PaperlayerView exact revision on arXiv

Source-linked AI summary

Security and Privacy on Blockchain

Rui Zhang, Rui Xue, Ling Liu

arXiv:1903.07602v2cs.CR

TL;DR

Blockchain’s growing use has made security and privacy properties central research concerns. This survey organizes those properties and reviews techniques for achieving them, concluding that complex applications generally require context-dependent combinations of technologies rather than a single solution.

  • Problem

    Blockchain applications need security and privacy in open environments, but existing research has focused mainly on attacks or narrow countermeasures rather than comprehensive analysis.

  • Method

    The paper surveys blockchain concepts, inherent and additional security and privacy attributes, consensus algorithms, and techniques including hash storage, mixing, signatures, encryption, zero-knowledge proofs, and trusted execution environments.

  • Results

    The survey characterizes blockchain security and privacy attributes and reviews representative techniques for achieving them across blockchain-based systems and applications.

  • Takeaways & Limitations

    Security and privacy techniques should be selected according to application requirements and context, with multiple technologies potentially combined in complex blockchain systems.

  • Takeaways & Limitations

    Proof of Work is extremely inefficient because it combines high computational complexity with a low probability of successfully generating the proof.

Abstract

from arXiv · show

Blockchain offers an innovative approach to storing information, executing transactions, performing functions, and establishing trust in an open environment. Many consider blockchain as a technology breakthrough for cryptography and cybersecurity, with use cases ranging from globally deployed cryptocurrency systems like Bitcoin, to smart contracts, smart grids over the Internet of Things, and so forth. Although blockchain has received growing interests in both academia and industry in the recent years, the security and privacy of blockchains continue to be at the center of the debate when deploying blockchain in different applications. This paper presents a comprehensive overview of the security and privacy of blockchain. To facilitate the discussion, we first introduce the notion of blockchains and its utility in the context of Bitcoin like online transactions. Then we describe the basic security properties that are supported as the essential requirements and building blocks for Bitcoin like cryptocurrency systems, followed by presenting the additional security and privacy properties that are desired in many blockchain applications. Finally, we review the security and privacy techniques for achieving these security properties in blockchain-based systems, including representative consensus algorithms, hash chained storage, mixing protocols, anonymous signatures, non-interactive zero-knowledge proof, and so forth. We conjecture that this survey can help readers to gain an in-depth understanding of the security and privacy of blockchain with respect to concept, attributes, techniques and systems.

1 INTRODUCTION

Blockchain has expanded from Bitcoin into a broad distributed-computing technology, while its security and privacy remain central deployment concerns. This survey reviews blockchain properties, techniques, and open challenges for researchers and application designers.

  • Background: Blockchain is a distributed database that organizes transaction records into a cryptographically secured chain of blocks without centralized authority.It is maintained through a peer-to-peer overlay network and decentralized use of cryptography.
  • Motivation: Blockchain research has grown across industry, government, and academia, with applications and platforms spanning financial systems, smart contracts, and distributed ledgers.The passage identifies Bitcoin, Hyperledger, Ethereum, and FileCoin as representative efforts.
  • Research gap: Prior security and privacy studies mainly examined attacks or proposed countermeasures for subsets of attacks, leaving few in-depth analyses of blockchain properties and implementation techniques.The survey addresses this gap with a comprehensive review.
  • Contribution: The survey characterizes inherent and additional security and privacy attributes, reviews techniques for achieving them, and identifies open challenges.Its stated audiences include nonsecurity experts, specialists, and researchers.
  • Organization: The paper proceeds from blockchain concepts to security attributes, consensus algorithms, security and privacy techniques, and concluding observations.This organization supports both introductory understanding and exploration of advanced techniques.

2 OVERVIEW OF BLOCKCHAIN

Blockchain is a cryptographically protected, consensus-governed ledger that records transactions as an expanding chain of linked blocks. In Bitcoin, it provides a public archive designed to verify transactions and prevent double spending without central authority.

