Designing to the Limit with Greater Efficiency
By Eric Black, Ph.D., CTO
The FCC has established limits on mmWave infrastructure that cap the allowable Effective Isotropic Radiated Power (EIRP) at 74 dBm. EIRP is the key figure of merit in a communications system for determining downlink performance. It makes sense then, that any vendor of mmWave infrastructure gear (meaning base stations or 5G gNBs), would design as close to this limit as feasible. This prevents a competitor from building a product with superior downlink performance.
Ever wonder why the commercially deployed gNBs in the US don’t offer more than 62 dBm EIRP? 12dBm is a significant gap, representing more than an order of magnitude difference between limits and actual products. EIRP is a critical metric that combines radiated power and antenna gain. Pivotal’s own second generation repeater added 5 dBm to increase range and power through windows.
Not Green
The 62 dBm line is rooted in the silicon-based beamforming integrated circuits (ICs) so common in mmWave today. There are roughly seven major semiconductor firms building nearly the same chip that pushes silicon to its high frequency limits. Combined, these ICs radiate one Watt of RF power, but consume over 100 Watts of input power. This efficiency of less than 1% stems from more than silicon’s struggle relative to more complex semiconductors to generate power at mmWave. More importantly, the individual power amplifiers behind each radiating element must be backed off from their nonlinear compression points (P1dB) by at least 8 dB to avoid signal distortion in large QAM constellations.
One way to achieve higher EIRP is to add more beamforming ICs, as Nokia, Ericsson, and ZTE did at Mobile World Congress in February 2024. These high-power demonstrations resulted in EIRPs higher than 65 dBm but did not address the efficiency issue. If it takes 100 Watts of input power to source one Watt of mmWave signal, it will take ten times that to improve by 10 dB. One kilowatt of power per gNB sector at mmWave is not a viable product direction.
Uncle Sam Says No
These vendors could switch from silicon to other materials, like GaAs or GaN, but this runs into a problem with defense-related regulations. The United States Munitions List defines any non-silicon-based, high power phased array technology as being subject to both ITAR and export controls, making this approach extremely unattractive for commercial applications.
Efficient Alternative
Pivotal uses Holographic Beam Forming® (HBF) instead of phased arrays. In HBF, the beamforming functionality is fully divorced from the power amplifier (PA), which allows the HBF to use more efficient GaAs and GaN power amplifiers. More importantly, the single-PA approach for HBF allows the use of digital pre-distortion. This way, the same PA can operate much closer to its P1dB point, radically increasing efficiency. Taking overall power efficiency from the 1% seen today to 4% using HBF would allow mmWave gNBs to produce 74dBm of EIRP while needing less than 250W of input power — like the power draw of 62 dBm EIRP phased array systems deployed today.