PoAh: A Novel Consensus Algorithm for Fast Scalable Private Blockchain for Large-scale IoT Frameworks
Deepak Puthal, Saraju P. Mohanty, Venkata P. Yanambaka, Elias Kougianos
PPoAh : A Novel Consensus Algorithm for Fast Scalable PrivateBlockchain for Large-scale IoT Frameworks
January 22, 2020
Deepak Puthal Saraju P. MohantySchool of Computing Dept. of Computer Science and EngineeringNewcastle University, UK University of North Texas, USA.Email:
Email: [email protected]
Venkata P. Yanambaka Elias KougianosSchool of Science and Engineering Department of Electrical EngineeringCentral Michigan University, USA. University of North Texas, TX 76203.Email: [email protected]
Email: [email protected]
Abstract
In today’s connected world, resource constrained devices are deployed for sensing and decisionmaking applications, ranging from smart cities to environmental monitoring. Those recourse constraineddevices are connected to create real-time distributed networks popularly known as the Internet of Things(IoT), fog computing and edge computing. The blockchain is gaining a lot of interest in these domains tosecure the system by ignoring centralized dependencies, where proof-of-work (PoW) plays a vital roleto make the whole security solution decentralized. Due to the resource limitations of the devices, PoWis not suitable for blockchain-based security solutions. This paper presents a novel consensus algorithmcalled Proof-of-Authentication (PoAh), which introduces a cryptographic authentication mechanism toreplace PoW for resource constrained devices, and to make the blockchain application-specific. PoAhis thus suitable for private as well as permissioned blockchains. Further, PoAh not only secures thesystems, but also maintains system sustainability and scalability. The proposed consensus algorithmis evaluated theoretically in simulation scenarios, and in real-time hardware testbeds to validate itsperformance. Finally, PoAh and its integration with the blockchain in the IoT and edge computingscenarios is discussed. The proposed PoAh, while running in limited computer resources (e.g. single-board computing devices like the Raspberry Pi) has a latency in the order of 3 secs.
Keywords—
Blockchain, Consensus Algorithm, Proof-of-Work, Proof-of-Authentication, Resourceconstrained distributed systems, Internet of Things (IoT).
The Internet of Things (IoT) has various definitions across the research community. A network connectingvarious devices called “
Things ” that are responsible for various tasks not limited to collecting data from theenvironment, is called “Internet of Things”. The devices referred as “
Things ” can be any object, sensor,human or other device. According to various definitions of the IoT, for a device in the network to be1 a r X i v : . [ c s . CR ] J a n onsidered a “ Thing ”, it should have communication capabilities. With their advantages, different IoTarchitectures have become a backbone for a wide variety of applications ranging from smart healthcare toindustrial IoT and smart cities [1, 2]. Specifically in a smart city, millions of sensors and things are criticalfor the IoT which build its various components [3, 4]. All these frameworks will have security requirementswhich this work envisions to be addressed through the use of blockchain technology.Based on the IoT architecture, individual devices are identified with their own ID in a network and it isnot necessary to follow the TCP/IP communication protocol stack. There are several other communicationprotocols for lightweight and low bandwidth systems to build to-end communication in an IoT network[5]. One of the major concerns in an IoT architecture is data collection, while processing is performed at alower level which uses low performance devices such as single board computers. Collecting environmentalvariables or data can be done using middleware for better performance in collection and transmission [6]. Ina server-based scenario, an edge layer is introduced for near real-time data processing [7], which includesEdge Datacenters (EDCs). These EDCs are not centralized and hence do not have any dependencies, whichmakes the IoT architectures suitable for emergency data processing [8, 9]. The integration of IoT anddistributed edge networks is a good fit for the application oriented deployment.With the help of Edge Datacenters, IoT architectures can be designed with such resource constrainedand low performance devices which are capable of low power consumption. Such architectures aidvarious mission-critical applications including military, industrial IoT, event detection. In such applications,attaining real-time data processing and secure data transfer are of paramount importance. Of all the attributesof such environments, security should be given highest priority. Research is already actively being conductedand many algorithms were developed for this purpose which involves cryptography [10, 11]. In the twocategories of cryptography, symmetric cryptography gives a 1000 times increase in performance comparedto asymmetric cryptography and this is one of the reasons it is widely adopted for IoT architectures [12, 13].For both of these cryptography approaches, a central entity plays a vital role during the security initializationphase, such as acting as a key distribution center for symmetric cryptography and acting as the certificateauthority for asymmetric cryptography. A failure at this single point can lead to a catastrophic failure in theentire system and compromise the entire network.
Blockchain Applications
Figure 1: Applications of blockchain.Page – 2-of-26he requirement of a central entity to organize and perform cryptographic operations or other taskscan be achieved with the implementation of a blockchain. In fact, the blockchain’s primary contributionis the resolution of the age-old problem of achieving a consensus via Proof-of-Work authentication. Theblockchain uses a decentralized public ledger for organizing the data and executing transactions. The ledgeris decentralized in nature, such that individual peers in the network possess a copy of the entire or part of theledger, depending on its architecture, after block validation by the miners. A blockchain is cryptographicallyanchored and tamper-proof and is a record of the different transactions that occurred among the participantsin the network [14]. There are various applications of the blockchain as shown in Fig. 1.Blockchain technology replaces a centralized entity with consensus algorithms among the participantsto secure the system in a decentralized fashion. This consensus allows the participants to trust each otherand complete the transfer or transactions among the participants [15]. All the blocks are connected to theprevious blocks with the help of cryptographic hash algorithms and hence the transactional values cannotbe tampered with once the block has been added to the blockchain. Proof-of-Work (PoW), Proof-of-Stake(PoS) and others are some of the consensus algorithms that are implemented in the blockchain. However,one of the major concerns of PoW and PoS is the cost of calculation. They require high computationalpower and energy which makes them unsuitable for IoT applications that choose the blockchain as a meansof decentralization. Hence this paper proposes a new algorithm, Proof-of-Authentication (PoAh), which canbe incorporated into resource constrained distributed systems. To the best of the authors’ knowledge, this isthe first attempt towards a lightweight blockchain consensus approach for scalable IoT deployment.The rest of the paper is organized in the following manner: Section 2 presents the contributions of thiswork. A brief overview of blockchain technology and relevant research is presented in Section 3. Theproposed novel Proof-of-Authentication (PoAh) algorithm is discussed in Section 4. A brief discussionon the analysis of the proposed consensus algorithm is presented in Section 5. Experimental results arepresented in Section 6. Conclusions and future directions are summarized in Section 7.
