Unfair Emission Distribution During Network Inactivity Periods

[distributedstatemachine]

Issue

During network inactivity (>5000 blocks), miners receive zero emissions while validators receive 100% of emissions based on stake. This creates an unfair distribution and penalizes miners during maintenance or network issues beyond their control.

Current behavior in run_epoch.rs:

if emission_sum == I32F32::from(0) {
    if is_zero(&active_stake) {
        // no active stake - miners get 0, validators get everything
        normalized_validator_emission.clone_from(&stake);
        normalized_combined_emission.clone_from(&stake);
    }
}

Impact

• Miners lose all earnings during maintenance periods
• Creates misaligned incentives during network recovery
• Unfairly penalizes miners for protocol-level issues
• Reduces network resilience by discouraging miner participation during recovery

Proposed Solutions

1. Minimum Miner Emission Floor

// Ensure miners get at least X% of emissions even during inactivity
let MIN_MINER_EMISSION_RATIO: I32F32 = I32F32::from_num(0.3); // 30%

if emission_sum == I32F32::from(0) {
    if is_zero(&active_stake) {
        // Split emissions between miners and validators
        normalized_server_emission = stake.iter()
            .map(|s| s.saturating_mul(MIN_MINER_EMISSION_RATIO))
            .collect();
        normalized_validator_emission = stake.iter()
            .map(|s| s.saturating_mul(I32F32::from_num(1.0) - MIN_MINER_EMISSION_RATIO))
            .collect();
    }
}

2. Time-Based Decay

// Gradually shift emissions to validators over time
let blocks_inactive = current_block.saturating_sub(last_activity);
let decay_rate = (blocks_inactive / activity_cutoff).min(1.0);
let miner_ratio = I32F32::from_num(0.5).saturating_mul(I32F32::from_num(1.0 - decay_rate));

3. Historical Performance Weighting

Use historical performance metrics to distribute inactive emissions instead of zeroing miner rewards.

Implementation Considerations

• Need configurable parameters for minimum miner emissions
• Should track historical performance metrics
• Consider subnet-specific settings
• Add emergency override capabilities
• Include monitoring for extended inactivity

Migration Plan

1. Implement minimum emission floor first
2. Add configuration parameters
3. Monitor impact
4. Consider more sophisticated solutions based on data

Questions

• What's the right minimum emission % for miners?
• Should this vary by subnet?
• How to handle very long inactivity periods?
• What metrics should determine historical performance?