PaperlayerView exact revision on arXiv

Source-linked AI summary

A Review on the Use of Blockchain for the Internet of Things

Tiago M. Fernandez-Carames, Paula Fraga-Lamas

arXiv:2402.00687v1cs.CR

TL;DR

IoT needs authentication, privacy, security, robustness, deployment, and maintenance capabilities as interconnected devices proliferate. This paper reviews blockchain fundamentals, BIoT applications, and the adaptations needed for IoT design and deployment. It identifies practical limitations and proposes research and development recommendations for future BIoT systems.

  • Problem

    IoT applications must address authentication, privacy, security, attack robustness, deployment, and self-maintenance as interconnected devices expand.

  • Method

    The paper conducts a holistic review of blockchain-based IoT architectures, applications, optimization aspects, practical limitations, and future recommendations.

  • Results

    The review identifies adaptations needed across BIoT design, including architecture, cryptography, consensus, privacy, security, energy efficiency, throughput, and adoption.

  • Takeaways & Limitations

    Future BIoT deployments require targeted optimization and further research before the next generation of applications can be deployed.

Abstract

from arXiv · show

The paradigm of Internet of Things (IoT) is paving the way for a world, where many of our daily objects will be interconnected and will interact with their environment in order to collect information and automate certain tasks. Such a vision requires, among other things, seamless authentication, data privacy, security, robustness against attacks, easy deployment, and self-maintenance. Such features can be brought by blockchain, a technology born with a cryptocurrency called Bitcoin. In this paper, a thorough review on how to adapt blockchain to the specific needs of IoT in order to develop Blockchain-based IoT (BIoT) applications is presented. After describing the basics of blockchain, the most relevant BIoT applications are described with the objective of emphasizing how blockchain can impact traditional cloud-centered IoT applications. Then, the current challenges and possible optimizations are detailed regarding many aspects that affect the design, development, and deployment of a BIoT application. Finally, some recommendations are enumerated with the aim of guiding future BIoT researchers and developers on some of the issues that will have to be tackled before deploying the next generation of BIoT applications.

I. INTRODUCTION

IoT is expanding toward interconnected devices, but centralized cloud architectures face growing costs, maintenance, privacy, and trust challenges. The paper reviews how blockchain can support decentralized IoT and analyzes BIoT architectures, applications, optimizations, and challenges.

  • Motivation: IoT growth requires architectures that can support increasingly many interconnected devices and services.Forecasts cited in the introduction include 26 billion IoT devices by 2020 and 3.3 billion M2M connections by 2021.
  • Motivation: Current IoT solutions commonly rely on centralized cloud servers, but expected growth motivates alternative architectural paradigms.Earlier decentralized proposals lacked some privacy and security capabilities later associated with blockchain.
  • Blockchain for IoT: Blockchain can track, coordinate, transact, and store information across many devices without requiring a centralized cloud.This capability is presented as relevant to IoT adoption challenges involving infrastructure costs, maintenance, privacy, and trust.
  • Paper scope: The paper reviews blockchain fundamentals, BIoT applications, architecture and deployment optimizations, practical limitations, and future recommendations.Its stated contribution is a holistic analysis rather than a narrow overview of blockchain basics or applications.

II. BLOCKCHAIN BASICS

A blockchain is a shared, cryptographically linked ledger maintained by network peers through transaction validation and consensus. Its variants differ in access, authorization, managed data, and supported logic.

  • Blockchain concept: A blockchain is a distributed ledger whose data are shared across network peers and appended through consensus.Bitcoin’s blockchain addressed the double-spend problem by seeking agreement among mining nodes.
  • Basic functioning: Nodes sign transactions with private keys, broadcast them to peers, and rely on validators or miners to order valid transactions into timestamped blocks.Peers verify transactions before retransmitting them, while consensus determines block formation.
  • Basic functioning: Nodes verify each broadcast block’s transactions and hash reference before adding it to their local chain.Blocks failing either validity or linkage checks are discarded.
  • Blockchain types: Blockchains may be public or private and permissionless or permissioned, representing distinct dimensions of access and authorization in IoT.The paper cautions that these terms are sometimes conflated, although IoT applications benefit from distinguishing authentication from authorization.
  • Blockchain types: Some blockchains track digital assets, whereas others support smart contracts or tokens without necessarily implementing a cryptocurrency.The taxonomy includes systems such as Bitcoin, Ethereum, Ripple, and Hyperledger.