Fig. 2 shows the process of generation, validation and addition of a block to the blockchain and the issuespresent in the blockchain. Once a transaction is initiated, it is broadcast to a network of devices. Each devicein the network will have a copy or part of the ledger in its respective local storage. Whether to maintaina copy of the entire ledger, or only the recent blocks is determined by the node administrators based onstorage and power considerations. However, miner nodes always contain the entire blockchain. The minerswill then validate each transaction and create a block comprising of multiple transactions, as shown in Fig.2. The transaction validation and block validation consume energy as they require the processing powerfrom various devices. Then, the validated block gets added to the blockchain. The local storage of thedevices also becomes a bottleneck as over time, a big block size consumes a lot of memory which is limitedin IoT devices.It is estimated that by 2022 there will be 18 billion devices spread across the global network, with themajority of them being smart devices and small sensors [16]. The widespread adoption of the IoT bringsforward new technological challenges in data privacy and security [17].The most promising solution to this issue is the implementation of blockchain technology into IoTnetworks. Blockchain is the technology that has maintained the integrity and security of cryptocurrenciessince the introduction of Bitcoin in 2009 [18]. This technology is predicted to be able to: • Maintain transaction data integrity, authentication and immutability as required in IoT networks [18].Page – 3-of-26 “Transaction” isbroadcastedto a P2P network Multiple Transactionsform Blocks
Transaction Validation
Verified Transactions A “Verified Transaction”
BlockValidation
A “Block”A “Validated Block” is added to the existing Blockchainwhich is unalterable New Block Old BlockLedger Transaction CompleteBlockchain (i.e. Ledger)B i+1 B i B i-1 B i-2 Figure 2: Process of block creation and addition into the ledger. Flames are used to graphically highlightcomputationally expensive processes. • Maintain the integrity of stored data within IoT networks, making the data difficult to modify [19]. • Increase reliability of IoT networks through decentralization of nodes [18].However, general IoT systems have different requirements than cryptocurrencies. Some everyday tasksand applications cannot be solved with a blockchain implementation [20]. Various issues must be resolvedfor successful integration of the blockchain into such systems. These can include: • Reducing the time to add blocks to the blockchain to improve system responsiveness [21]. • Improving physical Internet infrastructure to support increased Internet usage of decentralizednetworks [22]. • Accounting for power, CPU and memory constraints of less powerful computers.Blockchain implementation into the IoT is a field that is currently receiving a lot of attention from theacademic and industrial communities due to its promise, with many researchers designing new architecturesand solutions to implement the blockchain into IoT [23, 16, 24].The blockchain promises to be a possible solution to the privacy and security issues posed by IoTsystems. The primary goal of this research project is the creation of an original IoT architecture, withblockchain implementation to secure user access to data and the system.This paper focuses upon the utilization of a private blockchain network in an IoT system. This is dueto several advantages that private blockchains have over public blockchains, when implemented into IoTsystems. Private blockchain networks generally have lower network delay. This is due to the measures publicblockchains have to utilize to motivate peers. In addition there are also fewer issues of trust, resulting inlower security measures required between nodes [21]. This leads to a more responsive network, which mayPage – 4-of-26e desirable for some implementations of the IoT. Another advantage of a private blockchain implementationis that system data are contained entirely within the network, which allows network owners full control oftheir personal data [21].There are always security threats in the IoT from both inside and outside attackers [9, 10, 11].Considering the available cryptographic security solutions, both symmetric and asymmetric cryptographyrequire a centralized dependency for initialization. Therefore, one point failure can lead to compromise ofthe entire system. To overcome such possible hazards, the blockchain provides a decentralized frameworkto secure the system. The problem with existing consensus algorithms is that they are not scalable for IoTsystems with a very large number of nodes. We address this problem in this work and we are proposinga novel scalable blockchain consensus for the IoT. The proposed consensus approach is described in thefollowing section.
Integrating the blockchain with the IoT is still a challenging task due to the resource constrained natureof IoT devices. At the same time, the IoT system demands a decentralized security solution due to thedistributed and untrusted environment. To address these issues, this paper implements a novel blockchainconsensus algorithm called PoAh for IoT systems. The novel aspects of the proposed consensus algorithmtowards blockchain technology are listed as follows: • Proof-of-Authentication (PoAh) is introduced as a cryptographic authentication mechanism forlightweight blockchain. • PoAh is a new consensus algorithm and is validated for resource-constrained devices in IoT and edgecomputing. • Finally, PoAh is validated theoretically and evaluated in both simulation and real-time hardwaretestbeds for performance.
Bitcoin was the first implementation of the blockchain technology. It is a cryptocurrency which usesthe concept of a decentralized distributed ledger at its core to remove the central entity for performingtransactions between different participants in the network [14]. Since then many other cryptocurrencieshave appeared which use blockchain technology, such as Etherium which uses Smart Contracts [7]. Themain intention of all these blockchain consensus algorithms and cryptocurrencies is the removal of thecommon entity which oversees the transactions. Every participant in the network will have a complete orpartial copy of the ledger with the necessary transactions. They implement cryptographic hashes whichallow them to be irreplaceable or irreversible once added to the blockchain. All the participants in thenetwork use consensus algorithms for validating the transactions, and once a transaction is validated it isbroadcast to every participant in the network and everyone updates the blockchain ledger [15]. Along theselines, notwithstanding its remarkable strengths which incorporate decentralization, validation, transparencyand immutability, the blockchain stretches security and privacy at all points in time.In distributed networks participants can be any kind of client, establishment or association sharing a copyof the ledger containing their real transactions in a progressive succession. A sequence of blocks within ina ledger is shown in Fig. 3. Blocks are connected together with hash values in a sequential order for dataimmutability. Individual blocks comprise of sets of transactions, and individual transactions are digitallyPage – 5-of-26igned by the source and verified by other devices in the network before being added into the chain. Somefeatures of the blockchain are now discussed.
Prev_Hash PoAhTx1 Tx1 Tx1 … Prev_Hash PoAhTx1 Tx1 Tx1 … i th Block (i+1) th Block
Figure 3: Block processing with transactions and hashing.
Digital Signatures:
Individual nodes digitally sign the transactions and broadcast them to the networkfor verification by the miners of the network. Each block contains multiple transactions, and participantssign them if they wish to broadcast them to the network and include them to the blockchain.
Consensus:
The novelty of the blockchain is decentralized storage or management, where there are nocentral dependencies. The consensus algorithm plays a vital role in this operation to keep the transactionsin a block and validate them for further processing.
Proof-of-work:
This consensus algorithm searches for values generating a given cryptographic hash.This process essentially solves a cryptographic “puzzle” to validate blocks. As a result, PoW providesgreater security to the overall system. PoW requires intense resources i.e., electricity consumption isequivalent to 1.5 times the overall household electricity for one day in the USA [15]. It is estimatedthat bitcoin transactions will consume close to the electricity in Denmark by 2020 [25] because of thecomputational power needed.
Cryptographic Hashes:
After successful proof-of-work, the block is broadcast to the network and allthe participants compute a hash (such as SHA-256 for normal blockchain) to be used as the previous hash(“Hash_Prev” in Fig. 3) for the next block.
The blockchain can use several consensus algorithms, such as proof-of-stake, proof-of-work, proof-of-relevance and proof-of-activity. A lightweight consensus algorithm named Proof of Authentication hasbeen briefly introduced in [26, 27]. Fig. 4 shows the various blockchain consensus algorithms and theirclassification. The following section provides technical details about the proposed PoAh algorithm.
The major concern in the widespread adoption of IoT networks is the security of deployed systems [17]. Theuse of IoT networks is expected to grow in both the public and private domains and the abundance and valueof data from both public and private IoT networks will attract attackers who will attempt the theft of thisinformation. Within an insecure system, attackers can also potentially gain control of networked IoT deviceswhich can have disastrous effects [23]. As an example, within a private home IoT network, an attacker canpotentially gain control of lights, TV or connected cameras. Insufficient security can also lead to injury oreven death when IoT systems become trusted with the control of larger machines, such as self-driving cars.
