WiFi Regulatory Requirements
There is a very large body of regulatory and standards requirements which need to be met by any shipping product.
This page is meant to be a non-exhaustive introduction into the topic. This must not be treated as authoritative - any company or organisation wishing to create and ship a wireless device must engage the relevant regulatory bodies and create a suitable testing/certification plan.
There is a lot of uncertainty around what exactly is required for any given wireless device. Although the rules vary from country to country, here are some general guidelines:
(TBD: flesh out with examples to specific regulatory requirements):
- When an access point, laptop or other device is certified, the entire device is certified - this includes the laptop itself, the NIC and the antenna choice/configuration.
- Certification may involved BOTH the hardware AND software.
- Choosing an already certified NIC from a vendor does not bypass the requirement to certify the entire solution - if you place it in your own enclosure, with your own antenna configuration and customised firmware, you will have to re-certify the entire device.
Getting a NIC certified is a complicated task. The regulatory bodies require many things to be tested before certification is given. The following is not a complete, exhaustive list:
- The NIC must be able to emit RF at the correct centre frequency. This is typically tested using a continuous TX tone test, using a calibrated spectrum analyser to check the centre frequency.
- The NIC must meet some spectral emission mask - the requirements here may be different to the IEEE 802.11 spectral mask requirements. Typically this means the NIC must not distort transmissions and leak into adjacent channels; that the TX side has good TX filtering to meet these spectral mask requirements.
- When utilising MIMO antennas, the total radiated power will be not necessarily be the straight sum of the per-chain TX power level. There may be other effects at play (eg STBC) which can change the power density of the signal at each antenna.
- Since higher CCK/OFDM/MCS encoding rates have increasing power density, the total transmission power tends to be lower for each higher rate in order to meet the spectrum emission requirements.
In addition to the NIC requirements, there are typically device requirements. The following is not an exhaustive list.
- There are generally transmission power limits in different frequency bands - these may be specified across an entire transmission bandwidth (eg a dBi across a 20MHz channel) along with a spectral mask; others may specify a per-MHz maximum emission limit.
- The transmission power limits tend to be effective radation (EIRP) rather than the output from the NIC (in dBm.) This means the antenna choice and configuration will play a part in the certification process.
- When transmitting using multiple antennas (eg MIMO, phased array/smart antennas), the output radiated power is going to be a complex combination of both the per-chain transmission power and the antenna configuration/elements selected.
DFS device certification
When operating in the DFS bands in 5GHz, there are also strict regulatory requirements on a number of behaviours, including:
- Implementing CAC (Channel access check) correctly - not transmitting anything during CAC, which includes probe responses and ACK frames;
- Being able to accurately detect radar;
- Being able to cease transmissions quickly upon detecting radar, and only emitting CSA IE's in the next few beacons (which advertise the new channel to move to);
- Implementing some kind of channel spectrum management to "spread" 802.11 emissions across the whole band, rather than simply always using the lowest available channel number or some other non-fair mechanism.
This requires a separate certification for DFS compliance and again the entire device is tested and certified - including the processor board, wireless NICs, cabling, antenna selection, enclosure AND software.
Other reasons for transmission power limitations
There are three main reasons for limiting TX power:
- There's a specific limitation for that particular channel and TX rate. Transmitting at higher TX rates does increase the power density and can lead to distortion at higher TX powers.
- There may be an electronic reason for limiting TX power at higher rates - for example, the NIC may draw too much transient power and cause TX distortion or instability, or the output amplifier/filtering circuitry doesn't behave in quite as linear a fashion as the designers would have liked.
- There are requirements at band edges ("CTL" edges) for more stricter spectral masks (so the 802.11 emissions at these band edges don't "leak" into the adjacent, non-802.11 licenced channel.) Typically this results in a lower TX power at the band edge in order to ensure the spectral mask at said band edge is below the required emission level into the adjacent band.
WiFi standards certification
Country specific regulatory requirements
The Australian regulatory body is the ACMA (http://www.acma.gov.au). The relevant rules can be found via these links:
2009 ACMA Radio Spectrum Plan: http://www.acma.gov.au/WEB/STANDARD/pc=PC_2713
Radio Frequency Planning Topics: http://www.acma.gov.au/WEB/STANDARD/pc=PC_2709
There may also be some requirements for antenna configuration and radiation patterns for fixed microwave services (TBD):