Why massive MIMO radios strain passive heatsinks, what the industry uses today, and where solid-state airflow fits. Talk to our engineers.
TL;DR: A 64T64R massive MIMO site can draw roughly three times the power of an LTE radio, and almost all of that ends up as heat inside a sealed, pole-mounted box cooled by natural convection [1]. Fans fix the airflow problem but add a moving part to equipment that is expected to run unattended for years. This post walks through where the heat comes from, why passive fins are reaching their limits, and where solid-state airflow can help.
5G moved the power-hungry electronics up the mast. In a massive MIMO active antenna unit (AAU), the power amplifiers, transceivers and beamforming logic sit directly behind the antenna array, in one enclosure.
ABI Research estimates that a 5G RAN with a 64T64R massive MIMO configuration draws up to 2.7 kW in typical conditions, against about 0.8 kW for an LTE radio [1]. The same analysis notes that 5G is far more efficient per bit, but that absolute site consumption still rises because the radios are more power-hungry and networks are denser [1]. A peer-reviewed survey of 5G RAN energy efficiency identifies the radio unit, and in particular the power amplifiers, as the dominant consumer in a modern base station and the main target for energy-saving work [2].
For thermal engineers the per-bit figure does not matter much. What matters is the number of watts that must leave a closed box on a pole, in direct sun, with no service visit for years.
Most outdoor radio units are cooled by natural convection through large aluminium fin arrays. A published design study of AAU 5G and RRU 4G heatsinks lists the reasons: low cost, high reliability, silent operation and no dependence on other cooling circuits. It also names the main drawback directly: relatively low heat transfer [3].
That tradeoff shapes the whole product. When the heat transfer coefficient is low, the only levers left are:
Natural convection also depends on orientation and ambient conditions. Still air, direct solar load and a dense radio cluster on the same pole all push the operating point in the wrong direction at the same time.
Fans raise the heat transfer coefficient, and some radio designs use them. But a fan in an outdoor radio brings problems that do not exist indoors:
Liquid cooling is the other path the industry is exploring. ABI Research reports that liquid-cooled sites can be lighter and smaller than sites with active air conditioning and describes the technology as still at an early stage [1]. It addresses the cabinet and baseband side of the site well. It is a bigger architectural change for the radio head itself.
There is a gap between "passive fins only" and "a fan or a liquid loop". Solid-state airflow devices aim at that gap: they add forced convection without a rotating part.
The physics is well documented. Electrohydrodynamic (EHD) flow, often called ionic wind, uses an electric field to accelerate ions, which transfer momentum to neutral air. In a study by Purdue University and Intel, an ionic wind generator integrated on a heated surface in the presence of a bulk flow distorted the thermal boundary layer and enhanced the local heat transfer coefficient by more than a factor of two [4]. That experiment used a corona discharge. Dielectric barrier discharge (DBD) actuators use the same momentum-transfer principle with the discharge formed across a dielectric layer, which gives a thin, surface-mounted device. We explain the mechanism in detail in how DBD plasma actuators work.
For a radio head, the relevant property is not raw airflow. It is the ability to break up the stagnant boundary layer where natural convection is weakest:
Because a DBD actuator has no bearing and no rotor, it removes the wear-out mechanism that makes fans unattractive on a mast. Its own life is set by dielectric ageing, which still has to be characterised for outdoor duty cycles, and its high-voltage drive and ozone output need to be engineered for a sealed enclosure. We cover those questions openly in ionic wind reliability and ozone, explained.
Solid-state airflow is not a replacement for the main heatsink of a kilowatt-class massive MIMO radio. The fin array still carries the bulk of the heat to ambient. The realistic role is augmentation: raising the effective heat transfer coefficient where it is lowest, so that the same fin array handles more power, or the same power runs cooler. For a broader comparison of the options, see our complete guide to solid-state cooling and ionic wind vs fans for electronics cooling.
| Approach | Moving parts | Heat transfer | Main limitation on a mast |
|---|---|---|---|
| Natural convection fins | None | Low [3] | Size, weight and derating as power rises |
| Fans | Rotor and bearings | Higher | Wear-out, ingress, site visits |
| Liquid cooling | Pump | Highest | Architectural change, early stage for radio heads [1] |
| Solid-state airflow (EHD / DBD) | None | Locally enhanced [4] | Augments, does not replace, the fin array; life still being characterised |
If you are designing a radio unit or small cell and passive fins are running out of margin, we would like to compare notes on your hotspot map and enclosure constraints. Contact the YPlasma team to book a technical call.
How much power does a 5G massive MIMO base station consume? ABI Research estimates up to 2.7 kW for a 5G RAN with 64T64R massive MIMO in typical conditions, versus about 0.8 kW for an LTE radio [1]. Actual figures depend on configuration, hardware generation and traffic load.
Why are 5G radio units passively cooled? Natural convection is low cost, reliable, silent and independent of other cooling systems, which suits unattended outdoor equipment. Its main drawback is low heat transfer [3].
Can ionic wind replace the heatsink in a base station? No. In a kilowatt-class radio the fin array remains the main heat path. Ionic wind and DBD actuators can enhance local heat transfer where natural convection is weakest [4].