congress, the Deutsche Telekom CTO and CEO estimated the European 5G fiber infrastructure cost to be between 300 and 500 billion euros.

In contrast, by using the existing copper infrastructure’s unexploited waveguide modes to provide home internet and to backhaul 5G cell sites, could reduce costs of building the high-speed access networks of the future by a factor of five to ten. In fact, providing 10 Gbps at 500m could change capex planning for internet service providers and telecommunications operators significantly if a customer needs it and is willing to change providers or services for it.

Cioffi concluded his keynote by encouraging the industry to work together with pre-competitive cooperation for better measures and consequent calculations and projections. Then in the short-term, focus on 10Gbps at 100 meters, 5G small cells back and front haul, and data centers.

Read more about TDSL.