Executive Summary

ZC-1 seamlessly integrates Directed Acyclic Graph (DAG) ordering with Byzantine Fault Tolerance (BFT) checkpoints and Verifiable Random Function (VRF)-selected committees, further enhanced by AI-assisted rotation to ensure fairness and efficiency. The protocol delivers finality in under 2 seconds, targets scalability beyond 100,000 transactions per second (TPS), and enables seamless mobile participation through lightweight validation, data availability (DA) sampling, and delegated staking.

Designed with UK regulatory compliance in mind, it incorporates Environmental, Social, and Governance (ESG) incentives to promote sustainable and ethical network operations.

Design Objectives

  • Achieve sub-2-second finality with deterministic safety guarantees
  • Scale to over 100,000 TPS without compromising performance
  • Prioritize mobile-first accessibility for iOS and Android devices
  • Ensure interoperability with Ethereum and implement robust cross-chain bridges
  • Align with UK Financial Conduct Authority (FCA) and General Data Protection Regulation (GDPR) standards
  • Integrate ESG-linked staking rewards and regular audits for transparency and sustainability

System Architecture

The architecture supports applications built on an Ethereum Virtual Machine (EVM)-compatible execution layer, with the core consensus mechanism blending DAG-based ordering and BFT finality. Complementary modules handle Governance & ESG, Interoperability, and Compliance.

Security & Privacy features include post-quantum (PQ) cryptography and zero-knowledge (ZK) proofs for enhanced protection. Networking & Storage manages efficient gossip protocols, erasure-coded data availability, dynamic pruning, sharding, and edge gateways to optimize data distribution and access.

Consensus Protocol - Specification

Stake, Identity, and Committees

Participants bond stake to join the network, with optional Know Your Customer (KYC) integration via trusted custodians for enhanced identity verification. Each consensus round generates a cryptographic seed, from which VRF selects a committee of K validators. A lightweight reputation system subtly biases proposer selection to favour reliable participants, promoting long-term network fairness.

DAG Ordering

Validators produce blocklets that reference at least N - f prior-round blocks (where N represents the total validator set size and f the Byzantine fault tolerance threshold). A deterministic reduction process extracts a consistent DAG cut, forming a batch candidate for validation and inclusion.

BFT Checkpoints and Finality

Finality is achieved through a two-phase voting mechanism: prevote followed by precommit. Once 2/3 or more of the stake-weighted votes are secured, the checkpoint is finalized, a receipt is broadcast, the state machine advances, and obsolete DAG sections are pruned for efficiency.

Data Availability

Transaction batches are encoded using Reed-Solomon codes into k + r shares, distributed across validators for redundant storage. Light clients verify availability by sampling m shares; any detected withholding prevents finality and incurs slashing penalties on the responsible proposers.

Slashing & Fork Choice

Penalties are enforced for violations such as double-proposals, double-votes, invalid state transitions, or DA withholding. The fork choice algorithm prioritizes the chain extending from the most recent finalized checkpoint, ensuring network convergence.

Mobile Participation & Light Clients

Mobile devices (e.g., smartphones) engage via lightweight operations like signature verification, DA sampling, and proof checks. For active staking or validation, users can delegate through edge gateways or participate in micro-committees.

Adaptive duty-cycling adjusts activity based on battery life and thermal limits. Finality receipts encapsulate VRF epoch data, aggregate signatures, and DA proofs for verifiable transparency.

Threat & Fault Model

The system assumes an adversary controlling less than 1/3 of the total stake, capable of delaying, withholding, equivocating, or censoring messages. Clients and gateways are treated as untrusted entities. Bridge security is bolstered by light-client (LC) or ZK-based verification, supplemented by challenge windows to contest invalid transfers.

Network & Synchrony Assumptions

Operating under partial synchrony, the protocol uses gossip with a fan-out approximating √n and rotating peer connections. Timeouts dynamically adjust based on observed latency percentiles to maintain resilience in varying network conditions.

