Multi-path Packetized Routing for Payment Channel Networks

This direction is merged into another direction: channel liquidity depletion. Its last version stays readable and citable here.

Version 1 of 17 papers17 conceptsfrozen 2026-10-04

In traditional Payment Channel Networks (PCNs) like the Lightning Network, sending money is an "all-or-nothing" affair. If you want to send 1 BTC, you must find a single path from sender to receiver where every intermediate hop has at least 1 BTC of liquidity. If a single link along that path is depleted, the entire transaction fails. This "atomic" routing leads to high failure rates, fragmented liquidity, and a constant need for massive collateral.

A growing body of research suggests a different approach: treating cryptocurrency transactions like internet data packets. By decomposing large payments into smaller units and routing them across multiple paths simultaneously, these systems can achieve significantly higher throughput and better network balance.

From Atomic Blocks to Liquidity Streams

The fundamental shift proposed by researchers like Vibhaalakshmi Sivaraman and Shaileshh Bojja Venkatakrishnan in the Spider protocol (2018, 2020) is the "packetization" of value. Instead of trying to push a single large block of capital through the network, the sender breaks the transaction into many tiny "shares" or packets.

This decomposition solves the "shares issue" identified by Chang-Ting Lin and colleagues in Rapido (2018). In a typical PCN, a user might have enough total funds across various channels to pay a merchant, but no single path has the full capacity. Packetization allows the system to aggregate this fragmented liquidity from across the entire network, effectively bypassing individual channel bottlenecks.

Congestion Control for Money

If payments are packets, then payment channels are like network links with a unique constraint: bandwidth is finite and directional. When money flows from Alice to Bob, the capacity to send money back from Bob to Alice increases, but the capacity to send more to Bob decreases.

To manage this, researchers have adapted classic internet congestion control mechanisms to financial flows:

  1. Feedback Loops: In the Spider architecture, nodes monitor their local liquidity. If a channel is nearly depleted, the node signals this "congestion" back to the sender using binary marking or queue lengths as a proxy for imbalance.
  2. Dynamic Rate Control: As described in Splicer+ by Xuewen Dong and Sheng Gao (2025), systems can use window-based congestion control to adjust the rate of payment units. If a path is "congested" (low liquidity), the sender throttles the flow and redirects packets to more balanced paths.
  3. Liquidity Replenishment: Because packets can be routed independently over time, a channel that is temporarily depleted can be "refilled" by payments flowing in the opposite direction. This real-time replenishment allows the network to support a higher volume of transactions with much less locked collateral.

Redundancy and the Straggler Problem

Multi-path routing introduces a new challenge: what if 99% of the payment packets arrive quickly, but the last 1% gets stuck in a slow or congested channel? In a standard atomic setup, the whole transaction would hang.

Joachim Neu and Vivek Kumar Bagaria addressed this in Boomerang (2020) by introducing redundancy. Instead of sending exactly the amount needed, the sender transmits more partial payments than required. Using a cryptographic "reversible" contract, the transaction settles as soon as a sufficient quorum of packets reaches the destination. This "first-to-arrive" logic effectively bypasses straggling nodes and congested links, reducing latency by up to 40%.

Why It Works: The Intuition

The intuition behind this research direction is that liquidity is a dynamic resource, not a static one.

By moving away from "source routing" (where the sender pre-calculates a single fixed path) toward distributed, rate-controlled multi-path routing, the network transforms from a rigid set of pipes into a flexible fabric. Packetization allows the network to treat available funds as a dynamic bandwidth constraint, ensuring that capital is never sitting idle in one corner of the network while transactions fail in another.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_multi_path_packetized_routing_for_paymen_960e6d37,
  title = {Multi-path Packetized Routing for Payment Channel Networks},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/multi-path-packetized-routing-for-payment-channel-networks/rdr_960e6d37a9054223b82d03b0819522be},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-20},
}

This exact version (version 1, frozen text):

@misc{lacuna_multi_path_packetized_routing_for_paymen_960e6d37_v1,
  title = {Multi-path Packetized Routing for Payment Channel Networks},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/multi-path-packetized-routing-for-payment-channel-networks/rdr_960e6d37a9054223b82d03b0819522be/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-20},
}