Blockchain is the technology that enforces the integrity and value of Bitcoin and other cryptocurrencies.Since its introduction in 2008 by Satoshi Nakamoto (a pseudonym), the blockchain has proven to bePage – 6-of-26 lockchain Consensus Algorithm Validation Proof of WorkProof of ActivityProof of StackProof of RelevanceAuthentication Proof of Authentication
Figure 4: Taxonomy of blockchain consensus algorithms.reliable in protecting the value and integrity of cryptocurrencies [18]. The secure properties of blockchaintechnology complement well with the security and privacy vulnerabilities of the current generation IoTsystems [22].Blockchain technology features the decentralization of data, which means control of the system isdistributed amongst the peers within the network [18]. This ensures that the user maintains control oftheir data as there is no central authority being sent gathered information from sensors. Decentralizationof the data can also benefit the security of IoT networks as access control information is shared amongstall devices within the network [23]. If an attacker attempts to infiltrate a blockchain-secured IoT network,depending on the architecture of the system, more than half of the connected devices will need to be hackedsimultaneously.IoT systems with blockchain implementation is a new area of study, currently receiving significantresearch attention. The use of the blockchain in IoT systems has yet to be standardized. A detailed discussionon the advantages offered by the Blockchain technology when integrated into an IoT architecture and SmartCities is presented in [28]. The current challenge for engineers and researchers attempting to implement thissystem is the creation of functional architectures that can effectively merge these two technologies together,keeping the function of IoT systems while maintaining the privacy and security features of blockchaintechnology. A cross-chain framework which integrates multiple architectures of blockchain for robust andsecure data management for IoT environments is presented in [29]. Similar implementations were alsoproposed in [30], and [31]. A new blockchain based IoT architecture is proposed in [32] which can bedeployed in a smart home, but the overhead added to the communication is high, which makes it unsuitablefor various environments. A blockchain consensus algorithm which can authenticate the identity of the IoTdevice and maintain data protection was proposed in [33].
The fusion of blockchain and IoT technologies is still a new research topic and standards for implementingthese two technologies have not yet been determined. There have been many architectures and proposedsolutions for the implementation of the blockchain into the IoT and the design of these systems can varysignificantly. Page – 7-of-26 .3.1 Decentralized Access Control System
Novo, in [16] details an IoT system which avoids the integration of IoT devices into the blockchain networkand only the access control system of the network is managed and distributed by blockchain technology. Thiscontrol system allows for lower power devices that cannot participate in the blockchain to be used withinthe system. To describe this design very simply, sensor networks collect data and send this informationencrypted to the management hub. The management hub then transforms the information sent fromthe sensors into a format understandable by the rest of the blockchain network. The smart contract issecured with blockchain technology and is managed by public or private miners. It contains access controlinformation which determines the devices that can access the sensor network controlled by the managementhub.
The fair access framework [23] is an IoT system which allows a number of organizations or people to accesseach other’s data. The access control of the data is managed by the same blockchain network, in whichall members are connected to. Each member has a “Wallet” containing their access information and alsocontains all keys pertaining to the information they are allowed to view. To explain this system simply,if member A wants to request a resource that member B possesses, they would send their request and allrelated keys to the blockchain network. The blockchain network would then determine if member A haspermission to access the resource form member B through mining. If member A is authorized to receive theinformation, then member B would transmit the data.Blockchain implementation in IoT systems is a newly researched technology. While at the momentthis technology seems promising, proper implementation can prove to be not viable or not even possible.This increases the risk of implementing prototype projects as the lack of support and preexisting modulesavailable will significantly increase the difficulty of the task. Another problem with blockchain technologyis that it is currently designed around mining cryptocurrencies and can take up to 12 minutes to validate atransaction, depending on the protocol used [18]. While most delays are shorter, this waiting period canmake the blockchain nonviable for some IoT systems that require fast response.
The proposed Proof-of-Authentication is a new consensus algorithm proposed in this paper to builda lightweight and sustainable blockchain for resource-constrained devices. This consensus algorithmintroduces an authentication mechanism during block validation. In other respects, it follows traditionalcommunications.Fig. 5 provides a comparative overview of the proposed PoAh with the more common Proof-of-Workand Proof-of-Stake consensus algorithms. At the very beginning of the process, network precipitantsgenerate transactions (Trx) with the sensed or collected data assembled to form a block. In the figure blocksare represented as B = { T rx , T rx , , T rxn } . The nodes broadcast the blocks for further evaluationand/or validation by trusted nodes in the network. We follow the standard IoT deployment steps to createthe initial trusted node set based on geographical location. Trusted nodes are reachable from any part ofthe network. The proposed model adopts the ElGamal crypto-system for encryption and decryption, i.e. y = g x ( mod p ) where x is the private key PrK and y is the public key PuK. The large prime numbers formodulus operation p and generator function g are publicly known to all the network devices. Prior to blockbroadcast, the network user makes the public key PuK, i.e. y , available to the network and signs the blockusing its own private key PrK, i.e. x .The proposed protocol is specifically designed for resource constrained distributed networks, such asPage – 8-of-26 reate Block Solve Puzzle Broadcast the Proof of Work (PoW) B i-2 B i-1 B i Process Starts Again B i (a) Proof-of-Work (PoW) consensus algorithm Block of Transactions B i-1 B i Process Starts AgainPool of ValidatorsStake Randomly Chosen Validator Validated? Discarded (b) Proof-of-Stake (PoS) consensus algorithm
Proof-of-Authentication (PoAh)
Prof./Dr. Saraju P. Mohanty
Nodes form Block of Transactions Sign with a Private Key Transmit to Trusted Nodes Trusted Nodes Network
Prev-Hash PoAhTx-1 Tx-n … Prev-Hash PoAhTx-1 Tx-n … Authenticated?Yes No (c) The proposed Proof-of-Authentication (PoAh) consensus algorithm.
Figure 5: Various consensus algorithms which can be used in a blockchain technology.Page – 9-of-26he IoT, where there will be limited number of trusted nodes initialized during the network deployment witha minimum trust value and other network devices with a zero trust value i.e. ‘tr = 0’. With successfulblock authentication, an authenticated nodeâ ˘A ´Zs trust value is increased by 1. Similarly, each fake blockauthentication will result in decreasing the trust value by ‘tr = 1’. After node authentication by the trustednode, other untrusted nodes in the network can also identify the authenticated block to gain trust value i.e.‘tr = 0.5’. Importantly, identifying false block authentication can gain high trust value i.e. ‘tr = 1’. Withthis trusted evaluation process, a trusted node can be out of the process when its trust value is lower than athreshold ‘tr < th’, and a normal node can be part of the authentication process. Here, a threshold ‘th = 5’ isconsidered and an initial trust value ‘tr = 10’ is assigned to trusted nodes. The procedure for authenticationis as shown in Fig. 6. Consider a specific node for authenticationIs this a initially defined trusted node?Trust value is 5: i.e. tr >= thAssign a block to be part of chain after block authenticationYes YesYesNo NoNoSelect a node to be part of authentication process Is this a normal node received trust value?
Figure 6: Steps to select authenticated node for PoAh.Upon receiving the block for validation, the trusted node finds the source public key, i.e. y for signatureverification. This work uses asymmetric cryptography, where the public key is used for signature validation.An attacker cannot extract the private key, i.e. x , when all other things are publicly available. This is thediscrete log problem. After successful signature verification, the trusted node evaluates the Media AccessControl (MAC) address and compares it with the received one for a second round of evaluation. Aftersuccessful authentication, validated blocks are broadcast by the trusted nodes with PoAh identification.Subsequently, individual users in the network verify the PoAh information to add blocks into the chain.After acceptance of a valid block, the user computes a hash value to link the next block and retrieves theprevious block hash value to save into the current block. All the nodes in a network follow this property tomaintain the chain as shown in Fig. 5c. The technical steps of the PoAh consensus algorithm are presentedin Algorithm 1. Page – 10-of-26 lgorithm 1: Procedure of the Proposed PoAh.