C. DETERMINING THE NEED FOR USING A BLOCKCHAIN

Blockchain is not universally appropriate for IoT; its use should follow the application’s requirements for decentralization, peer exchange, transaction handling, logging, robustness, and traceability. The review then surveys representative BIoT applications and smart-contract uses.

  • Blockchain suitability: Developers should first determine whether an IoT application needs blockchain-specific properties rather than assuming blockchain is the best ledger.Traditional databases or DAG-based ledgers may better fit some scenarios.
  • Decision criteria: Blockchain is most relevant when applications require decentralization without a trusted central system and peer-to-peer exchanges.Mutual trust or predominantly gateway-mediated communication can reduce the need for blockchain.
  • Decision criteria: Payment systems, sequential transaction logging, and robust distribution alone do not necessarily justify blockchain adoption.Traditional payment systems, databases, and distributed computing can provide these functions under some conditions.
  • Decision criteria: Micro-transaction collection may require recording every transaction for traceability or auditing, while other applications can aggregate readings into fewer transactions.The paper notes that sidechains may help in some high-volume collection scenarios.
  • BIoT applications: Smart contracts are autonomous decentralized code that executes when specified conditions are met, supporting applications such as supply-chain monitoring and device management.Examples include temperature-range checks for pharmaceutical shipments and remotely configurable IoT devices.
  • BIoT applications: Reported BIoT applications span sensing, identity, timestamping, smart living, transportation, wearables, supply chains, agriculture, healthcare, cities, industry, and security.The review presents these as application areas rather than a single standardized deployment model.

IV. DESIGN OF AN OPTIMIZED BLOCKCHAIN FOR IOT

Adapting blockchain to IoT requires optimizing components beyond the consensus algorithm, with particular attention to performance and scalability. Existing evaluations examine consensus behavior as network size increases and compare PoW with BFT approaches.

  • Optimization scope: BIoT optimization studies analyze several influential blockchain components, but primarily focus on consensus algorithms.The paper frames adaptation as necessary because blockchain was not designed specifically for IoT environments.
  • Consensus performance: PBFT evaluations with up to 100 peers measure average consensus time and show it increasing as the number of peers grows.This pattern raises the possibility that PBFT becomes a bottleneck in larger networks.
  • Consensus performance: Comparisons of PoW and BFT consensus methods identify poor scalability as a limitation of Bitcoin-style consensus.The reviewed work suggests combining PoW with BFT and implementing consensus protocols in hardware to improve performance.

A. ARCHITECTURE

BIoT architectures must accommodate high IoT traffic while reducing dependence on centralized clouds and supporting distributed, low-latency processing. Proposed designs range from cloud, fog, and edge arrangements to hierarchical, layered, and lightweight blockchain architectures.