  • Bitcoin utility: In Bitcoin, the blockchain records and verifies transactions on an open peer-to-peer network while preventing double spending without trusted central authority.Transactions are stored in cryptographically secured blocks that are chained verifiably and persistently.
  • Blockchain concept: Blockchain organizes transaction records into a hierarchically expanding chain of cryptographically guarded blocks.New blocks enter the global chain only after completing a decentralized consensus procedure.
  • Blockchain concept: Each block stores its own hash and the preceding block’s hash, creating a cryptographic linkage across the chain.Consensus controls block admission, secure reading, and consistency of transaction data across copies.

2.1 How does the Blockchain Work

Bitcoin’s blockchain combines transaction broadcasting, verification, consensus, hash-linked storage, Merkle trees, and digital signatures to maintain a shared ledger. These mechanisms support integrity, authenticity, membership verification, and resistance to unauthorized changes.

  • Consensus: Bitcoin transactions are broadcast to network nodes, collected and verified by miners, and incorporated into blocks through a consensus protocol.Consensus seeks agreement on a single ledger state and helps prevent dishonest or malicious updates.
  • Hash Chained Storage: Hash-chained storage links each block to its predecessor and enables users to detect tampering by checking cryptographic hashes.Changing earlier data requires altering subsequent hash pointers, while the genesis block provides an initial reference point.
  • Merkle Tree: Merkle trees organize hashed nodes so that data membership can be verified by checking a node’s path to the root.The passage states that membership verification requires logarithmic time.
  • Digital Signature: Digital signatures use private keys to sign data and public keys to validate signatures, with secure schemes requiring verifiability and existential unforgeability.Bitcoin uses ECDSA over secp256k1, providing 128 bits of security according to the passage.
  • Digital Signature: Bitcoin users generate key pairs themselves, and hashes of their public keys serve as pseudonymous addresses without central public-key management.This decentralization supports pseudonymity but does not by itself establish a centrally registered identity.

2.2 Blockchain Level Transaction Models

Blockchain transaction models differ in how they represent ownership, balances, and transaction ordering. Bitcoin’s UTXO model offers granular outputs, while account-based models simplify state representation and support Ethereum’s mining-based ledger updates.

  • 2.2.1 The UTXO Model: The UTXO model represents a user’s balance as unspent transaction outputs, each associated with an owner and value.Combining outputs requires a new transaction that creates a consolidated output.
  • 2.2.1 The UTXO Model: UTXO transactions can provide potential privacy, scalability, and security benefits through selective disclosure, parallelizable outputs, and Merkle proofs of ownership.Bitcoin uses consensus confirmation and a universal ledger to resolve double spending.
  • 2.2.1 The UTXO Model: The UTXO model can leak information through change outputs and complicate balance calculation and wallet design.Strict input-output ordering also limits fully parallel transaction execution.
  • 2.2.2 Account-Based Online Transaction Model: Account-based models operate directly on sender accounts, improving consensus efficiency and block times at the cost of higher risk.They use account balances rather than individually tracked unspent outputs.
  • 2.2.2 Account-Based Online Transaction Model: Ethereum uses an account nonce to order transactions from an account and a proof-of-work nonce to satisfy the mining puzzle.Miners validate blocks by solving puzzles, broadcasting candidate blocks, and extending the chain selected by the network.
  • 2.2.2 Account-Based Online Transaction Model: Ethereum adjusts proof-of-work puzzle complexity to maintain approximately 12-second block validation.The passage describes routine adjustment when puzzles are solved faster or slower than the system-default rate.

2.3 CAP Properties in Blockchain

Blockchain systems must balance consistency, availability, and partition tolerance, while Bitcoin achieves these properties through eventual rather than simultaneous consistency.

  • CAP denotes consistency, availability, and partition tolerance as the three core properties of distributed systems.Consistency means nodes receive the latest write, availability keeps data accessible, and partition tolerance preserves operation despite node failures.
  • In a distributed ledger, consistency means identical updated ledgers, availability means accepting transactions, and partition tolerance means continued operation despite failed nodes.
  • Bitcoin does not achieve CAP properties simultaneously; consistency is reached after a delay while availability and partition tolerance continue.
  • Mining, consensus, and at least six confirmations provide Bitcoin’s mechanism for eventual consistency.

