Just as quantum mechanics imposes a rough floor size on chip process design, so more mundane laws of physics constrain miniaturization of radio antennas for small form factor IoT devices.
This is of course independent of the radio protocol, whether we are talking about 5G, Zigbee, LoRaWAN, NB-IoT, WiFi or any of the other options for either indoor or outdoor wireless communication over varying ranges and frequencies. There is no absolutely precise limit, but the rule of thumb is that the antenna cannot be shorter than one quarter of the wavelength of signals it is designed to receive.
This means that antenna can be smaller at higher frequencies and at 70 GHz, about the highest frequency practical for mobile and IoT usage in millimeter wave spectrum at present, that would be just one millimeter, so not a problem at all. However, most IoT radio protocols operate just below 1 GHz, when minimum antenna length is closer to seven centimeters.
Even this can be accommodated within a package design using surface mount technology (SMT) assembled onto printed circuit boards (PCBs). These can utilize the so-called PCB ground plane which boosts both efficiency and bandwidth. Where space is a constraint, laser direct structuring and similar processes can be employed to print customized antenna structures on the inside of a device’s plastic housing, which is conducive for low cost, high volume production.
However, an additional problem arises with hybrid multimode devices supporting several radios, as will be commonplace in the IoT just as it is for smartphones. Indeed, demand for compact IoT devices that serve multiple use cases will demand multimode radios.
This runs into another rough law of physics governing multiple radio coexistence. While the role of antennas is to transmit electromagnetic energy over the air, they will also exchange that energy increasingly with companion antennas the closer they are together. The basic rule is that the ratio of received target signal power to noise power must be above a certain threshold in order for data to be exchanged without loss. The problem is worse for longer range communication as in LPWAN, or indeed 4G/5G, because the received signal power is that much lower, so that the ‘noise floor’ also has to be lower.
This is leading R&D towards dedicated multimode components optimized for reliable operation, which involves trade-offs between size, receive efficiency, bandwidth or spectral coverage and power consumption. Improving one comes at the expense of some or all of the others.
Optimizing this balance in design of RF subsystems has become an important skill with the antenna adding an extra layer of complexity. Although that is not directly related to the wireless protocols involved that does interact with the overall design and hardware layout where the relevant protocol stacks play a part.
Although integrated systems will increasingly be available, there will be conflicting requirements to be met, with varying protocol sets and device battery life being more critical in some cases than others. The main point is that antenna design has become more important as devices have shrunk in size and as the number of radio protocols being supported has increased.