Inputs :
All nodes in the network follow
SHA − hash. Individual nodes have private ( P rK ) and public keys ( P uK ) . Outputs:
Validated Blocks which are added to the blockchain. ( T rx + ) → blocks; /* Nodes combine various transactions to form theblocks */ ( S P rK ) (block) → broadcast; /* Nodes sign blocks with private key andbroadcast to network */ ( V P uK ) (block) → MAC Checking; /* Trusted node verifies signature withsource public key */ if Authenticated then block || P oAH ( D ) → broadcast; /* Trusted node broadcasts theauthenticated block to network */ H ( block ) → Add blocks into chain; /* If nodes hear from trusted node, theyadd block to blockchain */ else DROP the block; /* If not authenticated, drop the block */ GOTO ( Step − for next block; The Proof-of-Authentication (PoAh) algorithm implements an authentication mechanism in contrast to thevalidation process used by other consensus algorithms. Authentication uses fewer resources and less energythan other mechanisms, which can be highly advantageous in case of a resource-constrained environmentlike IoT architectures.For the implementation of the PoAh algorithm, the block contains the payload data, which can beenvironmental data collected by the sensors, the identification of the node which is invoking the transactionand the time stamp at which the transaction is initiated. In the current paper, all the devices in the networkare connected through a wired or wireless network using a router. Hence the MAC addresses of the devicesare used as identification while initiating the transaction. Once a transaction is initiated, the trusted nodereceives the transactions and starts the validation process. After validating the transaction, the trusted nodeor the miner adds the block to the local blockchain and retransmits the block to the other nodes in thenetwork. All the other participants add the block only if the block is sent by the trusted node. Once thenodes get blocks from trusted nodes, they add them to the local blockchain ledger in the database. Thiswork makes the following claims for PoAh to validate its scalability, while Section 6 presents the simulatedand testbed evaluations of the PoAh algorithm.
Claim - 1:
PoAh utilizes minimal resources for block validation.
Proof:
PoW works as a traditional consensus algorithm for the blockchain, where individual nodescan generate data blocks and miners validate them before being added to the chain. The miner process isnothing but the computation of the inverse of a hash, and it takes the equivalent energy consumed by twohouseholds in one day to validate a block [15]. Utilizing this much energy is not feasible, when it comesto an IoT system. In the IoT, devices are resource constrained with minimal computing power and energysupply. PoAh is proposed to address this issue of evaluating blocks with minimal energy. PoAh introducesan authentication mechanism using a digital signature process. In a cryptographic context, digital signatureand hash (not inverse hash) computation is very fast and utilizes minimal energy. This fact is reflected in thePage – 11-of-26xperimental results in the following section.
Claim - 2:
PoAh requires minimal time compared to PoW without compromising security threats.
Proof:
From the above claim, it is found that PoW consumes significant energy while validating ablock. In addition to this, PoW takes approximately 10 minutes to evaluate a block [15]. This is notacceptable in any kind of IoT application. IoT applications are mainly deployed for real-time monitoringpurposes and 10 minutes to evaluate a block is not acceptable under any circumstances. PoAh proposes toaddress this issue to evaluate blocks in minimal time. As PoAh utilizes authentication to validate blocks,cryptographic authentication takes significantly less time. From experimental evaluations, it is found thatPoAh taking time in seconds, i.e. it is 1000 times faster. Integration of cryptographic authentication anddigital signatures ensure the security level of PoAh [5]. Hence, it can be concluded that PoAh is ideallysuited for IoT applications.
Claim - 3:
PoAh provides substantial security while integrating a blockchain based decentralizedsecurity solution to the IoT.
Proof:
By introducing PoAh as a consensus algorithm for the blockchain in IoT, the proposeddecentralized security solution provides sustainable security infrastructure. The IoT does not requires thesame level of security as required for cryptocurrencies [14]. IoT applications require real-time security withproper authentication of data and source, wherein a cryptographic solution is sufficient protection. PoAhintegrates with the existing cryptographic concept of PoW but ignores the block evaluation of computingthe inverse of a hash. With the SHA-256 hash (from Algorithm â ˘A ¸S 1) and digital signatures, the proposedPoAh provides a decentralized security solution by providing the same level of security as asymmetriccryptography.Furthermore, we are addressing two major weaknesses of current blockchain consensus, namely unstablenetwork connectivity (which may prevent all peers from communicating), and the 51% attack in thenetwork. In PoAh, all the network devices are eligible to generate blocks, whereas only trusted nodesare authenticating them. In any unfavorable situation arising from these weaknesses, are not broadcast toall peers. Only reachable (solves the unstable network problem) and already trusted peers (solves the 51%attack issue) can authenticate and add blocks into the chain. As a result, the 51% attack weakness of PoWis addressed due to the dynamic nature of trust values.
Claim - 4:
PoAh provides a better platform for IoT compared to other blockchain consensus algorithms.
Proof:
The consensus algorithm is the backbone of the blockchain and makes the network and theprocess decentralized. However, current consensus algorithms widely used, such as Proof of Work (PoW),Practical Byzantine Fault Tolerance (PBFT), Proof of Authority (PoAu), and Proof of Elapsed Time (PoET)are very expensive in terms of security and resource requirements [34]. PoW requires approximately 10minutes to valid a block and approximately 1 hour for the block to become accepted in the chain. PBFTrequires at least 2/3 of the network devices to behave honestly by signing transactions and message overheadmay increase significantly as the size of the network increases, affecting both speed and scalability. PoAuconsensus represents a more centralized approach most appropriate for governing or regulatory bodies, andis currently also proving popular with utility companies in the energy sector. PoET is developed by Intel,meaning that trust is still required towards a single authority [35].PoAh is developed to address the above consensus drawbacks for scalable IoT deployment.
The proposed PoAh consensus algorithm has been validated using a simulation environment (for large scalestudy), as well as a hardware based experimental test-bed for real-life scenarios. The details of both options,including the experimental setup to validate the proposed model’s sustainability, and results analysis arepresented in this Section. Page – 12-of-26 .1 Simulation Evaluation
The proposed PoAh algorithm is simulated to evaluate its sustainability in the Python programminglanguage. In this experimental setup, an environment of five nodes in the network has been used, where twonodes are trusted nodes to work as miners and 35 bytes are taken as the size of block. Multiple transactionsare incorporated into a block before being broadcast. This work uses the ElGamal cryptographic system forsignature, verification, MAC and encryption. In the simulation, the block format is created as < Source ID,“Signature”, MAC, Trx1, Trx2, > for PoAh testing. The output screenshot of the PoAh evaluation is shownin Fig. 7. Blockchain Validity: Valid ********************************************** Blockchain Validity : Valid ********************************************** N4: Adding the following transaction: Sender: N3, Receiver: N3, Amount: 5312. N4: Adding the following transaction: Sender: Nl, Receiver: N2, Amount: 4736. N4: Adding the following transaction: Sender: N2, Receiver: N2, Amount: 9906. N4: Adding the following transaction: Sender: Nl, Receiver: N3, Amount: 459. N4: Adding the following transaction: Sender: N2, Receiver: N2, Amount: 7672. N4: Adding the following transaction: Sender: Nl, Receiver: N2, Amount: 719. N4: Mining for the next block … N4: Done! block has been mined in 3.48132 seconds! --------new block info-------- Block Index 2 Transactions 0 : Sender: N3, Receiver: N3, Amount: 5312. Transactions 1 : Sender: N1, Receiver: N2, Amount: 4736. Transactions 2 : Sender: N2, Receiver: N2, Amount: 9906. Transactions 3 : Sender: N1, Receiver: N3, Amount: 459. Transactions 4 : Sender: N2, Receiver: N2, Amount: 7672. Transactions 5 : Sender: N1, Receiver: N2, Amount: 719. Hash : d07a039e478dla33623cbl2b121959bee3111c2 Proof of Authentication: 153988
Figure 7: Sample output of PoAh.The network users generate blocks and sign them using own their private key and broadcast them forvalidation. Trusted nodes use the source node public key to validate blocks. If blocks are authenticatedsuccessfully, the trusted node broadcasts the blocks again to add into the chain. In this simulationenvironment, results from 500 iterations show that the average time for PoAh is 3.34 seconds. The 500iteration results from the simulation are listed in Fig. 8, where they are computed from the Algorithm 1steps.
