BlockLoc: Secure Localization in the Internet-of-Things using Blockchain
BBlockLoc: Secure Localization in the Internet-of-Things usingBlockchain
Omar Cheikhrouhou
College of CIT, Taif UniversityTaif, Saudi ArabiaComputer and Embedded Systems LaboratoryUniversity of Sfax, Sfax, Tunisia [email protected]
Anis Koubˆaa
Prince Sultan University, Saudi ArabiaCISTER/INESC-TEC, ISEP, PortugalGaitech Robotics, China [email protected]
Abstract —Several IoT applications are tightly dependent onthe locations of the devices. However, localization algorithms canbe easily compromised by injecting false locations. In this paper,we propose a Blockchain-based secure localization algorithmfor the Internet of Things (IoT). The algorithm uses a publicledger (Blockchain) that contains nodes position and the listof their neighbor nodes. This ledger is shared among the IoTdevices. Once an IoT device is localized its new position andthe list of neighbor nodes are added to the Blockchain. Thisshared localization data will be used later by other IoT devicesfor their localization process. To avoid the attack where amalicious node sends a fake position, the correctness of theclaimed position are verified before adding it to the Blockchain.Moreover, data exchanged between nodes (IoT devices) are signedto guarantee their authenticity and integrity. The integration ofthese security mechanisms into the localization process permitsto exclude false data and therefore reduces the localization error.The simulation results show that adding the proposed securitymechanism improves the localization accuracy of the algorithmwhen running in the presence of malicious nodes.
I. INTRODUCTIONAccording to Cisco, fifty billion devices will be deployedin the Internet-of-Things (IoT) by 2020 [1], [2]. Several typesof devices are connected including smartphones, healthcaresensor devices, drones and robots [3], [4], vehicles [5], in-dustrial machines, to name a few. For all these applications,localization is essential considering that these applicationsare typically location-aware. In fact, localization is attractinginterest due to the emerging of context-aware applications inthe IoT [6], [7]. However, to take benefit from localizationservices, applications must trust the localization data and makesure that the positions are not manipulated by malicious nodes.Therefore, several research works are addressing the problemof secure localization in the Internet of Things, such as [8],[9], [10], [11], [12]. In fact, many localization algorithms suchas those based on triangulation [7], or RSSI-based algorithms[13], [14] are very dependent on the correctness of the locationof nodes participating in the localization algorithm. If anynode provides the wrong location information, the wholelocalization system will be compromised. Thus, it is cruciallyimportant to secure the localization algorithm to avoid thisproblem for location-aware IoT services and applications.
Related works:
Secure localization in IoT is still in itsinfancy, and several challenges are still open. Several recent works have tackled the problem [8], [9], [10], [11], [12]. In [8],the authors addressed the problem of localization of dronesin urban environments, which demands high precision andaccuracy in the selection of waypoints, and presented a novelsolution that is capable of securing the context information forsharing 3D waypoints between UAVs. The proposed approachachieves optimal localization through hierarchical context-aware aspect-oriented Petri nets while being powered by anew drone context-exchange protocol for security validations.The wormhole attack has been addressed in [9] and [10].Paper [9] proposed a label-based secure localization schemeto detect and defend against wormhole attack. The proposedwork addressed only the wormhole attack, and so it is stillvulnerable to other kinds of attacks. Moreover, the authorsassumed in their network model that there is no packet loss,which is not realistic in real scenarios.In [10], the authors proposed a secure localization algorithmfor DV-HOP that establishes the neighbor node relationshiplist (NNRL) between nodes to avoid wormhole attack. All thenodes get the ID numbers of their neighbor nodes throughNNRL. Then, the detection of a suspected node is done bycomparing the theoretical and the actual number of neighbornodes. These suspected nodes are eliminated from the local-ization process.The Sybil attack, where a malicious node generates severalidentities to hide its real identity, was addressed in [12].The authors proposed Sybil Free APIT (SF-APIT), a securelocalization scheme for hostile distributed wireless sensornetworks that can detect