Dust Expiry: Clean the UTXO set from spam

In a thoughtful exploration of optimizing Bitcoin's unspent transaction output (UTXO) management, a novel policy proposal emerges aimed at addressing scalability and node resource concerns without imposing additional fees on users. This approach introduces a scoring system for UTXOs based on their value in satoshis and their longevity within the UTXO set. Specifically, a UTXO's score decreases over time and once it falls below zero, it transitions into an accumulator. To spend these UTXOs, an accumulator-inclusion proof is required, yet this does not contribute to the block weight limits, thereby not incurring extra fees.

This mechanism is designed to be non-confiscatory, ensuring that all UTXOs remain spendable without increasing the associated costs. The maintenance of accumulator-inclusion proofs could potentially be burdensome due to the need for regular updates; however, this task can be outsourced or managed independently by reindexing the blockchain. It is anticipated that consumer-level hardware could feasibly maintain a complete set of proofs for UTXOs created over the past two decades with manageable cost implications.

Applying this rule would notably reduce the current UTXO set size by approximately 81%, significantly decreasing the storage requirements from around 12GB to between 2GB and 3GB. The criterion for UTXO discarding scales with value and time—UTXOs worth less than 4000 satoshis are dropped after two weeks, and those under 104,000 satoshis after a year. This strategy primarily affects older and lower-value UTXOs, which constitute a significant portion of the UTXO set but are rarely spent, thus mitigating the potential impact on most users.

The discussion also touches upon the utreexo concept, highlighting its challenge of requiring users to construct or maintain proofs for transaction broadcasts—a process that might become increasingly complex and costly, impacting privacy, centralization, and efficiency. By contrast, focusing on old or low-value outputs for accumulator inclusion minimizes these issues, as high-value or recent UTXOs continue to be stored individually, avoiding the massive shift to proof-based transactions while still offering a substantial reduction in bandwidth usage and database size.

Finally, the proposal suggests a possible reconfiguration of the scoring algorithm to predict when a UTXO will transition to the accumulator, providing a basis for efficient data management and potentially setting a calculable upper limit on the UTXO set size, given the Bitcoin protocol's existing constraints on supply cap and block size. This consideration underscores the proposal's intent to balance scalability improvements with practical implementation challenges.