The Proof-of-Authentication (PoAh) consensus algorithm was implemented in a real-time hardware testbedfor evaluation of performance. The testbed included 6 Raspberry Pi single board computers, where fiveboards were collecting the environmental data using various sensors. Among the five Raspberry Pis, threePage – 13-of-26 . . . . . . .
12 3 .
02 3 . . . .
51 3 .
11 3 2 . . .
02 3 .
83 3 .
21 2 . . . . . .
01 3 . . .
98 2 .
45 2 . . . . . . . . . . . . . . .
10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260 270 280 290 300 310 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460 470 480 490 500 Po A h Executions T i m e i n S e c o nd s Figure 8: Time taken by PoAh for authentication of blocks.of them were Raspberry Pi 1 Model B+ and the other two were Raspberry Pi 3 Model B. This was consideredto evaluate the performance of the consensus algorithm when different processing powers were utilized. TheRaspberry Pi 1 Model B+ is equipped with a 700 MHz Cortex single core processor with 512 MB of RAMwhereas the Raspberry Pi 3 Model B is equipped with a Quad Core 1.2 GHz ARM Cortex - A53 CPU and1 GB LPDDR2 RAM. These Raspberry Pi boards were used for collecting the data and for validating it. Inaddition, a trusted node was deployed in the network. As a trusted node, a Raspberry Pi 3 Model B+ wasdeployed which is equipped with a Quad Core 1.4 GHz ARM Cortex A53 CPU and 1GB of LPDDR2 RAM.The trusted module also has the capability of connecting to a 5 GHz wireless network which gave it theadvantage of better and faster connectivity to the network. This testbed is of sufficient complexity to allowreal-time evaluation of the PoAh algorithm and is not intended as a commercial deployment demonstration.For the experimental setup, the single-board computers that are connected using Ethernet cables arethe Raspberry Pi 1 Model Bs as they do not have wireless communication capabilities. All the devicesare connected through a router and are connected using wired and wireless connectivity wherever possible.Once all the devices are connected to each other, the PoAh algorithm will start authenticating the transactionsperformed by the nodes. The load of the transactions is the time stamp, the environmental data collectedand the identity of the node that is collecting the data. Every participant in the network stores the blockchainledger locally using an SQLite database. For real-time evaluation, 300 transactions were performed by all thenodes and various parameters were observed for each transaction as discussed in the following subsections.The Node-RED development tool is used for prototyping the PoAh blockchain consensus algorithmon the Raspberry Pi. Node-RED is a flow-based programming tool used for the development of IoTapplications. The interface can be accessed through a web browser which makes it optimal for a distributedwork environment like the implementation of the PoAh consensus algorithm. The server and the clientnode software was developed and deployed using the Node-RED interface on the Raspberry Pi single boardcomputers. The Node-RED development environment is not confined by the operating system, which makesthe development of PoAh deployable on various environments and not restricted to the Raspberry Pi. Fig.9 shows the implementation of the PoAh algorithm on a Raspberry Pi Mode 3B+. As shown in the figure,once the data is received from the network, the hash of the previous block is fetched from the blockchainand is sent cascading it to the data received. After authentication of the device that sent the data, the serverthen calculates the hash of the entire block which contains the data received and the previous hash. This isPage – 14-of-26hen added to the blockchain at the server and transmitted to the other nodes to be added to their respectiveblockchains. The details of the Node-RED program are shown in Fig. 10.
Participant 1 Participant 2 Participant 3Participant 4 Participant 5
Miner Miner
Figure 9: Block processing with transactions and hashing.
Node-RED
Full
Deploys everything in theworkspace
Modified Flows
Only deploys flows thatcontain changed nodes
Modified Nodes
Only deploys nodes that havechanged
Deploy ProjectsNewOpenProject Settings View Show sidebarDebug messagesLibraryibm watson-iotPi Cpu TempeTime-Injectplay audioTTS Say Hellopi sense-hatClockCompasspi sense-hat-simulatoClockCompassExamplesImportClipboardExportClipboardLibraryFlows AddRenameDeleteSubflowsCreate SubflowSelection to SubflowSearch flowsConfiguration nodesManage paletteSettingsKeyboard shortcutsNode-RED websitev0.19.5 Flow 1
Flow 2
CONVERTDatabase
RWC rehashgetPrevHash prevHash DEBUGUDP IN get
RWC
CONVERT Split join
UDP OUTtimestamp_before_hash timestamp_after_db join time_hashingtimestamp join time_communication Database
RWC time_consumed_db_entry join filter nodes input injectcatchstatuslinkmqtthttpwebsockettcpudpWatson IoTserial output debuglinkmqtthttp responsewebsockettcpudpWatson IoTplay audioserial Figure 10: Node-RED features in PoAh experiment.It needs to be noted that we tried to run PoW and PoS in the real-life resource-constrained IoT devicesavailable in our testbed. We observed that even one mining process could not be completed for these heavy-duty algorithms. We observed that the testbed hardware platform doesn’t complete even after running itfor days. This indicates that it is impossible to run these heavy-duty consensus algorithms in resource-constrained IoT devices. If we run these IoT devices under the assumption that they are deployed in possiblyremote locations in real-life applications, like smart cities [4] and powered by batteries, then obviously thebattery won’t last enough to even complete one mining process for these heavy-duty algorithms.
