MM8108 GB/EU 863–868 MHz — origin of the 2.8% duty cycle, and is LBT/PSA operation available?

General
1

Hi all,

I’m integrating the MM8108 (MF15457) on an ESP32-C5 for a marine data link in the GB regulatory domain, and I’m trying to understand the airtime limits precisely — particularly whether we’re restricted to a fixed duty cycle or can operate under listen-before-talk.

What I’ve found in the SDK. In mmregdb.c, the GB and EU channel tables specify a 2.8% duty cycle (value 280) on all 1 MHz channels across 863–868 MHz, 16 dBm EIRP. The API exposes the duty-cycle machinery — mmwlan_set_duty_cycle_mode() (SPREAD/BURST), mmwlan_get_duty_cycle_stats() — and mmwlan.h points to mmregdb.c for the regulatory limits. I don’t see an LBT/AFA configuration path in this build.

My questions:

  1. Origin of the 2.8% figure. This is an unusual number for the UK — the commonly quoted values are 0.1% / 1% / 10% per the EN 300 220 sub-bands. My understanding is that 2.8% (≈2.78%) corresponds to the Polite Spectrum Access allowance of 2.78% per 200 kHz channel. Is that the correct origin — i.e. is the GB profile derived from the PSA provisions rather than the flat L-band duty cycle? Knowing this helps us understand what headroom exists.
  2. Is LBT / Polite Spectrum Access operation available? EN 300 220 permits LBT+AFA as an alternative to a fixed duty cycle, with no fixed percentage cap. Since 802.11ah uses CSMA/CA (inherently listen-before-talk), the MM8108 should be technically capable. Does Morse Micro offer, or plan to offer, a GB/EU regulatory profile that operates under LBT/PSA rather than the fixed 2.8% duty cycle? If so, how is it selected and configured?
  3. Aggregation across channels. The PSA framework appears to allow a frequency-agile device using multiple 200 kHz segments to increase the aggregate airtime. Does the MM8108’s regulatory implementation support any form of multi-channel airtime aggregation in GB/EU, or is the 2.8% applied per-device across all TX regardless of channel?
  4. Sustained throughput. For a continuous data application in GB, what sustained application-layer throughput should we realistically expect under the current 2.8% profile at, say, MCS0–MCS5 on a 1 MHz channel? I plan to measure with your iperf example — any config caveats for a duty-cycle-limited region?
  5. Certification basis. Which harmonised standard(s) is the MF15457 GB/EU profile certified against (EN 300 220-2, the 802.11ah-specific standard, or other), and is a copy of the Declaration of Conformity / regulatory notes available? This affects what we can legally rely on.

Context: the bulk of our traffic is low-rate and fits comfortably within 2.8%. The open question is a higher-rate sensor stream where the airtime cap is the binding constraint, so whether LBT/PSA is available is the key decision point for our architecture.

I’ve seen the earlier thread on fixed-channel / no-duty-cycle operation — my question is specifically about the standard GB/EU regulatory path and the origin of the 2.8% figure.

Thanks,
Chamith

2

Hi Chamith,

  1. Yes the 2.8% figure is from HaLow being classifiable as PSA, derived from the 100s/h Max Tcum_on requirement stipulated in ETSI EN 300 220-2 Section 4.6.3.2 Table 18 Limits for Polite Spectrum Access timing parameters.

  2. Our EU/GB configured devices default to using PSA. The 2.8% is already the PSA maximum as stipulated by the EN 300 220.

  3. It’s not something that we’re able to support currently, we are looking into improving the throughput in the EU

  4. Applying the 2.8% Duty cycle would mean that MCS0 would be 9.17Kbps single channel, MCS5 would be 74.1Kbps single channel.

  5. There’s an EU DoC hosted on Mouser