Briefing
The core problem in modular blockchain architecture is the reliance on centralized Layer 2 (L2) sequencers, which introduces a single point of failure, censorship risk, and compromised credible neutrality. This research proposes Based Sequencing , a mechanism that delegates L2 transaction ordering rights to the most decentralized and economically secure actors → the Layer 1 (L1) proposer set → via specialized Gateways. The breakthrough lies in decoupling high-speed user experience (UX) through near-instant preconfirmations from the slower L1 settlement, enabling rollups to inherit L1 security and achieve a scalable, synchronous composability across the entire ecosystem. This framework shifts the security model from a single, trusted operator to the L1’s robust economic consensus, fundamentally resolving the centralization trade-off that has defined the rollup landscape.
Context
Prior to this work, L2 rollups achieved high performance by utilizing a single, centralized sequencer to order and batch transactions, which offered excellent latency but sacrificed the foundational principles of decentralization and censorship resistance. This centralized model created a single point of failure for liveness, a single entity with a monopoly on transaction inclusion, and a significant vector for Maximum Extractable Value (MEV) extraction, undermining the credible neutrality of the entire L2 layer. While “escape hatch” mechanisms existed to post transactions directly to L1, they resulted in a severely degraded user experience due to high cost and significant delays, failing to offer a practical, trust-minimized alternative to the centralized sequencer.
Analysis
Based Sequencing operates by mapping L1 slots to Gateway identities via a public Lookahead schedule, effectively establishing a rotating leader for L2 transaction ordering. L1 proposers, who are the most credibly neutral actors, delegate their sequencing rights to these high-performance Gateways, which are staked entities held accountable by slashing conditions for safety and liveness faults. The Gateways sequence transactions into small, signed fragments called Frags and distribute them via a fast, leader-aware gossip network.
This pipelined block construction allows follower nodes to optimistically construct a TempBlock and issue preconfirmations to users within milliseconds, conceptually advancing the L2 state ahead of its final L1 settlement. The L1 proposer set acts as the ultimate watchdog, enforcing the Gateway schedule and penalizing misbehavior through economically guaranteed slashing.
Parameters
- Execution Preconfirmation Latency → Milliseconds – The time for a user to receive a preconfirmation receipt from the Gateway leader, providing a UX comparable to centralized sequencers.
- Demonstrated Throughput → 300,000 Transactions per Second – Achieved by pipelined sequencers on a single core, demonstrating the raw performance potential of the design.
- L1 Slot Assignment → Lookahead Schedule – A public mapping from L1 slot numbers to Gateway identities, enforcing a leader rotation for sequencing rights.
- Slashing Conditions → Reneging on Preconfirmations – A key mechanism for economic security, penalizing Gateways that fail to honor their promises or settle batches on L1.
Outlook
The immediate next step for this research is the transition to Phase 3 → fully permissionless sequencing, which requires resolving complex, open-ended mechanism design challenges. Specifically, the community must formally specify protocols to prevent front-running by malicious Gateway leaders, establish robust fault attribution between L1 proposers and their delegated Gateways, and define clear, enforceable MEV policies that align incentives across the entire system. In the next 3-5 years, this theoretical model is poised to unlock truly synchronous composability across multiple rollups sharing the same sequencing set, enabling atomic transactions and shared liquidity that were previously impossible, thereby realizing the full potential of the modular blockchain thesis.