Fig. 11 shows the time taken by a transaction to reach the miner or trusted node. The transactions havetaken around 100ms to 200ms and this depends on the connection type that was used for the transaction.Fig. 12 presents the time taken by the miner or the trusted node to authenticate the data sent and add to theblockchain at the validation node locally. Once the validation is complete, the trusted node will broadcastthe transaction back to the network nodes at this point. Once the other nodes obtain the transaction from thetrusted node or the miner, they add the transaction to their local blockchain. Once this is done, a transactionis considered complete. The time taken for the transaction to be completed is presented in Fig. 13.Fig. 14a and Fig. 14b show the time required for 300 transactions to be added by different RaspberryPis. Fig. 14a shows the data pertaining to the Raspberry Pi 2 Model B+. Once the Raspberry Pi 2 ModelB+ receives the data from a miner or validator node, the time consumed by that single board computer toPage – 15-of-26 Tr a nsaction T i m e i n M illi s ec ond s Figure 11: Time taken for the transaction to reach validator from the ClearPi.check the identity of the sender and add to the blockchain is shown in Fig. 14a. Similarly, in the case of aRaspberry Pi 1 Model B, the time taken by the single board computer to check the received data and add itto the blockchain is presented in Fig. 14b.As shown in Fig. 10, various timestamps were generated during the process for calculating the timeconsumed for various processes. At the server node, a timestamp t sr is generated when the message isreceived at the node. When the message traverses through various functions in NodeRed, timestamp t sh is generated at the server and timestamps t cr and t cv are generated at the client. The time required forauthentication by the server is: δt sa = t sh − t sr , (1)where t sh is the timestamp after validation and hashing are complete at the server.The time required by the client to check if the message is sent by the server and add it to the blockchainis: δt ca = t ch − t cr , (2)where t cr is the timestamp generated when the authenticated message is received at the client and t ch isthe timestamp generated when the message is added to the blockchain at the client node. The mean andstandard deviations presented in Table 2 and Table 3 are using δt sa and δt sa calculated over a period of 300transactions.The timestamps t sr , t sh , and t cr are also used for calculating the time required for communicationbetween the devices during the transactions. This helps in finding the latency and the overhead added bydifferent communication mechanisms. When the sensor data is collected by the client node and a transactionis initiated, an initial timestamp is added to the block, t i . The time taken to complete one transaction is thefollowing: δt tx = t ch − t i , (3)where t ch is the corresponding client node timestamp where the transaction is added to the blockchain. Besides the time consumed and the processing power of the entire system, another major challenge inthe case of a blockchain network is the power consumption. Fig. 15 shows the experimental setup forPage – 16-of-26 Tr a nsaction T i m e i n M illi s ec ond s Figure 12: Time taken for the block to be validated and to be added to the blockchain at the BlackPi.measuring the power consumption of the system. The minimum and maximum power consumption ofdifferent Raspberry Pis is listed in Table 1. An electrical meter is used to measure the power consumptionof the Raspberry Pis as shown in the figure. It is connected to the power outlet and the Raspberry Pi boardpower supply is connected through the electrical meter. Two measurements were considered during theexperiment. Once when the system is idle and once when the Raspberry Pi is processing or adding the blockto the blockchain.Fig. 16 shows the power consumption results. The minimum power consumption is when the RaspberryPis are waiting for the data to arrive. In the case of a client node, they are collecting the data and transmittingat a 10 sec interval. The minimum power consumption is when the nodes are neither collecting the data northey are adding the blocks to the blockchain. The maximum power consumption is when the CPU usageis at its maximum which indicates that the nodes are collecting the data and adding a received block to theblockchain. Fig. 16 results are abstracted from the testbed as shown in Fig. 15 and the energy consumption islisted in Table 1. In the case of a miner or validator node, the minimum power consumption is when it is idle,i.e. when no communication reaches the node and it is not adding any new blocks to the blockchain. Themaximum power consumption is when the block is being validated, added to the blockchain and broadcastto the rest of the network. The times required to add blocks are tabulated in Table 2 while the times requiredfor communication are tabulated in Table 3. The frequency plots of the measured results for the blockaddition and communication are shown in Fig. 17 and Fig. 18, respectively. A comparative analysis of theproposed PoAh algorithm to other consensus algorithms is given in Table 4, in which, we have presented themining process, possible attacks and power consumption in comparison to PoAh. It is clear that the PoWmining process took around 538 KW energy and is impossible to deploy in recourse constraints IoT devices.Further we tested PoS, which is cutting the energy consumed by a hundredfold (i.e. around 99%). As aresult, the PoS mining process is taking around 5.5 KW and IoT applications sill demands novel blockchainconsensus for resource constraint devices. Compared to the above most promising blockchain consensus,the proposed PoAh is introducing authentication mechanisms and taking a maximum of 3.5 Watts energy forblock authentication. PoAh is much faster compared to PoW and PoS, and suitable for IoT devices, whichis experimented and validated in the in-lab hardware testbed.Page – 17-of-26 Tr a nsaction T i m e i n M illi s ec ond s Figure 13: Time taken for the transaction to reach ClearPi from block creation and after validation.
50 100 200 250 300150 Tr a nsaction T i m e i n M illi s ec ond s (a) For a Raspberry Pi Model 2B (a Clear Pi)
50 100 150 200 250 300 Tr a nsaction T i m e i n M illi s ec ond s (b) For Raspberry Pi Model 1 (a Clear Pi) Figure 14: Time taken to check if Black Pi sent the transaction and add it to the Blockchain (Spikes whentraffic is high). Table 1: POWER CONSUMPTION OF DIFFERENT RASPBERRY Pis.
Power Types Raspberry Pi 1 Raspberry Pi-2 Raspberry Pi-3
Max Power consumption in Watts 1.8 2.5 3.6Min Power consumption in Watts 1.5 2 3.1Table 2: TIME TAKEN BY PoAh FOR ADDING THE BLOCKS RECEIVED.
Single Board Computer Mean(Milliseconds) Standard Deviation(Milliseconds)
BlackPi 843 26ClearPi (Raspberry Pi 2 Model B+) 85 36ClearPi (Raspberry Pi 1 Model B+) 162.4 98.6Page – 18-of-26 a) Maximum power consumption ofRaspberry Pi 3 Model B+ (b) Maximum power consumptionof Raspberry Pi 2 Model B+ (c) Maximum power consumptionof Raspberry Pi 1 Model B+
Figure 15: Experimental setup for power consumption measurement of different versions of Raspberry Pi. Raspberry Pi 3 Model B+ Raspberry Pi 3 Model B Raspberry Pi 1 Model B+ P o w e r C on s u m p ti on i n W a tt s Maximum Minimum
Figure 16: Power consumption of different Raspberry Pi versions while running the PoAh blockchain.Page – 19-of-26able 3: TIME TAKEN FOR COMMUNICATION.
Single Board Computer Mean(Milliseconds) StandardDeviation(Milliseconds)
BlackPi 89 20ClearPi (Raspberry Pi 2 Model B+) 116.35 12.3ClearPi (Raspberry Pi 1 Model B+) 246 33
This paper provides a novel consensus algorithm named Proof-of-Authentication (PoAh) for sustainableand lightweight blockchain for resource-constrained distributed systems, such as IoT and edge computing.Furthermore, the proposed algorithm bypasses centralized dependencies by building a lightweight decen-tralized security solution. The proposed algorithm is validated in terms of security and sustainability in threesteps: (i) theoretical validation, (ii) simulation results, and (iii) real-time test-bed deployment. The proposedPoAh, while running in limited computer resources and using minimal energy (e.g. single-board computingdevices like the Raspberry Pi) has a latency in the order of few secs. Thus PoAh is scalable for largescaleIoT deployed in the smart cities. In future work, the PoAh algorithm will be evaluated in big data scenariosfor large scale networks. We are also planning a comparative study of various benchmark security solutionsin large scale networks, as well as evaluation of various threat attack scenarios.
A preliminary version of this study has been presented at following [26, 27].The authors would like to sincerely thank Dr. Gautam Das for his feedback on conference version ofthis work.
