This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

ETSI 300.220 LBT + AFA duty cycle

Hi,

In order to increase the TX duty cycle, AFA & LBT can be used in most cases for the 863 to 870 MHz ISM band. However, in the 869,400 MHz to 869,650 MHz g3 band (500 mW), the duty cycle limit is 10% which is 360 sec in one hour (if not using LBT+AFA). But if you choose  to use LBT & AFA, the duty cycle will actually decrease! The maximum on time for AFA+LBT is 100 sec in any 200 kHz spectrum (which basically means the whole sub g3 band).

This is what I find a bit confusing. If your system is "polite" and uses LBT and you implement AFA (changing channels), you will have less TX on-time compared to if you use a single channel with no LBT... One intention with using LBT+AFA is to allow a greater duty cycle - which is normally the case for the rest of the ISM band.

So my question is: would it be possible to use LBT and AFA but with the 10% duty cycle definition for this sub 3g band?

Argument1: I guess when it comes to LBT, no one could blame you for being "polite" and use LBT - the device can transmit anyway whenever it wants.

Argument2: But, when using AFA, you actually change channels. The question is whether the duty cycle limitation (as apposed to the AFA+LBT) imposed by the EN 300.220 actually refers to single channel devices.

One could also imagine a system which has just AFA (no LBT). Then I guess the duty cycle limit applies (normally the stricter rule). So combining this with argument1 I guess one should be allowed to apply the normally "stricter" duty cycle rule which turns out to be a better option while still implementing LBT and AFA if one chooses to do so anyway.

Anyone who has any insight to this special case?

Regards

  • For LBT devices without AFA, the duty cycle limit applies. In the 869.4 - 869.65 MHz band this is 10%. Of course, there is nothing preventing you from being "polite" and implementing LBT as long as you comply with the duty cycle restriction.

    An LBT+AFA device will have a maximum "duty cycle", which is around 2.7% (100 seconds per hour) per 200 kHz of spectrum, see 9.2.5.2.3 on page 56. Thus, in your case there is no benefit in using LBT+AFA.

    There are of course scenarios where LBT+AFA gives higher duty cycle. As an example, if you utilize 3 MHz of spectrum, you can have a maximum duty cycle of 41.6% (as long as all channels are utilized equally much).

  • In my current proprietary FHSS mesh network application I utilize a large spectrum due to the latter example you provide. However, in my new application I wish to use narrow band for increased range in addition to the higher E.R.P. I still wish to use AFA (or actually FHSS+LBT) due to its benefits (avoiding noisy channels etc), so in this case AFA is a benefit. But I still would like to have 10% duty cycle.

    I guess that if I define the system as FHSS and not AFA, the 10% could be valid.

    In order to distinguish AFA from FHSS, ETSI TR 102 313 provides a definition of AFA:

    "Frequency agility can be defined as the ability to select an unoccupied channel for operation in order to avoid other
    users within the same band and hence minimize interference."

    "If the channel is occupied then the device must either wait until the channel is free or look for another channel. Having chosen a free channel the device should not change channels unless contention is detected."

    Since FHSS will not stay on a free channel, one could say its not AFA and hence the 10% rule could apply :-)

    What do you think?

  • In my opinion, ignore the FHSS section in EN 300 220, because there are absolutely no circumstances that you can get any advantage by declaring your device as FHSS, compared to a low duty cycle device or an LBT+AFA device.

    For example, FHSS (without LBT on every hop) is restricted to 0.1% duty cycle for the entire duration (not for each channel), or 1% if you only use the 865-868 MHz frequency band.This is exactly the same as any non-FHSS low duty cycle device would have if it transmits on a single frequency. However, a frequency agile device without LBT (i.e. a low duty cycle device) is allowed to have a duty cycle of 0.1% per channel, see 7.10.3 on page 42. Thus, you can increase the maximum duty cycle of your FHSS device from 0.1% to 5% (assuming 50 channels) by simply not declaring it FHSS. Furthermore, you would of course not have to follow the FHSS rules anymore.

    Similarly, both FHSS with LBT on every hop and LBT+AFA devices would have the same maximum "duty cycle", which is around 2.7% (100 seconds per hour) per 200 kHz of spectrum, see 9.2.5.2.3 on page 56. Thus, if you utilize 3 MHz of spectrum, you can have a maximum duty cycle of 41.6% (as long as all channels are utilized equally much). Again, there is no advantage of declaring your device as FHSS, just more rules to follow.

    Something is not adding up: You started by asking about 869.4 - 869.65 MHz band and now you refer to FHSS. FHSS in the band you mention is not really possible.

  • I agree with your first statement. If you would like to change channels dynamically, define it as AFA.

    There is space for 19-20 channels @ 12.5 kHz in the g3 band. I am aware that this is less than the minimum required number of channels for defining the system as FHSS. Anyway, good for AFA.

    Basically I want to have frequency hopping (AFA) with LBT and still use the 10% duty requirement. Don't you agree that it is a bit odd that the duty cycle requirement becomes more strict when utilizing LBT+AFA (for sub g3)?