Formal Properties

  • Safety - No two conflicting checkpoints can both reach finality
  • Liveness - The protocol guarantees eventual commitment under partial synchrony
  • Accountability - Faulty behavior produces verifiable proofs, automatically triggering slashing mechanisms

Key Parameters (Recommended Defaults)

Parameter Default Notes
Validator set (n) 400–1,200 Stake-weighted selection with support for gradual validator churn
Committee size (K) 64–256 Dynamically selected via VRF per round
Fault tolerance (f) < n/3 Standard Byzantine fault tolerance bound
Batch size 1–4 MB Optimized with erasure coding for compact storage and transmission
Checkpoint interval 250–500 ms Balances latency and finality; adjustable based on network load
Gossip fan-out (g) ≈ √n Employs rotating peer sampling for efficient propagation
DA samples (m) log(k + r) + δ Ensures probability of undetected withholding is below 2⁻⁴⁰
Slashing amount 0.5–5% stake Tiered penalties scaled by violation severity
Unbonding period 14–21 days Prevents long-range attacks by delaying stake withdrawal

Cryptography & Privacy

The protocol employs post-quantum signatures (e.g., Dilithium or Falcon) for future-proof security. zkSNARKs and zkSTARKs facilitate efficient verification. Selective-disclosure techniques enable privacy-preserving KYC attestations, allowing users to reveal only necessary information during audits or compliance checks.

Economic Model, Mempool & DoS Controls

Transactions incorporate a base fee plus optional tip; validators earn tips alongside a portion of base-fee rebates, aiming for a 6–12% annual staking yield. Maximal Extractable Value (MEV) is mitigated through fair DAG ordering and encrypted mempool phases.

Denial-of-Service (DoS) defenses include adaptive fees, per-origin rate limits, peer reputation scoring, greylisting, and optional Proof-of-Work (PoW) challenges. Snapshot-based synchronization accelerates node bootstrapping.

Compliance (UK FCA / GDPR) & ESG

Pseudonymous on-chain commitments are complemented by off-chain KYC through custodians. Selective-disclosure proofs support regulatory audits without compromising user privacy. ESG integration offers staking multipliers or bonus token emissions for validators demonstrating sustainable practices, such as energy-efficient operations.

Interoperability & Bridges

Primary bridges leverage LC or ZK verification for trustless cross-chain interactions. As a fallback, threshold-signed bridges include slashing incentives, challenge protocols, and time-bounded proofs to minimize risks.

Upgrade & Governance Process

Network upgrades proceed in two phases: initial proposal with mandatory independent audit, followed by a stake-weighted on-chain referendum requiring quorum and time-locks. Emergency hot-patches are strictly limited in scope, protected by multisig controls, set to expire automatically, and mandate comprehensive post-mortem analyses.

Testnet Roll-out & Benchmarking Plan

  • Phase 0: Localnet testing with simulated faults and edge cases
  • Phase 1: Devnet deployment limited to ≤200 validators for initial validation
  • Phase 2: Incentivized testnet scaling to ≥600 validators for real-world stress testing

Benchmarking focuses on key metrics: TPS throughput, latency distribution (CDF), finality durations, reorg depth (zero after commitment), energy consumption per transaction (measured on Android/iOS), bridge verification times, and DA failure incidence.

Implementation Roadmap & Repository Mapping

  • consensus/ConsensusLayer.py - event loop, DAG ordering, checkpointing, votes, slashing hooks
  • consensus/messages.py - Blocklet, Checkpoint, Vote, FinalityReceipt
  • consensus/vrf.py - VRF sortition (replace dummy with production VRF)
  • consensus/storage.py - in-memory DAG + checkpoint stores
  • networking/ - gossip, discovery, sync, finality receipts
  • privacy/ - ZK verifiers; off-chain proving
  • governance/ - delegation, referenda, params
  • interop/ - bridge protocols (LC/ZK or threshold + slashing)

High-Level Pseudocode

while True:
    seed = derive_seed_from_last_checkpoint()
    committee = vrf_sortition(validators, seed, size=K)
    if self in committee:
        msg = build_blocklet(pending_txns, refs=dag.select_refs(N_minus_f))
        msg.zk_attest = optional_zk_attestations(msg)
        broadcast(msg)
    dag.update(received_messages)
    if dag.has_checkpoint_candidate():
        cand = dag.best_checkpoint()
        votes = bft_prevote_precommit(cand, threshold=two_thirds_stake)
        if votes.commit:
            finalize(cand)
            emit_finality_receipt(cand)
            prune_old_dag(cand)
            rotate_reputation_scores()