Cysic Auction Mechanism

1. Motivation

Efficient task assignment in decentralized computation requires a fair, incentive-compatible, and penalty-enforced bidding system. The auction mechanism ensures that tasks are allocated to provers with sufficient computational resources, while maintaining economic alignment between requesters, provers, and verifiers.

2. Design Goals

• Fairness: Guarantee that tasks are allocated based on competitive bidding under a capped maximum price.

• Efficiency: Encourage timely completion by rewarding faster provers with higher payouts.

• Security: Ensure that dishonest or underperforming provers are penalized through slashing.

• Incentive Alignment: Incorporate token reserves into the reward distribution, aligning long-term commitment with higher earnings.

3. Task Definition

A task is published by the Requester with the following parameters:

• Task Difficulty (task_difficulty): Expressed in cycles.

• Maximum Acceptable Bid (bid_max): The upper bound of unit price, in CYS per million cycles.

• Task Deadline (task_ddl): The maximum allowable computation time.

4. Prover Requirements

• Token Reservation: Each prover must lock a minimum amount of tokens to be eligible.

• Bidding Configuration: Provers specify their bidding price in the configuration file, which serves as the reference for subsequent auctions.

• Performance Evaluation: By running the client software, a prover’s computational capacity is automatically benchmarked.

5. Auction Flow

1. Task Broadcast: The requester publishes the task, which is propagated to all registered provers.

2. Capability Assessment: Each prover determines whether it can complete the task within task_ddl.

3. Bid Submission: Provers submit bids reflecting their desired compensation.

   1. Bids exceeding bid_max are automatically discarded.

4. Bid Selection: Upon bidding closure, the system sorts valid bids by price.

   1. The two winning provers are chosen starting from the second-lowest bid.

      • If two provers submit the same price, the more reserved one will be chosen.

   2. The selected bid price (bid_select) is the lowest bid among these winners.

6. Reward Distribution

6.1 Prover Rewards

The total reward pool for provers is defined as: $task\_reward\_prover = bid\_select \times task\_difficulty \times 80\%$

This reward is distributed based on completion speed among the three winning provers:

• Fastest prover: 80% of task_reward_prover

• Second fastest: 20%

6.2 Failure and Slashing

• If a prover fails to meet the deadline, the task is reassigned to a backup prover.

• The failing prover incurs a penalty:

  $slash = \beta \times bid\_max \times task\_difficulty$

• where $\beta = 1$.

6.3 Verifier Rewards

• 20 verifiers are randomly selected to validate the proof.

• They share the remaining 20% of the task reward equally:

  $verifier\_reward = \frac{bid\_select \times task\_difficulty \times 20\%}{20}$

7. Reserve-Weighted Incentives

To strengthen long-term commitment and discourage opportunistic behavior, prover rewards are further adjusted by token reserves:$$final\_task\_reward\_prover = task\_reward\_per\_prover \times \big(1 + \gamma \times R\big)$$ Where:

• $\gamma$: Initially set to 0.25, but subject to adjustment in the future depending on the overall reserve distribution.

• $R = \frac{reserve\_token}{total\_reserve\_token}$: relative reserve ratio.

This ensures that provers with larger token commitments receive proportionally higher rewards.

8. Security Considerations

• Sybil Resistance: Token reservation reduces the risk of Sybil attacks by requiring economic commitment.

• Collusion Prevention: Pricing determined by the second-lowest bid mitigates collusion among provers.

• Reliability Enforcement: Slashing enforces accountability, ensuring underperforming provers bear economic costs.

• Verifier Integrity: Distributing rewards among multiple verifiers enhances the robustness of proof validation.