NYU emulator advances 5G technology towards reality
I recently heard about an amazing piece of tech wizardry by members of CATT and NYU WIRELESS at the Tandon School of Engineering at NYU: a prototype that can emulate the wireless channel and as a bonus, multi-antenna front ends as well.
The members of this NYU team, Dr. Aditya Dhananjay, post-doctoral research fellow, along with faculty members Sundeep Rangan and Dennis Shasha, have been able to achieve what no one in the industry has been able to do to date. Some industry experts have considered a 5G emulator, but the cost would be prohibitive, especially due to the unwieldy complexity of hundreds of mmWave antenna arrays with phased-array beamforming that require a multitude of cable runs to connect phased array antennas to the channel emulator. Complicating matters further is the fact that it is nearly impossible to connect phased array antennas to cables in the first place.
Channel emulators historically have been the gold standard for testing and designing wireless systems. I spoke to Dr. Dhananjay to see how they had been able to accomplish this amazing technical feat—completely in software. The only hardware required will be a chassis made by National Instruments commercial-off-the-shelf Software Defined Radio (SDR) hardware/software, containing ADCs, DACs, FPGAs, and so on. The NI system provides tremendous flexibility and baseband processing power to enable developers to push the envelope of mmWave system prototyping.
Image courtesy of NYU Wireless, Tandon School of Engineering
Dhananjay commented that emulation has been used for the past 15 years as a standard tool in the design and testing of wireless devices, whether it be Wi-Fi, 3G, or 4G, channel emulation has been absolutely essential. Field trials are very time-consuming and you are able to only test a small fraction of scenarios within which you want your devices to work. The whole point of the emulator, says Dhananjay, is that instead of the transmitter sending a signal into the air and the receiver picking up that signal from the air, you can cable the transmitter to an emulator. So, instead of the signal going out of the antenna, they go out over cables from the transmitter to the emulator, then the emulator modifies the signals as if they had gone over a programmable channel.
Then you feed the resulting signals back to the receiver again using cables. The researchers are able to sit in their lab and program the emulator to emulate the scenario where, say, the transmitter is the side of a railway track and the receiver being inside of the moving train at 200 mph to see how the transmitter and receiver function in such a challenging propagation environment.
The 5G standardization process is moving along at a break-neck speed, and a part of this process is the acceptance of channel models, which are essentially mathematical representations of the signal propagation. Those mathematical models have been already put into the standard; we have a statistical understanding of how these signals are expected to propagate through the air. This is why now is the right time to be using the NYU emulator. Companies and researchers in academia will be testing out their 5G designs and it makes sense to do that over a standardized emulated channel since in a matter of minutes one can emulate literally multiple thousands of wireless propagation scenarios.