2.4 Classification and Evolution of Blockchains

Blockchains are classified by participation and write permissions, and their evolution has expanded from digital currencies toward smart contracts and broader decentralized collaboration.

  • Classification of Blockchains: Public blockchains allow anyone to read, submit transactions, join consensus, and participate in deciding which blocks are added.
  • Classification of Blockchains: Consortium blockchains restrict write permissions so only a pre-selected participant set can influence network decisions.
  • Evolution of Blockchains: Blockchain development progressed from digital currency systems to smart contracts and other decentralized collaborations.
  • Evolution of Blockchains: As applications expanded from blockchain 1.0 to blockchain 3.0, understanding blockchain security and privacy properties became more critical.

3 SECURITY AND PRIVACY PROPERTIES OF BLOCKCHAIN

Blockchain security and privacy properties address transaction risks including inconsistency, tampering, double-spending, service disruption, identity linkage, and data exposure. The paper distinguishes Bitcoin’s inherent protections from additional properties that many applications require.

  • Online transaction requirements include integrity, availability, double-spending prevention, minimal disclosure, and protection against authentication and identity-disclosure risks.
  • Basic Security Properties: Bitcoin’s signed, replicated transactions make undetected tampering practically impossible, supporting immutable audit trails and reliable patient-data records.
  • Basic Security Properties: Bitcoin’s decentralized construction and consensus allow transaction processing to continue when several nodes go offline, making large-scale DDoS attacks difficult.
  • Additional Security and Privacy Properties: Bitcoin provides pseudonymity but not unlinkability or full anonymity because publicly traceable addresses can connect transactions and reveal identities.
  • Additional Security and Privacy Properties: Bitcoin does not provide transaction confidentiality: transaction contents and addresses remain publicly viewable despite pseudonymous addressing.

4 CONSENSUS ALGORITHMS

Blockchain consensus requires coordinated agreement despite malicious actors and faulty processes. Current blockchain consensus algorithms provide probabilistic solutions to this agreement problem, with emphasis on their security and privacy properties.

  • Consensus protocols coordinate a group to reach agreement, but malicious actors and faulty processes can tamper with coordinated consensus.
  • Current blockchain consensus algorithms provide probabilistic solutions to the Byzantine Generals Problem.

4.1 Proof of Work (PoW)

Proof of Work reaches decentralized agreement by making valid block production computationally difficult and verification easy, while hash chaining and consensus make transaction tampering costly. Its security depends on economic incentives, honest-majority assumptions, and resistance to majority-hash-power attacks.

  • Consensus mechanism: PoW addresses the Byzantine Generals Problem by using computational work to coordinate agreement on transaction validity.A valid proof is difficult and time-consuming to produce but easy for others to verify.
  • Consensus mechanism: Bitcoin repeatedly appends a nonce, hashes the augmented message, and iterates until the result satisfies a preset threshold.Finding the nonce is computationally expensive, while recipients can verify the resulting hash condition.
  • Protocol design: Bitcoin limits block creation to roughly one block every 10 minutes by adjusting PoW difficulty according to observed solving time.The protocol also makes it harder to predict which miner will generate the next block.
  • Limitations: PoW has major limitations: high computational inefficiency, dependence on mining rewards and an honest majority, and potential concentration of hashing power.The survey reports that more than 70% of hash rate was divided among five mining farms in the cited 2018 snapshot.
  • Security incentives: PoW security relies on anti-centralization and economic incentives, including block rewards and computational costs that discourage cartels and collusion.Persistence and liveness additionally depend on an honest majority and sufficient chain depth represented by ω.
  • Tamper resistance: Hash chaining blocks backward and redoing PoW for subsequent blocks make retrospective transaction tampering difficult.Backward attacks require altering hashes back to the genesis block; forward attacks require rebuilding later blocks and producing a longer winning chain.

4.2 Proof of State (PoS)

Proof of Stake replaces PoW’s computational competition with stake-based validator selection and penalties or rewards. The survey presents multiple PoS designs that address centralization, chain splits, adversarial control, and persistence or liveness.

