RF and microwave design engineers inhabit a mysterious world where circuits that “should” work don’t, where simulation tools get you just so far, and where signal propagation varies with the time of day, the weather, and yes, the phase of the moon. This year, the vagaries of the ether will test the best minds in the business of “fields and waves,” as applications will need their talents more than ever. There’s much to be done, a limited time to do it, and none of it will be easy.
The fifth generation of cellular, 5G, is a terrific example because it’s the most wide-ranging change since wireless visionary Martin Cooper made the first cellular call using a handheld phone in 1973 (Figure 1). Since then, cellular networks have moved up in frequency from UHF to microwave, but 5G will take them more than an order of magnitude higher, to millimeter wavelengths, where almost everything about communications becomes more difficult and more expensive.
Figure 1: Dr. Martin Cooper poses with the 1973 DynaTAC prototype, the first cell phone. (Source: Rico Shen/CC BY-SA 3.0)
A signal “up there” travels only as far as the first thing it meets, whether it’s a tree, sign post, or building. After that, the signal is scattered and thus severely reduced in strength as to be almost undetectable. But this is where cellular is headed, as little spectrum remains at lower frequencies to accommodate the huge bandwidths required to transfer the zettabytes of data generated by streaming video, gaming, and virtual reality. To make millimeter-wave networks a reality, designers will need all available tools, from higher-order modulation techniques to multi-user “massive” MIMO, and phased array antennas, currently the exclusive domain of defense radar systems.
They will also have to avail themselves of the few semiconductor technologies that work at these frequencies, primarily silicon germanium (SiGe) and RF CMOS, both of which remain to some degree works in progress. That is not just in the base stations; smartphone designers have the onerous task of incorporating massive MIMO in a pocket-sized device, along with even more frequency bands and the seemingly impossible problem of keeping a user’s hand from absorbing a millimeter wave signal.
Speaking of works in progress, there’s the IoT, to which RF and microwave technology is linked at the hip. The challenge here is not high frequencies, higher data rates, or obscure technologies; just the opposite. IoT networks operate at much lower frequencies, minimize data rates and channel bandwidths, and use existing technologies to create dirt cheap “edge devices” the size of a matchbook that can operate for up to 10 years on a coin cell battery.
This might be less a technical challenge than operating at frequencies in the wilds of the millimeter-wave frontier, but it’s not trivial. The designer’s tasks are compounded by the fact that too many competing solutions exist for connecting edge devices, and to no one’s surprise, those are mostly incompatible. Regardless of the connectivity solution, such as Bluetooth 5, Zigbee, or the other dozen competitors, a mesh network will be required that lets them communicate with each other.
The data that devices generate must be transferred to a gateway where the data is aggregated, minimally processed to reduce its bulk, and sent onward to the Internet via either cellular networks or Low-Power Wide Area Networks (LPWANs). This is difficult enough in a modest home automation system, but throughout a smart city, 5,000-acre farm, or manufacturing facility, it becomes a massive challenge.
We also can’t fail to mention the frenetic pace of autonomous vehicle development that combines the aforementioned challenges with mobility and the need to provide extraordinarily precise, near-real-time vehicle situational awareness. Vehicles need to communicate with each other and with external sensors such as cameras mounted on “street furniture,” ranging from street lights to traffic signals. As governments, automakers, and device manufacturers are eager to make driverless cars a reality, a communication solution that all stakeholders can agree on must be found, and soon.
Ideally, this would be a simple process: Allocate frequencies, wireless access methods, and other communication specifications, and move forward. Reality being far less than ideal, two camps are now duking it out to reign supreme. The government has its 18-year-old commitment to Dedicated Short-Range Communications (DSRC) for which spectrum at 5.9GHz was allocated in 1999. It requires a dedicated communications infrastructure and “roadside units,” neither of which currently exists. The cellular industry has ubiquitous infrastructure and advanced data communications already available, and considers DSRC archaic. The cellular industry will probably win this war, but not easily.
IoT, 5G, and autonomous vehicles share another immense communications challenge as well: Reducing latency to near the vanishing point. Latency is the round-trip elapsed time between when a signal is sent and when it arrives at the desired destination. This isn’t simply a matter of advancing the state of the art; it pits designers against the immutable laws of physics. That is, latency over a given path can be reduced only so much, which is fundamentally dictated by the distance between the two points. This means that to achieve the 1ms latency proscribed by 5G and needed by some IoT applications, the communication path must be very short, necessitating the use of enormous numbers of small base stations (small cells). That will be extremely expensive and exceedingly complicated.
These are just some of the problems that the engineering, scientific, and academic communities must solve for 5G and IoT to meet their fantastic promises, and for autonomous vehicles to function in an unpredictable world while keeping us safe. There are many others in which RF and microwave technology play a less prominent but still important role. In short, big technological gaps exist that must be filled, and the stopwatch is running.
Source : Mouser Electronics
