rbxii3/firmware
1
0
Fork 0
mirror of https://github.com/meshtastic/firmware.git synced 2026-01-28 20:52:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

Enhance congestion scaling coefficient for online nodes with power-law scaling

Browse Source
This commit is contained in:
Ben Meadors
2026-01-28 06:34:47 -06:00
parent 4eb4c4b584
commit ed063568b5
1 changed files with 69 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

89
src/mesh/Default.h
View File

@@ -1,5 +1,6 @@
#pragma once
#include <NodeDB.h>
#include <cmath>
#include <cstdint>
#include <meshUtils.h>
#define ONE_DAY 24 * 60 * 60
@@ -57,31 +58,79 @@ class Default
static uint32_t getConfiguredOrMinimumValue(uint32_t configured, uint32_t minValue);
private:
// Note: Kept as uint32_t to match the public API parameter type
/**
* Calculates a congestion scaling coefficient based on the number of online nodes.
*
* Uses power-law scaling (exponent 1.2) which provides a soft start that accelerates
* as node count increases - matching the superlinear growth of flood routing traffic.
*
* Scaling starts at 20 nodes (simulator shows congestion problems emerging early).
* Different modem presets have different channel capacities based on airtime per packet.
*
* Examples for LongFast (capacityMultiplier = 1.0):
* 20 nodes: 1.0x, 50 nodes: ~3.0x, 100 nodes: ~6.9x, 200 nodes: ~15.8x
* Examples for ShortFast (capacityMultiplier = 0.5):
* 20 nodes: 1.0x, 50 nodes: ~2.0x, 100 nodes: ~4.0x, 200 nodes: ~8.4x
*/
static float congestionScalingCoefficient(uint32_t numOnlineNodes)
{
if (numOnlineNodes <= 40) {
return 1.0;
} else {
float throttlingFactor = 0.075;
if (config.lora.use_preset && config.lora.modem_preset == meshtastic_Config_LoRaConfig_ModemPreset_MEDIUM_SLOW)
throttlingFactor = 0.04;
else if (config.lora.use_preset && config.lora.modem_preset == meshtastic_Config_LoRaConfig_ModemPreset_MEDIUM_FAST)
throttlingFactor = 0.02;
else if (config.lora.use_preset &&
IS_ONE_OF(config.lora.modem_preset, meshtastic_Config_LoRaConfig_ModemPreset_SHORT_FAST,
meshtastic_Config_LoRaConfig_ModemPreset_SHORT_TURBO,
meshtastic_Config_LoRaConfig_ModemPreset_SHORT_SLOW))
throttlingFactor = 0.01;
// Start scaling at 20 nodes - meshes show congestion problems earlier than 40
if (numOnlineNodes <= 20) {
return 1.0f;
}
// Use power-law scaling (p=1.2) - soft start that accelerates with node count,
// matching the superlinear growth of flood routing traffic
float baseScale = powf(static_cast<float>(numOnlineNodes) / 20.0f, 1.2f);
// Apply modem-specific capacity multiplier based on relative channel capacity.
// Capacity is inversely proportional to airtime - faster modems can handle more
// traffic before congestion, so we scale their intervals less aggressively.
// Airtime values are for a typical 237-byte packet (max payload).
float capacityMultiplier = 1.0f;
if (config.lora.use_preset) {
switch (config.lora.modem_preset) {
case meshtastic_Config_LoRaConfig_ModemPreset_SHORT_TURBO:
capacityMultiplier = 0.3f; // ~28ms airtime, BW500 SF5
break;
case meshtastic_Config_LoRaConfig_ModemPreset_SHORT_FAST:
capacityMultiplier = 0.5f; // ~50ms airtime, BW500 SF7
break;
case meshtastic_Config_LoRaConfig_ModemPreset_SHORT_SLOW:
capacityMultiplier = 0.7f; // ~100ms airtime, BW500 SF8
break;
case meshtastic_Config_LoRaConfig_ModemPreset_MEDIUM_FAST:
capacityMultiplier = 0.7f; // ~100ms airtime, BW250 SF7
break;
case meshtastic_Config_LoRaConfig_ModemPreset_MEDIUM_SLOW:
capacityMultiplier = 0.85f; // ~200ms airtime, BW250 SF8
break;
case meshtastic_Config_LoRaConfig_ModemPreset_LONG_TURBO:
capacityMultiplier = 0.85f; // ~150ms airtime, BW250 SF9
break;
case meshtastic_Config_LoRaConfig_ModemPreset_LONG_FAST:
capacityMultiplier = 1.0f; // ~300ms airtime, BW250 SF10 (baseline)
break;
case meshtastic_Config_LoRaConfig_ModemPreset_LONG_MODERATE:
capacityMultiplier = 1.0f; // ~350ms airtime, BW125 SF9
break;
case meshtastic_Config_LoRaConfig_ModemPreset_LONG_SLOW:
capacityMultiplier = 1.3f; // ~700ms airtime, BW125 SF10
break;
case meshtastic_Config_LoRaConfig_ModemPreset_VERY_LONG_SLOW:
capacityMultiplier = 1.3f; // ~1400ms airtime, BW62.5 SF11
break;
default:
capacityMultiplier = 1.0f;
break;
}
}
#if USERPREFS_EVENT_MODE
// If we are in event mode, scale down the throttling factor
throttlingFactor = 0.04;
// Event mode: more aggressive throttling for dense temporary meshes
capacityMultiplier *= 1.5f;
#endif
// Scaling up traffic based on number of nodes over 40
int nodesOverForty = (numOnlineNodes - 40);
return 1.0 + (nodesOverForty * throttlingFactor); // Each number of online node scales by 0.075 (default)
}
return 1.0f + (baseScale - 1.0f) * capacityMultiplier;
}
};