  • PoS principles: PoS breaks the dependency on mining rewards by using stake-based participation and penalty-oriented security mechanisms.Validators qualify by placing bets, while reverting transactions can incur penalties much larger than interim rewards.
  • Chain-based PoS: Chain-based PoS pseudo-randomly selects a validator per time slot to create and link a block, allowing blocks to converge into one growing chain.Early reward-only designs created a “nothing at stake” problem because validators could support multiple competing chains without additional cost.
  • Chain-based PoS: Explicit penalties in newer PoS protocols deter validators from creating conflicting blocks on multiple chains.Casper is presented as a representative penalty-based PoS protocol using stake deposits, rewards, and transaction fees.
  • Validator selection: Coin age constrains validation competition by requiring coins to remain unspent for a defined period and resetting their age after signing.With lower and upper age limits, the approach makes large-stake domination more difficult.
  • PoS protocols: Snow White provides provable security under sporadic participation and past-member corruption when money does not switch hands too quickly.The protocol can attain security even when an adversary controls a minority of the stake.
  • PoS protocols: Ouroboros defines persistence and liveness guarantees and uses rewards to show that honest behavior approximates Nash equilibrium against adversarial strategies such as selfish mining.The protocol is described as making authentic transactions approved and permanent.

4.3 BFT based Consensus Algorithms

BFT consensus addresses agreement among distributed blockchain nodes despite Byzantine failures without relying on a central authority. The survey reviews classical, randomized, and asynchronous approaches, while noting that consensus suitability depends on the application and network setting.

  • BFT foundations: Byzantine fault tolerance is the ability of a system to reach agreement despite Byzantine failures, assuming a majority of participants are honest.Each participant may begin with a different value, but the protocol must produce one agreed value.
  • BFT foundations: Blockchain applies BFT without a central authority by having network nodes maintain replicated ledgers and verify candidate blocks before insertion.Candidate blocks serve as the values that participants collectively validate and agree to place in the chain.
  • Application scope: Neither PoW nor PoS is a perfect solution to BFT in decentralized peer-to-peer systems, and existing protocols may not fit every blockchain application.The survey highlights healthcare blockchain as an example requiring attention to application-specific feasibility.
  • BFT protocols: PBFT was proposed for high-performance Byzantine state-machine replication, extending the family of practical BFT protocols.The survey places PBFT within the historical development of Byzantine fault-tolerant algorithms.
  • BFT protocols: AlgoRAND reduces participation to a cryptographically selected subset of users through random sortition while providing player replaceability in adversarial settings.Selecting fewer officials is intended to avoid involving every system user in the protocol.
  • BFT protocols: HoneyBadgerBFT provides practical asynchronous BFT without timing assumptions, with throughput close to available network bandwidth under unstable conditions.Its broadcast protocol achieves activity and optimal asymptotic efficiency without careful parameter tuning.

4.4 Other Consensus Algorithms

Other consensus algorithms address different blockchain settings through mechanisms such as changing participant availability, trusted execution, and validator reputation. Their security and performance properties depend on assumptions including honest-majority participation, trusted hardware, timing models, and reputation-based authorization.

  • Sleepy Consensus: Sleepy consensus supports participants that switch between awake and asleep modes and is resilient when honest participants are the majority.Its protocols use PKI and, in one construction, weak synchrony and collision-resistant hashing, but support only static corruptions and a static online schedule.
  • Proof of Elapsed Time (PoET): PoET uses SGX to provide fairness and low computing consumption by selecting the node with the shortest randomly chosen waiting period to create a block.SGX generates a verifiable proof that the winner selected a random waiting time and completed it.
  • Proof of Elapsed Time (PoET): SGX is not entirely reliable because resourceful strategic adversaries may attack blockchain protocols implemented on the platform.Proposed countermeasures alter the probability distribution or statistically reject blocks generated by a given fraction of nodes.
  • Proof of Authority (PoA): PoA restricts transaction and block approval to validators who accumulate reputation, while preventing validators from approving two consecutive blocks.These mechanisms are presented as incentives for honest verification and as protection against centralized trust.
  • Proof of Reputation (PoR): PoR extends PoA by calculating reputation through predefined formulas and voting verified high-reputation nodes into the network as authoritative nodes.After admission, the authoritative node operates like a PoA validator.

