Why the Same 12-Byte Reading Has Two Efficiencies
Ada re-derives this chapter’s own numbers step by step, at full precision
ADA · CALCULATION AUDIT
Why the Same 12-Byte Reading Has Two Efficiencies
The same 12-byte sensor reading measures 41.4% efficient over LoRaWAN but only 22.2% over MQTT/TCP/IP — nearly a 2x gap for identical data. The difference is entirely the fixed header each stack drags along, from a 13-byte LoRaWAN MAC header to the stacked TCP and IP headers of MQTT. This audit separates payload from overhead to explain why one reading has two efficiencies.
Companion to the chapter Lab: Network Performance — every number here comes from that chapter.
Ada: This lab reports that a 12-byte reading is 41.4% efficient over LoRaWAN but only 22.2% over MQTT/TCP/IP, and that LoRaWAN is “nearly 2x” better. Same payload, two answers — the difference is entirely the fixed header each stack drags along, so let me separate payload from overhead.
Over LoRaWAN the reading carries a 13-byte MAC header and a 4-byte MIC:
- Total on air:
12 + 13 + 4 = 29bytes. - Goodput efficiency:
12 / 29 = 0.413793, i.e. 41.4%; the overhead is the complement,17 / 29 = 0.586207 = 58.6%.
Over MQTT the same 12 bytes ride a 2-byte MQTT header, a 20-byte TCP header, and a 20-byte IP header:
- Total:
12 + 2 + 20 + 20 = 54bytes. - Efficiency:
12 / 54 = 0.222222, i.e. 22.2%. - LoRaWAN’s advantage:
(12 / 29) / (12 / 54) = 54 / 29 = 1.862, so about 1.9x — the “nearly 2x” claim holds.
The design meaning is that efficiency is a property of the payload-to-header ratio, not of the protocol name. The 17-byte LoRaWAN overhead that swallows 58.6% of a 12-byte reading shrinks to just 17 / (100 + 17) = 0.1453 = 14.5% of a 100-byte reading (100 / 117 = 85.5% efficient). The lesson for a sensor designer is blunt: batching or enlarging the payload lifts efficiency far faster than switching protocols, because the fixed header is paid once per packet no matter how little it carries.
Every number above is taken from the chapter’s own material and re-derived step by step.