Despite the endless 5G hype, we still need work on mmWave modulation.
Marketers worldwide are hyping the idea that 5G networks will be deployed in 2019. Indeed, I’m guilty of contributing to this impression. When we examine the true definition of millimeter-wave (mmWave) technology, we discover that signals don’t have millimeter wavelengths until they reach 30 GHz. So, let’s take a closer look at what is going on in the industry for true mmWave communications.
With the exception of a few trials conducted by AT&T at 39 GHz, the currently deployed 5G technology is centered around the 28 GHz band—not quite 30 GHz, but close enough. When we take a step back from the marketing engine surrounding 5G and look at mmWave technology up close, we recognize that true mmWave commercial communication deployments are still on the way. Plus, numerous challenges and opportunities to fully utilize mmWave spectrum remain.
The 3GPP has focused on frequencies below 52.6 GHz for Release 15 and 16, and the consensus is that frequencies below this somewhat arbitrary threshold will use an orthogonal frequency-division multiplexing (OFDM)-based waveform (Figure 1). But what about frequencies beyond 52.6 GHz? Will the 3GPP and broader communications community take the time and effort to define a new waveform?
Just a few years ago, before Release 15 was frozen, numerous proposals for better, more efficient physical layers (PHYs) for 5G were under consideration. The rapid push to get 5G out the door and into the hands of consumers, however, largely prohibited any meaningful evaluation of these PHYs. Plus, new chipsets had to be developed for 100 MHz bandwidths. For better or worse, by sticking with OFDM, increasing the bandwidth was a straightforward and relatively fast way to deliver 5G technology to market. Thus, limited PHY innovations were incorporated into the first phase of 5G.
Millimeter wave provides us with a second chance to reevaluate the PHY and presents new opportunities for innovation. Although 52.6 GHz isn’t the exact breaking point for OFDM, the communications community agrees that OFDM is not a viable option past a certain point, whether frequency, bandwidth, or both. It’s too difficult to create components with the peak-to-average power ratio (PAPR), gain flatness, and the efficiency needed for OFDM at mmWave frequencies and extreme bandwidths. The 802.11ad specification is a good example of a mmWave communications protocol that doesn’t use OFDM; it uses a single-carrier modulation scheme instead.
Various companies have shown interest in evaluating single carrier waveforms for mmWave communications systems. Nokia and NI, for example, created a 73 GHz prototype that uses 2 GHz of instantaneous bandwidth and single carrier modulation. With a 2-channel MIMO system, a throughput rate of 14.6 Gbps was achieved in over-the-air demonstrations—roughly five times the throughput possible with the 28 GHz 5G NR specification today.
We haven’t seen whether a single carrier type modulation is best or if some other type of PHY works better than OFDM-based PHYs. The 3GPP lists waveforms above 52.6 GHz as a potential study item for Release 17. This debate creates an opportunity for PHY research to flourish and a chance to define a new, non-OFDM modulation scheme for use in the 3GPP.