Two decades ago, most engineers would have said it was not practical to use millimeter waves (30 GHz to 300 GHz) for mass market communications purposes. Signal attenuation was simply too great, and we could not affordably apply digital signal processing cheaply enough to overcome those obstacles.
As with many other problems, such as real-time compression of high definition TV streams, Moore’s Law changed the economics of signal processing, making possible what previously was economically impossible.
Back in the late 1980s, for example, some might have said that processors powerful enough to process and compress 45 Mbps to 100 Mbps data streams to 6 MHz of bandwidth, in real time, might be observed in random and small quantities in chip foundries, but might not have been commercially reproducible in quantities big enough to be applied in consumer mass markets.
But Moore’s Law held, processor power has continued to double about every 18 months, and mass production of the more-powerful chips has been possible.
And make no mistake, it is Moore’s Law that underpins the notion that previously-unusable millimeter frequencies in the millimeter region will be useful for communications purposes.
There are other angles. Because of physics, signals in higher frequency bands can carry more information in any unit of time than signals at lower frequencies.
Basically, the issue is frequency, the number of times any waveform passes across a “zero” point between high and low states. A waveform that crosses the zero point millions of times per second cannot carry as much information as a wave crossing the zero point billions of time.
There always is a trade-off between frequency and reach, so millimeter waves operating in the gigaHertz range generally will be used for shorter-range communications (small cells) than megaHertz range.
But the bandwidth advances will be prodigious. It is common these days to hear proponents argue that the fifth generation mobile air interface will not so much be about the air interface as it is device access to all available network access platforms, or perhaps the new applications that are seen as characteristic of 5G (Internet of Things).
Nevertheless, 5G also is likely to include new spectrum and much-faster access, much as prior generations have provided.
There is another possible implication as well. As all networks are capable of interworking, as apps move to the cloud, and then latency performance drops to perhaps a millisecond, cloud computing will be even more ubiquitous than it is today.
But paradoxically, many apps will shift to the edge of the network again, simply to match server and device latency performance requirements. But standards bodies now are working on new global standards for millimeter wave technology.
That means, among other things, a gigabit to every device, potentially. We will be discussing 5G implications of millimeter communications at Spectrum Futures in Singapore, in September 2015.