The mobile industry has come a long way since I joined in 1986. It would have been difficult for even the most optimistic person back then to imagine how far the industry has come.
First generation cellular combined an analog radio with a switch, and the voice market began to move around without wires. Rapid growth led to quality issues, and customers began to experience “Snap, Crackle, and Drop”. In response, the industry transitioned to Digital with 2G, where two camps of industry players argued the merits of FDMA vs. TDMA. Late to the party, Qualcomm cleverly contributed CDMA, and the industry transitioned to digital using CDMA and a few flavors of TDMA. As the industry continued is hypergrowth, 3G attempted to harmonize to a single CDMA voice standard, but ironically we still ended up with two different CDMA approaches – CDMA 2000 and WCDMA.
When the Internet Protocol (IP) emerged as the dominant network architecture for the rapidly growing data market, Flash-OFDM (later called LTE and 4G) was invented to allow the cellular networks to mobilize the internet. The takeoff of the iPhone and smartphones fast exposed the limitations of 3G. To enable the massive demands of the mobile internet, 4G/LTE was rapidly deployed to enable harmonized networks on a global scale.
Continued demands for faster mobile networks, enterprise automation and consumer applications which require very low latency, and billions of connected devices and sensors, are driving us to 5G. So what is 5G?
5G takes a 4G radio airlink (with some subtle changes) and tries to “reuse it” many times in the same geographic area, thereby creating far greater capacity from the same spectrum. This is accomplished with “beam forming”, which the industry calls Massive Mu-MIMO (Multi-user MIMO). It’s very similar to what the Fiber industry did when they created WDM, allowing a single fiber to be used over and over by separating wavelengths, or “colors”. This created tremendous leverage as a single fiber strand could function like many, many strands. Fiber optic capacity increased dramatically without needing to redeploy new cables in new trenches. That same leverage can be delivered to spectrum through smart antennas, or Massive Mu-MIMO. The concept is simple, but the deployment issues are complex in today’s “Time/Frequency” model.
Enter Cohere’s “Delay/Doppler” model. Years ago, Cohere’s founders realized that the Delay/Doppler model could deliver on the promise of Massive Mu-MIMO without the problems that are currently surfacing as operators try to deploy 5G. Cohere’s approach provides an accurate channel estimate and prediction, which are the essential ingredients for beam forming. That channel estimate ages slowly and gracefully, and it is fully independent of the modulation. It allows Massive Mu-MIMO on both TDD and FDD, meaning it can deliver many of the benefits of 5G on 4G networks as well. And perhaps most importantly, it enables an overall network architecture that is consistent with the evolution of the cloud. Embracing a Cloud-based architecture is the only way we will achieve the goals for cost-per-bit that we need as an industry to continue on our growth curve.
These are exciting times for our industry. Another major quantum step forward is now possible. Operators can afford to deploy 5G, providing far greater capacity, far higher peak rates, and opening the future to new applications that can all coexist on a single channel using the “Delay/Doppler” approach.
GSMA is at the forefront in this discussion and we look forward to working with the industry as we address 5G network challenges. This is going to be fun!
Click here to view Ray Dolan’s interview with GMSA
– Ray Dolan, Chairman and CEO of Cohere Technologies
Cohere Technologies is headquartered in Santa Clara, California and was founded in 2009. Follow Cohere and OTFS on Twitter @CohereOTFS www.cohere-technologies.com.