  • Architecture requirements: BIoT architectures must be adapted to the traffic generated by IoT applications rather than directly using traditional cloud-based designs.The paper identifies traffic volume as an architectural concern and presents cloud evolution toward edge and fog computing.
  • Architecture vulnerabilities: Cloud-centered IoT architectures have a single point of failure, while blockchain systems distribute validation and can reject updates from compromised devices.Central-cloud failure can stop the whole system; cryptographic verification supports rejecting malicious blockchain updates.
  • Edge and fog architectures: Edge and fog computing offload processing from the cloud to support distributed, low-latency, QoS-aware applications and reduce network traffic.Fog uses local gateways for fast responses, while edge architectures may include cloudlets for high-speed computation.
  • Alternative BIoT designs: Proposed BIoT designs include lightweight three-layer home-automation architectures, hierarchical multilayer blockchains, and blockchain deployment across ecosystem levels.These approaches target reduced communication overhead, heterogeneous-device connectivity, or lower deployment complexity.
  • Resource constraints: Resource-constrained IoT devices motivate cloud and fog architectures because hosting a regular blockchain directly on such devices is difficult.One evaluated design used Intel Edison IoT nodes and IBM Bluemix as the blockchain technology.
  • Alternative BIoT designs: Other explored approaches include SDN-controlled fog nodes, hierarchical distributed automation platforms, and architectures spanning centralized to fully distributed designs.The surveyed alternatives vary in how computation, storage, and blockchain services are distributed.

B. CRYPTOGRAPHIC ALGORITHMS

BIoT cryptography must balance security and privacy with the limited computation, memory, and energy budgets of IoT devices. The review therefore contrasts RSA and ECC and identifies efficiency requirements for hash functions.

  • Public-key cryptography: Resource-constrained IoT devices struggle with modern public-key cryptography, particularly because RSA is slow and power consuming on such hardware.Cryptographic selection must account for computational load, memory requirements, and energy consumption.
  • Public-key cryptography: Common public-key cipher suites include RSA and ECDHE, both recommended by NIST for TLS.RSA suites use RSA for key exchange, whereas ECDHE suites use ephemeral Diffie-Hellman over elliptic curves.
  • Public-key cryptography: 2048-bit RSA keys are secure but impose heavy certificate and computation overhead that most IoT nodes cannot easily accommodate.Smaller 768-bit and 1024-bit RSA implementations were broken in 2010, making them unsuitable as secure alternatives.
  • Public-key cryptography: ECC is presented as a lighter alternative to RSA that has shown better speed and power consumption on resource-constrained devices.The passage also notes a 2015 NSA recommendation to stop using Suite B because of progress in quantum cryptography.
  • Hash functions: IoT hash functions must avoid collisions, execute quickly, and consume minimal energy because blockchain transactions require hashing for signatures.The review discusses SHA-256d, SHA-256, and Scrypt as widely used blockchain hash functions.

C. MESSAGE TIMESTAMPING

Blockchain timestamping links transactions through hashes and can distribute timestamping to avoid a single failure point, but decentralizing the service across IoT devices remains insufficiently studied. Consensus mechanisms address agreement and attack resistance through different trade-offs in energy, latency, participation, and fault tolerance.

  • Message timestamping: Hash-linked timestamps preserve transaction order and make inserting fake transactions into the chain difficult, while distributed timestamping avoids a single point of failure.The passage notes that Bitcoin uses block linking and Proof-of-Work to address Sybil attacks in distributed timestamping.
  • Message timestamping: Decentralized timestamping among IoT devices remains insufficiently studied despite proposals for decentralized services and distributed timestamping keys.This identifies a specific research boundary rather than a resolved deployment strategy.
  • Consensus: Consensus determines when agreement has been reached on validating and adding blocks, corresponding to the Byzantine Generals Problem.The review frames consensus as central to proper blockchain operation.
  • Consensus: Proof-of-Stake reduces computational power and energy use relative to Proof-of-Work, while Delegated Proof-of-Stake speeds validation by involving fewer nodes.DPoS also allows dishonest delegates to be replaced and permits adjustment of block parameters.
  • Consensus: PBFT tolerates fewer than one-third malicious nodes and requires at least two-thirds support for the selected leader’s block ordering.Its participants must be known by the network, making the assumption explicit.
  • Consensus: Ripple reduces latency by relying on trusted subsets, whereas SCP waits for important subsets rather than querying every participant.Both approaches reduce the communication or participation required for consensus.
  • Consensus: Tendermint commits a block only after more than two-thirds of validators pre-commit it and tolerates up to one-third failures.Validation occurs through pre-vote and pre-commit stages.
  • Consensus: Bitcoin-NG experiments with 1,000 nodes found scalability limited by node bandwidth and network propagation latency.The reported conclusion concerns the tested Bitcoin-NG setting.

