Skip to content

Rotating Reference-Peer Protocol

Protocol — instantiates Decentralized Phase Locking

Temporarily lets one peer act as the timing reference to bootstrap a shared phase, then rotates and removes the role so no permanent master forms.

Purely mutual coupling can struggle to start from cold — an incoherent population may take a very long time to find any shared phase at all. Rotating Reference-Peer Protocol offers a scaffold: let one peer act briefly as the timing reference the others entrain to, then hand the role on and require that it be removable without collapse. Its defining safeguard is what keeps it decentralized despite using a reference at all — the role is temporary, auditable, and rotating. At any instant someone may be the reference, but no one stays the reference, and the population must prove it can hold phase once the reference is gone. That is the line between a helpful bootstrap and the very hidden-master, single-clock arrangement the archetype exists to avoid.

Example

A cycling paceline wants a smooth shared cadence, but a permanent leader is both a single point of failure and unfair — the front rider does all the work into the wind and tires first. So the front rotates: one rider sets the pace briefly, the group entrains to it, then that rider peels off and drifts to the back while the next takes over. At every moment there is a reference, but the role circulates continuously, and the paceline's rhythm survives any single rider dropping out. The protocol also fixes who is in the paceline and who is eligible to take the front — the population boundary — so a temporary local reference can never be mistaken for a fixed outside master.

How it works

What distinguishes a rotating reference from a leader:

  • Temporary election. A reference is chosen by rotation, token, or rule for a bounded turn — never open-ended.
  • Entrain, then hand off. The population locks toward the current reference, which then rotates on schedule or trigger; influence circulates rather than concentrating.
  • Removability is required and tested. The protocol periodically drops the reference and checks the population still holds — proving it is a scaffold, not a dependency.

Tuning parameters

  • Rotation cadence — how long each peer holds the reference. Longer gives a steadier bootstrap but drifts toward master-like; shorter is more clearly decentralized but choppier.
  • Reference influence strength — how strongly the reference pulls. Strong bootstraps fast but entrenches; weak preserves autonomy but starts slowly.
  • Eligibility rule — who may take the role — the most accurate clock, a round-robin, a stake-weighted draw.
  • Removal test — how often the reference is dropped to confirm the population no longer depends on it.
  • Handoff overlap — a clean swap versus briefly overlapping references to smooth the transition.

When it helps, and when it misleads

Its strength is bootstrapping coherence from a cold or fragmented start — or re-seeding after a mass disruption — without installing a standing master. It gives a decentralized population the one thing pure mutual coupling lacks: a fast, legitimate way to begin.

It misleads when "temporary" quietly becomes permanent. If rotation stalls or the reference's influence stays too strong, the scaffold silently becomes the de facto single clock — the hidden command path the archetype warns against — and a compromised reference can steer the whole population during its turn.[1] The classic misuse is the reference held "just for now" that never rotates. The discipline is to enforce rotation, cap the reference's influence, and actually run the removal test, so the population demonstrably holds phase with the reference gone.

How it implements the components

  • rotating_reference_peer — the protocol is this component: a reference role made temporary, auditable, and rotated so timing influence never permanently concentrates in one unit.
  • oscillator_population_boundary — it fixes who participates and who is eligible to hold the reference, so a temporary local reference is never confused with a hidden fixed master or common forcing.

It does not compute the ongoing peer corrections that sustain the lock — that is Nearest-Neighbor Phase Nudging — nor bridge separated clusters, which is bridge_oscillator_link; it seeds the phase, hands off, and gets out of the way.

Notes

The removal test is the whole crux. A reference that is never dropped and re-tested cannot be shown to be temporary — and an untested "rotating" reference tends to ossify into a master by default. Treat the periodic drop-and-check as mandatory, not optional: it is the only evidence that the lock is genuinely the population's and not the reference's.

References

[1] A single point of failure is a component whose loss halts the whole system. A permanent timing reference is exactly that; rotating the role and proving the population survives its removal is what converts a bootstrap aid from a liability back into a scaffold. ↩