References [1] M. S. Castanho, F. A. F. Ferreira, E. G. Carayannis, and J. J. M. Ferreira, “SMART-C: Developinga “Smart City” Assessment System Using Cognitive Mapping and the Choquet Integral,”
IEEETransactions on Engineering Management , pp. 1–12, 2019.[2] J. Corbett, K. Wardle, and C. Chen, “Toward a sustainable modern electricity grid: The effects of smartmetering and program investments on demand-side management performance in the US electricitysector 2009-2012,”
IEEE Transactions on Engineering Management , vol. 65, no. 2, pp. 252–263, May2018.[3] Y. Hsiao, M. Wu, and S. C. Li, “Elevated Performance of the Smart City-A Case Study of the IoT byInnovation Mode,”
IEEE Transactions on Engineering Management , pp. 1–15, 2019.[4] S. P. Mohanty, U. Choppali, and E. Kougianos, “Everything You wanted to Know about Smart Cities,”
IEEE Consum. Electron. Mag. , vol. 5, no. 3, pp. 60–70, July 2016.[5] A. Dorri, S. S. Kanhere, R. Jurdak, and P. Gauravaram, “Blockchain for IoT Security and Privacy:The Case Study of a Smart Home,” in
Proceedings of the IEEE International Conference on PervasiveComputing and Communications Workshops (PerCom Workshops) , 2017, pp. 618–623.[6] M. V. Moreno, F. Terroso-Saenz, A. Gonzalez-Vidal, M. Valdes-Vela, A. F. Skarmeta, M. A.Zamora, and V. Chang, “Applicability of Big Data Techniques to Smart Cities Deployments,”
IEEETransactions on Industrial Informatics , vol. 13, no. 2, pp. 800–809, April 2017.Page – 20-of-26
58 874 890 906 922 938 954 970 986 1002
Time taken to add a received block by Raspberry Pi 3 N u m b e r o f T r a n s ac ti on s (a) BlackPi
38 54 70 86 102 118 134 150 166 182
Time taken to add a received block by Raspberry Pi 2 N u m b e r o f T r a n s ac ti on s (b) ClearPi (Raspbery Pi 2 Model B+)
30 90 150 210 270 330 390 450 510 570
Time taken to add a received block by Raspberry Pi 1 N u m b e r o f T r a n s ac ti on s (c) ClearPi (Raspbery Pi 1 Model B+) Figure 17: Frequency plot of time required for block addition.[7] W. Shi, J. Cao, Q. Zhang, Y. Li, and L. Xu, “Edge Computing: Vision and Challenges,”
IEEE Internetof Things Journal , vol. 3, no. 5, pp. 637–646, 2016.[8] A. Zanella, N. Bui, A. Castellani, L. Vangelista, and M. Zorzi, “Internet of Things for Smart Cities,”
IEEE Internet of Things journal , vol. 1, no. 1, pp. 22–32, 2014.[9] D. Puthal, S. Nepal, R. Ranjan, and J. Chen, “Threats to networking cloud and edge datacenters in theinternet of things,”
IEEE Cloud Computing , vol. 3, no. 3, pp. 64–71, 2016.[10] D. Minoli, K. Sohraby, and J. Kouns, “IoT Security (IoTSec) Considerations, Requirements, andArchitectures,” in
Proceedings of the 14th IEEE Annual Consumer Communications & NetworkingConference (CCNC) , 2017, pp. 1006–1007.[11] J.-H. Lee and H. Kim, “Security and Privacy Challenges in the Internet of Things [Security and PrivacyMatters],”
IEEE Consumer Electronics Magazine , vol. 6, no. 3, pp. 134–136, 2017.[12] T. Xu, J. B. Wendt, and M. Potkonjak, “Security of IoT Systems: Design Challenges andOpportunities,” in
Proceedings of the IEEE/ACM International Conference on Computer-AidedDesign , 2014, pp. 417–423.[13] D. Puthal, S. Nepal, R. Ranjan, and J. Chen, “DLSeF: A Dynamic Key-Length-Based Efficient Real-Page – 21-of-26
Time taken for communication to reach Raspberry Pi 3 N u m b e r o f T r a n s ac ti on s (a) BlackPi
110 115 120 125 130
Time taken for communication to reach Raspberry Pi 2 N u m b e r o f T r a n s ac ti on s (b) ClearPi (Raspbery Pi 2 Model B+)
220 230 240 250 260 270
Time taken for communication to reach Raspberry Pi 1 N u m b e r o f T r a n s ac ti on s (c) ClearPi (Raspbery Pi 1 Model B+) Figure 18: Frequency plot of time required for communication.Time Security Verification Model for Big Data Stream,”
ACM Transactions on Embedded ComputingSystems (TECS) , vol. 16, no. 2, p. 51, 2017.[14] D. Puthal, N. Malik, S. P. Mohanty, E. Kougianos, and C. Yang, “The Blockchain as a DecentralizedSecurity Framework,”
IEEE Consumer Electronics Magazine , vol. 7, no. 2, pp. 18–21, 2018.[15] G. Zyskind, O. Nathan, and A. S. Pentland, “Decentralizing Privacy: Using Blockchain to ProtectPersonal Data,” in
Proceedings of the IEEE Security and Privacy Workshops . IEEE, 2015, pp. 180–184.[16] O. Novo, “Blockchain Meets IoT: An Architecture for Scalable Access Management in IoT,”
IEEEInternet of Things Journal , vol. 5, no. 2, pp. 1184–1195, 2018.[17] Z. Huang, X. Su, Y. Zhang, C. Shi, H. Zhang, and L. Xie, “A Decentralized Solution for IoT DataTrusted Exchange Based-on Blockchain,” in
Proceedings of the 3rd IEEE International Conference onComputer and Communications (ICCC) , 2017, pp. 1180–1184.[18] A. Nayak and K. Dutta, “Blockchain: The Perfect Data Protection Tool,” in
Proceedings of theInternational Conference on Intelligent Computing and Control (I2C2) , 2017, pp. 1–3.[19] N. Kshetri, “Can Blockchain Strengthen The Internet of Things?”
IT Professional , vol. 19, no. 4, pp.Page – 22-of-26able 4: COMPARATIVE PERSPECTIVE OF PoAh WITH OTHER RELATED WORKS.
ConsensusAlgorithm Year BlockchainType Mining Prone ToAttacks PowerConsumption
Proof-of-Work [36] 2008 Permission-Less Based onComputationPower Bribe attack,Sybil attack,51% attack 538 KWhElectricityconsumptionProof-of-Stake [37] 2012 Permission-Less Validation DoS, Sybilattack 5.5 KWhElectricityconsumptionRipple [38] 2014 Permissioned Vote BasedMining DoS attack,Sybil attacks -Proof-of-Vote [39] 2017 Consortium Vote BasedMining - -Proof-of-Trust [40] 2018 PermissionBased Probabilityand VoteBased Mining DDoS attack -Proof ofPUF-EnabledAuthentication(PoP) [41, 42] 2019 PermissionBased Authentication No KnownAttacks 5 W (UltraLow PowerConsumption)Proof-of-Authentication(PoAh) (TheCurrent Paper)
Communications of the ACM , vol. 62, no. 5, pp. 64–69, 2019.[21] W. Xin, T. Zhang, C. Hu, C. Tang, C. Liu, and Z. Chen, “On Scaling and Accelerating DecentralizedPrivate Blockchains,” in
Proceedings of the IEEE 3rd International Conference on Big DataSecurity on Cloud (bigdatasecurity), IEEE international conference on High Performance and SmartComputing (HPSC), and IEEE International Conference on Intelligent Data and Security (IDS) , 2017,pp. 267–271.[22] A. Kuzmin, “Blockchain-Based Structures for a Secure and Operate IoT,” in
Proceedings of theInternet of Things Business Models, Users, and Networks , 2017, pp. 1–7.[23] A. Ouaddah, A. Abou Elkalam, and A. Ait Ouahman, “FairAccess: A New Blockchain-Based accessControl Framework for the Internet of Things,”
Security and Communication Networks , vol. 9, no. 18,pp. 5943–5964, 2016.[24] S. Biswas, K. Sharif, F. Li, S. Maharjan, S. P. Mohanty, and Y. Wang, “PoBT: A Light WeightConsensus Algorithm for Scalable IoT Business Blockchain,”
IEEE Internet of Things Journal , pp.1–1, 2020.[25] S. King and S. Nadal, “Ppcoin: Peer-to-Peer Crypto-Currency with Proof-of-Stake,” self-publishedpaper, August , vol. 19, 2012.[26] D. Puthal and S. P. Mohanty, “Proof of Authentication: IoT-Friendly Blockchains,”