Sybil nodes. The detection mecha-nism is based on the received signal strength. To prove thecorrectness of their idea the APIT [15] localization algorithmwas used as a reference. In [11], the authors proposed theSecure Location of Things (SLOT) framework to mitigatethe spoofing attack. They reformulate the location estimationproblem as a stochastic censoring model and then proposedtwo algorithms to calculate the MLE (Maximum LikelihoodEstimation) for the tags location. The first algorithm is basedon a mixture model and the second on a time-difference-of-arrival. The authors in [16] replaced all the fixed anchors witha single drone that flies through a sequence of waypoints.At each waypoint, the drone acts as an anchor and securelydetermines the positions. This approach completely eliminates a r X i v : . [ c s . CR ] A p r he need for many expensive anchors. They propose three pathplanning algorithms that allow a drone to respectively measureand verify with a guaranteed precision a set of positions in asecure manner. Contributions:
As compared to previous works on se-cure localization, in this paper we leverage the use of theBlockchain technology to prevent attacks that can potentiallycompromise the localization algorithms. In fact, Blockchainis fully decentralized, and the verification of security is per-formed collaboratively using trusted entities and does not relyon third parties. All these advantages lead us to design anew secure localization algorithm for IoT devices. Besides,the Blockchain technique provides security at two levels ofprotocol execution. Indeed, it allows the protection of theexchanged localization data, and it guarantees the correctnessof the provided localization data.The remainder of this paper is organized as follows. SectionII gives an overview of the Blockchain technology and itsusage in IoT applications. Then, in Section III, we explainthe possible threat models and attacks. Then, our proposedsecure localization method is presented in Section IV and itsperformance evaluation in Section V. Finally, we conclude andgive some future works.II. T HE B LOCKCHAIN T ECHNOLOGY AND ITS U SAGE IN I O TIn this section, we introduce the Blockchain technology andits usage in the IoT field.
A. Blockchain Background
The Blockchain technology is a distributed ledger sharedbetween nodes in a peer-to-peer network. Basically, a ledgeris simply a database that is maintained and updated by everynode in the network.Each node in the network contains a copy of this ledger.The security of the Blockchain comes from the fact that blocksare cryptographically linked in a way that the alteration of oneblock requires the modification of all subsequent blocks in thechain. Moreover, as each node has a copy of the Blockchain,the attacker needs to make changes in at least 51% of nodesin order to pass fake information, which makes the attackextremely much harder.When a node has data to send (a transaction), it first signsit and then broadcasts it on the network. Each peer receivingthis transaction first verifies the signature and then forwardthe transaction to other nodes in the network. Special nodeson the network called miners try to packet this transactioninto a new block. For this purpose, first the miners verify thedata, and then, they compute a valid nonce that gives a hashthat satisfies a particular condition (generally that hash beginswith a specific number of zeros). The first miner that foundthe required nonce broadcasts this new block on the networkto be added to the Blockchain. To ensure that only validblocks are propagated on the network, before re-transmittingthe new block, a node makes extensive verification includingthe correctness of the nonce and the hash value and that the new block is linked to the latest block in the chain (i.e., itcontains the hash value of the latest block in the chain).
Consensus and Proof of Work:
In order to guarantee aspecific extension rate (number of blocks added per second),and to make tampering with a block a difficult task, Blockchainsystem introduces a consensus protocol that defines rules foradding new blocks to the chain. The main used consensusmechanism is Proof of work (PoW). In the proof of workconsensus protocol a miner needs to find a nonce (a randomnumber) that produces a hash satisfying certain condition(generally that the hash is less than a threshold). The task offinding such nonce is difficult (energy and time consuming);however, its verification by other nodes is easy. The difficultyin the PoW mechanism is updated every 2016 new blocksin such a way to guarantee the desired inclusion rate (in theBitcoin network it is fixed to 10 minutes) [17].