PROTOCOL STACK

Blockchain-enabled IoT may require protocol-stack adaptations, including the five-layer Internet of Money stack. However, the need for specialized stacks and their performance relative to traditional OSI stacks remain open research questions.

  • Protocol stack: IoT firmware updates can be distributed securely through blockchain rather than relying only on manual or Over-The-Air updates.The proposed use targets bug correction, attack prevention, and functionality improvements across large device deployments.
  • Protocol stack: The Internet of Money proposes five layers operating over TCP/IP to adapt protocol-stack organization to blockchain technologies.Its layers are Ledger, Payment and Exchange, Pathfinding, Contract, and Application.
  • Internet of Money layers: The Ledger Layer creates ledgers and issues assets, while the Payment and Exchange Layer handles payment and exchange functions.These layers separate ledger creation from value-transfer operations.
  • Internet of Money layers: The Pathfinding Layer calculates atomic operations for a desired transfer or exchange, and the Contract Layer controls balances through running code.The Application Layer supports application and user-interface development.
  • Open questions: More research is needed to determine whether specific stacks are necessary and to compare their performance with traditional OSI-based stacks.The paper presents this as an unresolved protocol-stack question.

V. CURRENT CHALLENGES FOR BIOT APPLICATIONS

BIoT applications face interconnected challenges involving privacy, identity, access management, robustness, and the computational cost of privacy-preserving mechanisms. The review surveys blockchain-based approaches while noting important limitations for resource-constrained IoT devices.

  • BIoT development is complex because IoT technologies, especially in mission-critical scenarios, already face several technical challenges.Adding blockchain introduces further operational and technical requirements.
  • Privacy: Blockchain transparency complicates privacy because shared transactions can expose user or device identities and reveal sensitive behavioral data.IoT data may be stored on blockchains whose privacy requirements differ across countries.
  • Identity management: Permissioned blockchains can distribute identity management, rotate locally generated asymmetric keys, and improve protection against attacks.These systems address concerns about centralized identity providers blocking authorized entities.
  • Identity management: Blockchain-based smart-home systems can automate authentication by extracting appliance signatures for scalable, heterogeneous, and mobile IoT environments.
  • Privacy-preserving mechanisms: Mixing techniques and fresh transaction addresses can make data analysis more difficult, although mixing services require trust and can be de-anonymized.
  • Privacy-preserving mechanisms: Zero-knowledge proofs and homomorphic encryption can preserve privacy during authentication or transaction processing, but some techniques remain vulnerable or computationally demanding.Zero-knowledge proofs can face statistical disclosure attacks, while homomorphic encryption lets services process data without seeing it unencrypted.

B. SECURITY

Blockchain security is framed through confidentiality, integrity, and availability. The review describes supporting mechanisms but emphasizes that centralized dependencies, exceptional reversibility, and majority attacks remain relevant concerns.

  • Information-system security requires confidentiality, integrity, and availability, while centralized IoT infrastructure can concentrate access to stored data.
  • Confidentiality: Private-key management is central to confidentiality because attackers can use a private key with a public key to impersonate users or steal assets.
  • Confidentiality: Certificate authorities provide trust to third parties but have failed in some cases, motivating distributed monitoring and auditing frameworks.Certificate Transparency is described as a near-real-time monitoring and auditing approach.
  • Integrity: Blockchain foundations make stored information difficult to alter, but the MintPal incident shows that hard forks can exceptionally reverse recorded damage.
  • Availability: Distributed blockchain designs support continued operation when some nodes are attacked, but a 51-percent attack can block transactions and compromise integrity.