IEEE PotentialsMagazine , vol. 38, no. 1, pp. 26–29, January 2019.[27] D. Puthal, S. P. Mohanty, P. Nanda, E. Kougianos, and G. Das, “Proof-of-Authentication forPage – 23-of-26calable Blockchain in Resource-Constrained Distributed Systems,” in
Proceedings of the 37th IEEEInternational Conference on Consumer Electronics (ICCE) , 2019.[28] J. Sun, J. Yan, and K. Z. K. Zhang, “Blockchain-Based Sharing Services: What Blockchain TechnologyCan Contribute to Smart Cities,”
Financial Innovation , vol. 2, no. 1, p. 26, Dec 2016.[29] Y. Jiang, C. Wang, Y. Wang, and L. Gao, “A Cross-Chain Solution to Integrating Multiple Blockchainsfor IoT Data Management,”
Sensors , vol. 19, no. 9, 2019.[30] G. Sagirlar, B. Carminati, E. Ferrari, J. D. Sheehan, and E. Ragnoli, “Hybrid-IoT: Hybrid BlockchainArchitecture for Internet of Things - PoW Sub-Blockchains,” arXiv , 04 2018.[31] O. Alphand, M. Amoretti, T. Claeys, S. Dall’Asta, A. Duda, G. Ferrari, F. Rousseau, B. Tourancheau,L. Veltri, and F. Zanichelli, “IoTChain: A Blockchain Security Architecture for The Internet ofThings,” in
Proceedings of the IEEE Wireless Communications and Networking Conference (WCNC) ,April 2018, pp. 1–6.[32] A. Dorri, S. S. Kanhere, R. Jurdak, and P. Gauravaram, “Blockchain for IoT Security and Privacy:The Case Study of a Smart Home,” in
Proceedings of the IEEE International Conference on PervasiveComputing and Communications Workshops (PerCom Workshops) , March 2017, pp. 618–623.[33] D. Li, W. Peng, W. Deng, and F. Gai, “A Blockchain-Based Authentication and Security Mechanismfor IoT,” in
Proceedings of the 27th International Conference on Computer Communication andNetworks (ICCCN) , 2018, pp. 1–6.[34] M. Andoni, V. Robu, D. Flynn, S. Abram, D. Geach, D. Jenkins, P. McCallum, and A. Peacock,“Blockchain technology in the energy sector: A systematic review of challenges and opportunities,”
Renewable and Sustainable Energy Reviews
Ripple Labs IncWhite Paper , vol. 5, 2014.[39] K. Li, H. Li, H. Hou, K. Li, and Y. Chen, “Proof of Vote: A High-Performance ConsensusProtocol Based on Vote Mechanism & Consortium Blockchain,” in
Proceedings of the IEEE19th International Conference on High Performance Computing and Communications; IEEE 15thInternational Conference on Smart City; IEEE 3rd International Conference on Data Science andSystems (HPCC/SmartCity/DSS) , Dec 2017, pp. 466–473.[40] J. Zou, B. Ye, L. Qu, Y. Wang, M. A. Orgun, and L. Li, “A Proof-of-Trust Consensus Protocol forEnhancing Accountability in Crowdsourcing Services,”
IEEE Transactions on Services Computing ,pp. 1–14, 2018.[41] S. P. Mohanty, V. P. Yanambaka, E. Kougianos, and D. Puthal, “PUFchain: Hardware-AssistedBlockchain for Sustainable Simultaneous Device and Data Security in the Internet of Everything(IoE),” arXiv Computer Science , vol. 1909.06496, Sep 2019.[42] S. P. Mohanty, V. P. Yanambaka, E. Kougianos, and D. Puthal, “PUFchain: Hardware-AssistedBlockchain for Sustainable Simultaneous Device and Data Security in Internet of Everything (IoE),”
IEEE Consumer Electronics Magazine , vol. 9, no. 2, pp. 8–16, March 2020.Page – 24-of-26 uthors’ Biographies
Deepak Puthal (M’16) received the Ph.D. degree in computer science from theUniversity of Technology Sydney (UTS), Australia. He is currently a Lecturerat School of Computing, Newcastle University, Newcastle upon Tyne, UK. He isan author/co-author of more than 100 peer-reviewed publications in internationalconferences and journals, including ACM and IEEE transactions. His researchinterests include cyber security, Internet of Things, distributed computing, andedge/fog computing. He has been a Program Chair and a Program Committeemember in several IEEE and ACM sponsored conferences. He was a recipientof the 2017 IEEE Distinguished Doctoral Dissertation Award from the IEEEComputer Society and STC on Smart Computing. He served as a Co-Guest Editorof several reputed journals, including Concurrency and Computation: Practice andExperience, Wireless Communications and Mobile Computing, and Information Systems Frontier. He is anAssociate Editor of the IEEE Transactions on Big Data, and IEEE Consumer Electronics Magazine.
Saraju P. Mohanty (SM’08) received the bachelor’s degree (Honors) inelectrical engineering from the Orissa University of Agriculture and Technology,Bhubaneswar, in 1995, the master’s degree in Systems Science and Automationfrom the Indian Institute of Science, Bengaluru, in 1999, and the Ph.D. degree inComputer Science and Engineering from the University of South Florida, Tampa,in 2003. He is a Professor with the University of North Texas. His researchis in “Smart Electronic Systems” which has been funded by National ScienceFoundations (NSF), Semiconductor Research Corporation (SRC), U.S. Air Force,IUSSTF, and Mission Innovation Global Alliance. He has authored 300 researcharticles, 4 books, and invented 4 U.S. patents. His has Google Scholar citationswith an h-index of 34 and i10-index of 124 with 5000+ citations. He was arecipient of 11 best paper awards, IEEE Consumer Electronics Society Outstanding Service Award in 2020for leadership contributions, the IEEE-CS-TCVLSI Distinguished Leadership Award in 2018 for services tothe IEEE and to the VLSI research community, and the 2016 PROSE Award for Best Textbook in PhysicalSciences and Mathematics category from the Association of American Publishers for his Mixed-SignalSystem Design book published by McGraw-Hill. He has delivered 9 keynotes and served on 5 panelsat various International Conferences. He has been serving on the editorial board of several peer-reviewedinternational journals, including IEEE Transactions on Consumer Electronics (TCE), and IEEE Transactionson Big Data (TBD). He is currently the Editor-in-Chief (EiC) of the IEEE Consumer Electronics Magazine(MCE). He has been serving on the Board of Governors (BoG) of the IEEE Consumer Electronics Society,and has served as the Chair of Technical Committee on Very Large Scale Integration (TCVLSI), IEEEComputer Society (IEEE-CS) during 2014-2018. He is the founding steering committee chair for the IEEEInternational Symposium on Smart Electronic Systems (iSES), steering committee vice-chair of the IEEE-CS Symposium on VLSI (ISVLSI), and steering committee vice-chair of the OITS International Conferenceon Information Technology (ICIT). He has mentored 2 post-doctoral researchers, and supervised 11 Ph.D.dissertations, 26 M.S. theses, and 10 undergraduate projects.Page – 25-of-26 enkata P. Yanambaka (M’19) received the Bachelor of Technology degree inelectronics and communications from the JNTU, India, in 2014. He obtained hisPh.D. at the System Electronic Systems Laboratory (SESL) at the Department ofComputer Science and Engineering, University of North Texas. He is currentlyan Assistant Professor in the School of Engineering and Technology, CentralMichigan University. His research interests are in Security in Internet of Things(IoT), Energy-Efficient Circuits and Systems, and Application-Specific SystemsDesign. He has authored of a 12 research articles which include multiplejournals/transactions articles. He has a regular reviewer of various peer-reviewedjournals and conferences.