B. Blockchain in IoT applications
Blockchain is an emerging technology that is introduced inmany fields especially to provide security and distributed trustbetween peer-to-peer nodes. In what follows, we discuss someworks that leveraged the use of Blockchain technology forthe security of IoT applications. Blockchain technology wasintegrated into IoT to provide authentication and access controlin [18], [17], [19]. In [18], the authors proposed an accesscontrol mechanism based on Blockchain called
FairAccess .The proposed solution is a fully decentralized pseudonymousand privacy-preserving authorization management frameworkthat enables users to own and control their data. To fit theirmodel, the authors adapted the Blockchain into a decentralizedaccess control manager, and they used it to store the accesspermissions to resources. A Blockchain-based authenticationmechanism for IoT was proposed in [19]. The proposedsolution provides authentication of communicating things andthe integrity of transmitted and stored data through the creationof secure virtual zone called bubbles of trust . Before anycommunication can occur between two nodes of the samezone, a transaction must be transmitted and validated bythis Blockchain. This rule presents a main weakness of thissolution as it introduces a big latency (depends on the inclusionrate the Blockchain) in the communicating system.The authors in [20] proposed a distributed trust manage-ment scheme for VANET security based on clustering andBlockchain. Before adding a new block, the scheme requiresthe verification of the correctness of the message based onthe vehicle behavior which is controlled by the miner and thecredibility of the message decided by a Cluster Header.In [21], Dorri et al. addressed the heavy computation loadof the Blockchain technology and provided a lightweightBlockchain solution for IoT to secure smart home. The pro-posed solution eliminates the Proof of work and cryptocurren-cies concepts.III. T
HREAT M ODEL FOR S ECURE L OCALIZATION
Malicious and fake nodes in the IoT positioning systemcould intentionally send corrupted or fake information toisturb the localization system. Several attacks with variousimpact target the localization scheme in IoT, including: • Eavesdropping the devices position : in some localizationsystems the position of devices is sent to other nodes.If this information is sent without encryption, outsiderattacker might eavesdrop the communication and disclosethe IoT device position. This fact breaches the privacyof the user and makes the confidentiality of the user indanger [22], [23]. • Message forging : when a node sends its position in thenetwork, an attacker can intercept the message and forgeit by putting a false position. This behavior could disturbthe whole localization system in the network [24]. • Wormhole attack : In the wormhole attack, two maliciousnodes (called wormhole nodes ), strategically placed atdistant regions, collude to create a wormhole link. Thiswormhole link can be created using out-of-band or evenwired link. Through this wormhole link the maliciousnodes make victims (called affected nodes ), in one region,believe that they are close to the far apart nodes inthe distant region, which is a deceptive belief to attractand sniff victims data [25]. The wormhole attack has asignificant impact especially for localization mechanismsbased on RSSI or on the topology the network [9]. Paper[9] explains the impact of the wormhole attack at theDV-Hop localization algorithm. • Sybil attack : In this attack, the malicious node illegiti-mately presents several addresses to hide its real identityor to gain more access to the network resources [26].The Sybil attack has several forms; fabricated identities,stolen identities, simultaneous and non-simultaneous, etc.The Sybil attack has a severe impact on the localizationof nodes and might totally disturb the operation of thenetwork. The impact of this attack on the localizationsystem was explained in [12].IV. B
LOCK L OC : B LOCKCHAIN L OCALIZATION A LGORITHM FOR I O T A
PPLICATIONS
A. System Model and Assumptions
In our application, we consider a set of IoT devices (callednodes) that collaborate together to determine their position. Inour model, we have the following requirements: • Decentralization : In our model, there is no central entitythat computes the localization position of nodes: Allnodes are peers that collaborate in the localization system. • P2P Communications : In our proposed localizationscheme, the IoT devices communicate with its neighbors’nodes to determine their positions, in a peer-to-peernetwork architecture. • No Central Trust : The proposed network model does notrequire a central trusted entity that manages the securitybetween nodes or detects the existing of malicious nodes.The trust is provided thanks to the use of the Blockchaintechnology. The aforementioned characteristics of an IoT network modelmake the use of a Blockchain a necessity as this latter providesa secure distributed ledger and can ensure a distributed trustbetween the different peers IoT devices. Moreover, a success-ful security protocol needs to fit the constrained resourcesof IoT devices. Indeed, IoT networks generally consist ofheterogeneous devices such as smartphone, watch, wirelesssensor nodes, etc. These later have low computation, lowmemory, and low energy power [27]. Therefore, securityprotocols need to be lightweight and low-power consuming.
