Permissioned blockchain

On this page 7

A permissioned blockchain is a blockchain in which an access control layer decides who may participate: only identified, approved organizations run validating nodes, and read or write rights can be restricted as well. Among the Blockchain fundamentals, it is the main alternative to an open public blockchain, keeping the shared, append-only ledger and cryptographic linking of blocks while replacing anonymous participation with known, accountable members.

Permissioned networks are common in Crypto & blockchain projects built for enterprises and consortia, such as trade finance, supply chain tracking, interbank settlement and tokenized assets, where participants already have legal relationships and need privacy, predictable performance and regulatory compliance.

How a permissioned blockchain controls participation

A permissioned blockchain controls participation through a membership service that issues and verifies digital identities, usually X.509 certificates from one or more certificate authorities run by the consortium. Every node and every client signs its messages, and nodes reject messages from unknown identities, so there is no open admission as on a public network.

Because validators are known, the network does not need costly Sybil resistance such as proof of work. It can instead use voting-based consensus, in which a fixed set of validators exchanges signed messages and agrees on each block. Misbehavior is deterred by contracts, governance and the threat of removal rather than by economic penalties alone.

The table below summarizes the building blocks that distinguish a permissioned network from an open one.

ComponentRoleTypical implementation
Identity and membershipAdmits and identifies participantsConsortium certificate authorities
Permission policiesDefine who may read, write or validateRole and channel configuration
Consensus mechanismOrders transactions into blocksRaft or Byzantine fault tolerant protocols
Privacy layerLimits data visibility to involved partiesChannels, private data collections, private transactions
Governance processAdds members and changes rulesConsortium agreement plus on-chain voting
Building blocks of a permissioned blockchain

Consensus mechanisms used in permissioned networks

Permissioned networks choose consensus according to the trust assumptions among members. If all validators are trusted to follow the protocol and only crashes are expected, a crash fault tolerant protocol is sufficient and fast. If some members might act maliciously, a Byzantine fault tolerant protocol is required, which tolerates up to one third of faulty validators.

The following table compares the most common options and their practical properties.

Protocol familyFault modelToleratesFinalityTypical scale
RaftCrash faults onlyFewer than half of nodes failingImmediate3 to 7 ordering nodes
PBFT and variantsByzantine faultsFewer than one third maliciousImmediate4 to about 30 validators
IBFT and QBFTByzantine faultsFewer than one third maliciousImmediate4 to a few dozen validators
Proof of authorityTrusted signers in rotationDepends on the variantProbabilistic or fastSmall validator sets
Consensus options and their trade-offs

Two practical consequences follow. First, immediate finality means a confirmed transaction cannot be reorganized away, which suits settlement. Second, Byzantine fault tolerant protocols need at least four validators to tolerate one faulty node, and their message overhead grows quickly as the validator set expands.

Permissioned blockchain versus public blockchain and distributed ledger

The essential difference between a permissioned and a public blockchain is who decides membership. On a public blockchain anyone may run a node and validate; on a permissioned blockchain a governing body does. This choice shifts the network’s position in the blockchain trilemma, trading decentralization for throughput and control.

PropertyPublic blockchainPermissioned blockchainCentralized database
Who validatesAnyone meeting protocol rulesApproved membersOne operator
Participant identityPseudonymousKnown and verifiedKnown to the operator
Data visibilityPublic by defaultRestricted by policyPrivate to the operator
ThroughputLimited at the base layerHundreds to thousands of transactions per secondVery high
Censorship resistanceHighLow to moderateNone
Native tokenUsually requiredUsually absentNot applicable
Comparison with neighboring concepts

The terms are often used loosely. Distributed ledger is the broader category, covering any replicated ledger, including designs that do not chain blocks. A permissioned blockchain is one kind of distributed ledger. Some sources also separate private blockchains, run by a single organization, from consortium blockchains, governed by several; both are permissioned.

When a permissioned blockchain is the right choice

A permissioned blockchain makes sense when several organizations need a shared record but none should control it alone, participants must be identifiable, and some data must stay confidential. If one party is trusted by everyone, a conventional database is simpler and cheaper.

SituationRecommended choiceReason
One company records its own dataCentralized databaseNo need for multiparty agreement
Regulated firms share settlement dataPermissioned blockchainKnown parties, privacy and finality
Open participation and censorship resistance matterPublic blockchainNo gatekeeper decides membership
Consortium wants public auditabilityPermissioned chain anchored to a public chainPeriodic hashes give external proof
Decision guide for choosing a ledger architecture

Widely used frameworks include Hyperledger Fabric, which separates execution, ordering and validation and supports private channels, R3 Corda, which shares data only between the parties to a transaction, and permissioned Ethereum clients such as Hyperledger Besu, which keep Ethereum smart contract compatibility while restricting validators.

Limitations and common misconceptions

The main criticism is that a permissioned blockchain reintroduces trusted intermediaries. Security depends on the honesty of a small validator set and on the governance that admits them, so collusion among members can rewrite rules in ways a public network would resist.

  • Not automatically private: every member node may still see all data unless channels or private transactions are configured.
  • Governance is the hard part: onboarding, cost sharing, upgrades and dispute resolution often take longer than the technology.
  • Immutability is relative: a sufficient majority of validators can agree to change history or roll back.
  • Network effects are limited: value grows only as competitors agree to join, and many pilots stall for that reason.
  • Interoperability is weak: different frameworks rarely exchange data without custom bridges.

The idea in brief

A permissioned blockchain is a shared notebook that only invited members may use. Everyone keeps an identical copy, every entry is linked to the one before it, and nobody can quietly change an old page. The difference from public networks such as Bitcoin is the guest list: a group of organizations decides who may join, write and check entries.

It is like a members-only club with a jointly kept ledger. The members know each other, so they agree on new entries quickly without the energy-intensive puzzles of public networks, but the club must trust its own members and its own rules.

A practical scenario

Five banks and two logistics firms set up a trade finance network to replace emailed copies of bills of lading and letters of credit. Each bank runs a validating node, the consortium operates its own certificate authority, and a Byzantine fault tolerant protocol with seven validators tolerates up to two faulty or dishonest nodes.

Each shipment is recorded once, and its status changes as goods move and documents are signed. Private channels keep pricing visible only to the bank and the client involved, while every party sees the same shipment history. Because blocks reach immediate finality, a bank can release payment as soon as the arrival event is confirmed.

After a year, the consortium reports that document reconciliation fell from several days to under one hour per shipment. The hardest work was not the software but agreeing on data standards and on how new members would be admitted and pay their share.