5G will give a big boost to self-driving autonomous vehicles. Intel goes so far as to say that 5G is like oxygen for autonomous vehicles. Really? Yes, I have to agree. And if it is, then millimeter wave (mmWave) frequencies are like the quarks that make up the atoms of that oxygen.
As 5G comes to life over the coming years, it will enable many things that were science fiction just a decade ago, and self-driving cars are high on that list. (Other things on the list? Smart communities, sophisticated healthcare at lower cost, smart grids, and smart factories.)
Why does 5G matter to autonomous vehicles? Broadly speaking, the automotive AI brain gets smarter when it gets more input. It needs mapping and positioning data from satellite and cloud services, it uses sophisticated sensors (lidar, radar, cameras) to see and feel the immediate area, and it needs as much situational awareness as it can get.
This is where 5G comes in. 5G isn’t just about faster cellular service. Sure, the design goals of 5G promise 10 times more throughput for users, but they also aim for 10 times more connections, 100 times more total capacity in a cell and ultra-low delay connections. Not to mention a variety of peer-to-peer (V2x) connectivity schemes. With those levels of connectivity and capacity, every truck and car will be communicating with every other vehicle in its vicinity; and pedestrians; and roadways (Left lane closed ahead? Got it); and traffic lights (that one just over the hill is turning red in 12 seconds).
Achieving the full 5G vision brings tons of challenges to be solved. Networks will need to be more responsive (that’s where NFV and network slicing come in, enabling networks that can scale up and down to instantaneous load and changing traffic patterns). The front-haul links to the cell radios and the backhaul network interconnections have to get faster. And building, managing and continuously assuring those technologies is a whole area of innovation in itself.
For me, though, the coolest stuff is going on in the air. Radio spectrum is finite. It’s just like real estate – Mark Twain said, “buy land, they’re not making it anymore.” If you and I are each trying to use our radios in the same physical space, we can’t both use the exact same frequency at the exact same time. We can share it and divide it up, but at the end of the day, like real estate, we can’t both build a house on the same plot of land.
And the 5G use cases that we’re just starting to imagine - autonomous vehicles and drones, smart cities and farms and factories, 8K resolution 360°video - will consume more than you might imagine. In the U.S. last summer, the FCC allocated about 11 GHz of spectrum. That’s more than every band of 2G/3G/4G allocated today. Way more. About 1.5 GHz for current cellular vs. 11 GHz for 5G.
This need for speed is what’s driving us into the mmWave frequencies of 29 GHz, 38 GHz and higher. At higher frequencies, there is more bandwidth available per channel. For example (oversimplifying), in a 4G LTE band at 2 GHz, a bandwidth of 0.01% of the carrier frequency yields a 20 MHz channel. Up at 38 GHz, a 100 MHz channel is achieved with only 0.003% deviation of the carrier. All sorts of extreme engineering are now going into making better use of that precious spectrum resource than we were able to in the past. The 5G radio design (cleverly dubbed 5G NR, for “New Radio”) has more sophisticated signal modulation (it can pack more bits into each kHz of frequency), and it includes arrays of antennas transmitting simultaneously to create directional beams, so that spectrum can be used concurrently in various directions within a given area.
This push to mmWave brings the cellular industry to a place it never had to deal with in 2G, 3G or 4G. And if 5G is the oxygen that self-driving cars and other futuristic applications will need to breathe, then mmWave frequencies are the quarks, or the building blocks, that will synthesize it.
For more on 5G and Spirent’s RF channel emulation solutions for mmWave, Massive MIMO and Beamforming, click here.