The protocol runs among n parties in synchronized epochs over a synchronous network with authenticated broadcast and pairwise private authenticated channels; honest parties erase old shares after refresh.

1. threat model

   Mobile mixed adversary

   defined

   The adversary may eventually compromise all parties across the system lifetime, but the stated passive, active, and recovery-dependent thresholds must hold within each refresh period.

   Problem, claimed novelty, thresholds, and communication · Synchronous network, channels, epochs, refresh, and deletion assumptions
2. assumption

   Communication and reset assumptions

   assumed

   Synchrony, a global clock, authenticated broadcast, private authenticated channels, restoration to a pristine state, and deletion of obsolete shares are environmental assumptions rather than guarantees constructed by this scheme.

   Synchronous network, channels, epochs, refresh, and deletion assumptions
3. assumption

   Verifiability assumption

   computational assumption

   Active-fault checking uses homomorphic commitments instantiated in the announcement by Feldman VSS, whose hiding/security properties rely on discrete-log hardness over the selected field group.

   Secret sharing, verifiable sharing, proactive refresh, recovery, and discrete-log commitment assumption

DM-Share, DM-Reconstruct, DM-Refresh, and DM-Recover jointly implement sharing, reconstruction, proactive rerandomization, and replacement of shares lost by rebooted or faulty parties.

1. method

   Additive encoding with increasing degrees

   specified

   DM-Share writes s as the sum of n random summands and verifiably shares the summands with polynomials whose degrees increase across the family; reducing the maximum degree reserves interpolation capacity for share recovery.

   Additive-summand and increasing-degree polynomial blueprint · DM-Share and DM-Reconstruct
2. protocol

   DM-Refresh

   specified

   Each party verifiably distributes a family of random refreshing polynomials whose constant terms sum to zero; adding their evaluations rerandomizes every local share while preserving the encoded secret, after which old shares are deleted.

   DM-Refresh and preservation of the shared secret
3. protocol

   DM-Recover

   specified

   Non-recovering parties mask their current polynomial shares with random polynomials that vanish at the recovering party's point, allowing that party to interpolate only its replacement shares rather than the secret or current sharing polynomials.

   DM-Recover, privacy intuition, and recovery-dependent thresholds

The announcement states separate thresholds for passive corruption, active corruption, mixed corruption, lost-share recovery, and batched communication; these guarantees are not interchangeable.

1. claim

   Passive dishonest-majority confidentiality

   claimed with protocol intuition

   Without lost-share recovery, the paper states confidentiality for t < n passive corruptions in a refresh period; with e simultaneous lost shares or faults, the stated passive threshold becomes t < n - e.

   Problem, claimed novelty, thresholds, and communication · DM-Recover, privacy intuition, and recovery-dependent thresholds · Conclusion, non-robust active security, communication, and open questions
2. claim

   Active and mixed corruption bounds

   claimed non robust

   The non-robust construction states security for t < n/2 - e active corruptions and for mixed adversaries whose total may be a majority provided fewer than n/2 - e corruptions are active; detected cheating may cause abort.

   Problem, claimed novelty, thresholds, and communication · Conclusion, non-robust active security, communication, and open questions
3. claim

   Communication cost

   asymptotic claim

   The announcement reports O(n^4) communication for one secret and O(n^3) communication when multiple secrets are batched; it leaves lower communication for dishonest-majority PSS open.

   Problem, claimed novelty, thresholds, and communication · Conclusion, non-robust active security, communication, and open questions

The three-page source specifies the model, encoding blueprint, and protocol mechanisms and gives security intuition, but it does not contain the full definitions, games, lemmas, or complete proofs of the corresponding SCN paper.

1. evidence

   Mechanism-level support

   construction inspected

   The source explains why additive summands resist passive reconstruction, why zero-sum refresh polynomials preserve the secret, and why vanishing recovery masks reveal only the recovering party's replacement evaluations.

   Additive-summand and increasing-degree polynomial blueprint · DM-Refresh and preservation of the shared secret · DM-Recover, privacy intuition, and recovery-dependent thresholds

The active guarantee is non-robust, thresholds decrease with parallel recovery, the scheme assumes synchrony and secure erasure, and the brief leaves complete proofs and optimal communication outside its three pages.

1. limitation

   Open communication and asynchrony questions

   open problem

   The authors ask whether batched communication can fall below O(n^3) and note that no construction in the represented work attains dishonest-majority security up to n - 1 over asynchronous networks.

   Conclusion, non-robust active security, communication, and open questions