4.5 Comparison of Consensus Algorithms

Consensus algorithms are selected according to blockchain deployment constraints. Strong-consistency protocols target smaller private or consortium networks, whereas eventual-consistency protocols target large public blockchains where complete consistency is difficult.

  • Strong Consistency Consensus: Strong-consistency algorithms such as BFT, PBFT, Paxos, and Raft are mainly used in private and consortium blockchains with relatively few nodes and stronger consistency requirements.BFT and PBFT address Byzantine faults, while Paxos and Raft do not consider Byzantine failure.
  • Eventual Consistency Consensus: Eventual-consistency algorithms such as PoW, PoS, and DPoS are used in public blockchains with many nodes where 100% consistency and correctness are difficult to achieve.The paper states that consensus selection is highly related to the application scenario.

5 PRIVACY AND SECURITY TECHNIQUES USED IN BLOCKCHAIN

Blockchain security and privacy techniques address anonymity, access control, confidentiality, and smart-contract protection through mixing, anonymous signatures, encryption, secure computation, and trusted execution. The survey emphasizes that no single technique is universally sufficient, so deployments should combine techniques according to application requirements while managing defects and efficiency trade-offs.

  • Mixing: Bitcoin transactions are publicly verifiable but pseudonymous, allowing address analysis to relate transactions and limiting user anonymity.Mixing obfuscates coin ownership from observers, but the discussed mixing services do not protect against coin theft.
  • Mixing: Mixcoin expands the anonymity set for simultaneous users, addresses passive and active adversaries, and adds accountability to detect stealing.It is designed for anonymous payments in Bitcoin and bitcoin-like cryptocurrencies.
  • Mixing: CoinJoin combines users’ payments into one transaction to reduce the probability of linking inputs and outputs or tracing a user’s money movement.Early centralized implementations required trust in the operator and could leak privacy through transaction logs recording participants.
  • Mixing: Incorrect CoinJoin implementations can weaken anonymity, as the CoinJoin Sudoku tool identified SharedCoin transactions and relationships between payments and payees.The cited analysis concluded that SharedCoin could not provide strong transaction privacy.
  • Mixing: CoinShuffle extends CoinJoin with a decentralized mixing protocol that avoids a trusted third party and is claimed to provide security against theft.It uses the accountable anonymous group communication protocol Dissent to support anonymity.
  • Anonymous Signatures: Group signatures hide a signer’s identity while allowing verification of group membership, but require a group manager for membership and dispute handling.This setup makes group signatures suitable for consortium blockchains, including platforms such as JUZIX.
  • Anonymous Signatures: Ring signatures provide signer anonymity without a group manager or additional setup, making them applicable to public blockchains.In CryptoNote, a ring of n members gives an adversary a 1/n probability of guessing the real sender.
  • Attribute-Based Encryption: ABE has seen few applications because core concepts and efficient implementation remain poorly understood, and it had not been deployed for real-time blockchain operation.A decentralized ABE scheme proposed representing permissions through ownership of access tokens.

6 CONCLUDING REMARKS

The survey characterizes blockchain security and privacy attributes and reviews techniques for achieving them in blockchain systems and applications. It concludes that deeper understanding of these properties is important for trust and for developing robust defenses, while noting that only a small part of platforms achieves the stated security goals in practice.

  • Contributions: The survey divides blockchain security and privacy attributes into inherent and additional categories for online transactions.It also reviews consensus algorithms, mixing, anonymous signatures, encryption, secure multiparty computation, non-interactive zero-knowledge proofs, and smart-contract verification.
  • Conclusion: Only a small part of blockchain platforms can achieve the survey’s stated security goals in practice.The authors argue that in-depth understanding of blockchain properties supports greater trust and technological innovation in defense techniques and countermeasures.
Loading 1903.07602v2…