C. ENERGY EFFICIENCY

Energy efficiency is a major BIoT challenge because end-nodes are battery-powered while blockchain mining and peer-to-peer communication can consume substantial energy. The review discusses alternative consensus, communication, hashing, and scaling strategies, alongside throughput and latency constraints.

  • IoT end-nodes are typically battery-powered, making energy efficiency essential for long-lasting deployments.Blockchain energy consumption is driven mainly by mining and peer-to-peer communications.
  • Consensus: Proof-of-Space, Proof-of-Stake, and Practical Byzantine Fault Tolerance are proposed as alternatives that can reduce energy consumption relative to Proof-of-Work.
  • Communication and storage: Mini-blockchains can reduce communication updates and computational requirements by allowing IoT nodes to interact with a blockchain while retaining only recent transactions.
  • Hashing: Faster hashing algorithms such as Scrypt and X11 may reduce mining energy consumption, but their performance on IoT devices requires further analysis.
  • Throughput and latency: 7 transactions per second is Bitcoin’s stated theoretical maximum, creating a throughput limitation for IoT deployments requiring many transactions per time unit.VisaNet is cited as handling up to 24,000 transactions per second.
  • Throughput and latency: Bitcoin block creation has a 10-minute mean, and merchants may wait about an hour for confirmation, whereas VISA latency requires only a few seconds.

INFRASTRUCTURE

BIoT infrastructure must accommodate blockchain storage, bandwidth, adoption, usability, interoperability, governance, and legal enforcement. The review highlights that IoT constraints make traditional blockchain designs difficult to deploy broadly.

  • Storage and bandwidth: Traditional blockchains can exceed IoT storage capacity because they grow over time and require nodes to retain data irrelevant to them.
  • Storage and bandwidth: Mini-blockchains address storage growth by retaining an account tree and only the most recent transaction for each user.
  • Storage and bandwidth: Transaction and block sizes must match IoT bandwidth limits because many small transactions increase communication energy, while large payloads may exceed device capabilities.
  • Infrastructure: BIoT infrastructure requires decentralized storage, communication protocols, mining hardware, address management, and network administration.Mining hardware has evolved from CPUs toward GPUs, FPGAs, and ASICs.
  • Adoption: Pseudoanonymity may hinder adoption because governments may require stronger links between real-world and online identities, especially for international IoT transactions.
  • Adoption and usability: Miner adoption affects transaction capacity, while user-friendly APIs are needed to simplify blockchain and account management for developers.
  • Governance and interoperability: BIoT deployments may need multiple chains for different data types, but forks complicate cross-chain transactions and miners can censor transactions under certain conditions.
  • Governance and interoperability: Legal rules for enforcing smart contracts and resolving disputes remain underdeveloped.

VI. FURTHER CHALLENGES AND RECOMMENDATIONS

Future BIoT systems require coordinated technical, infrastructural, organizational, and testing efforts. The review recommends evaluating requirements and blockchain choices systematically while recognizing that no single solution fits every application.

  • Technical challenges: BIoT development still faces scalability, security, cryptographic, stability, transaction-capacity, validation, smart-contract, and centralization challenges.
  • Interoperability and standardization: Interoperability with legacy systems requires stakeholder agreement, collaborative implementations, and international standards for data, policy, access control, authentication, and authorization.
  • Blockchain infrastructure: A comprehensive trust infrastructure is needed to satisfy IoT requirements, including support for inter-domain policies and public-interest use cases.
  • Governance and regulation: Regulatory and legal frameworks must address decentralized ownership and international jurisdiction, while governance should account for risks such as falsified blockchain performance claims.
  • Rapid field testing: Blockchain selection should proceed through standardization and testing against agreed criteria covering requirements such as privacy, security, and energy efficiency.
  • Recommendations: The review evaluates practical limitations, studies architecture, cryptography, and consensus, and offers recommendations before next-generation BIoT deployment.
  • Conclusion: Because BIoT applications differ across stakeholders and requirements, the review concludes that no one-size-fits-all solution exists.
Loading 2402.00687v1…