B. The BlockLoc Algorithm
Blockchain is considered as a framework to secure thedata exchanged between a set of peer nodes and to providetrust between them. In our work, we use Blockchain for twopurposes: (1) first to protect the localization data exchangedbetween IoT devices, (2) second to guarantee the correctnessof the given position data.In what follows, we propose BlockLoc, a secure localiza-tion scheme that uses the Blockchain technology to protectthe exchanged localization information and to guarantee thecorrectness of the claimed node’s position. In the BlockLoclocalization method, nodes collaborate to determine their po-sition. More precisely, a node needs to communicate withat least three anchors (i.e., nodes with known positions) inorder to determine its position. Triangulation is an exampleof a localization technique that uses at least three anchorsto determine the location of a fourth node. BlockLoc isalgorithm agnostic, which means it can be applied to anydistributed localization algorithms, that is based on locationdata exchange. However, instead of sending positions in anunprotected message that could be forged by attackers, thenode gets the neighbors positions from the secure Blockchainledger. This permits avoiding forging attacks. Moreover, amalicious node can provide a fake position to disturb thenetwork. To mitigate against this behavior, every claimed nodeposition is verified before adding it to the Blockchain ledger.For this purpose, the claimed node is required to send, inaddition to its position, the list of its neighbor nodes. Byverifying that the list of neighbor nodes are really in thevicinity of the claimed node, the localization scheme canexclude malicious data.We assume that each node has two keys: ( i. ) one key ispublic and known by all nodes and ( ii. ) the other key is keptprivate and secret.IoT applications generally use heterogeneous devices. Somedevices can be equipped with GPS and so can determine theirpositions (e.g. smartphone, car, etc.). Other devices might notbe equipped with a GPS, and so they need to run the proposedsecure localization scheme to determine their positions. Inorder to work properly, the BlockLoc scheme requires thata node knows the positions of at least three nodes and thecorresponding distance to them. These nodes will play therole of anchors. These anchor nodes can be either neighborsor not. If the anchor node is neighbor, the RSSI technique isused to estimate the distance between the node and this anchor Nonce 0x0123… Node Identity Position Position Source Neighbor List Hash=0x0567… Nonce Previous Hash Node Identity Position Position Source Neighbor List Hash=0x0123… Nonce 0x0567… Node Identity Position Position Source Neighbor List Hash Block N Block N-1 Block N+1
Fig. 1: The BlockLoc BlockChain Structure[7]. Otherwise, the DV-Hop technique is used [28]. Note thatRSSI stands for
Received Signal Strength Indicator and is usedas a link quality estimator in wireless communication [29] andalso used to estimate the distance between two nodes.The distance between two nodes can be deduced from thereceived signal power of nodes using the following equation[30]:
RSS ( d )( dBm ) = P tr − P loss ( d ) − τ log dd + X σ , (1)where d means the distance between the transmitting andreceiving nodes, RSS ( d ) indicates the signal power as receivedat a node located across a distance of d from the transmittingnode, d is the reference distance, P tr denotes the transmittedsignal’s power, P loss ( d ) means the signal power loss across thereference distance d , τ is the path loss exponent whose valuedepends on the medium of propagation, and X σ is the noise,which is described as a Gaussian random variable with zero asits mean and σ as the standard deviation. For more informationabout the RSSI and the DV-Hop methods, the reader can referto the work [28].More precisely, the proposed BlockLoc secure localizationscheme works as follows.
1) Initialization:
In our network model, we suppose the ex-istence of fixed anchors with known positions. These anchorsare generally relatively powerful nodes that could play the roleof miners; nodes responsible for adding new blocks to theBlockchain. Therefore, at the initialization phase, new blockscontaining the anchors’ positions are added to the Blockchain.These initial anchors’ positions serve in the verification of thefirstly localized nodes’ positions.
2) Blockchain construction:
First, each node knowing itsposition adds it to the Blockchain. For this purpose, it createsa block containing; its address, position and the list of neighbornodes, then it broadcasts this block to the network. Figure 1,shows the Blockchain structure.The message is sent signed. This means that the nodecomputes a digital signature of the message using its privatekey. Other nodes in the network verify the security of themessage by checking this signature using the sender publickey. This guarantee the authenticity (the message is actuallysent by the claimed node) and the integrity (the data has notbeen altered during transmission) of the message. Moreover, itavoids the Sybil attacks and the identity usurpation attacks aseach public key is associated with one address (a node cannot claim to have different identity). More precisely, the identityof a node is the hash value of its public key.When miner nodes receive the new block, they first verifythe message signature using the sender public key. Then, theminers verify the correctness of the claimed position. Thislatter verification consists in verifying that the claimed positionis in the vicinity of the given neighbor nodes. More precisely,they verify that the distance between the claimed positionand the position of a neighbor node is less or equal to thevalue of the communication range. If one of these verificationoperations fails, the block is ignored and the node position isexcluded from the localization system. In case both verificationoperations succeed, miner nodes compete to find the goodnonce (the nonce that satisfies the required PoW consensus).The first miner that computes the required nonce broadcast thenew block to the network. The new block contains in additionto the received data (node address, position and neighbor list)a hash code and a copy of the hash code of the last block inthe chain (this permit to link blocks between them).
3) Node Localization:
In most existing localizationschemes, the localization is based on the existence of anchornodes. More precisely, a node with unknown location (wecall it unknown node ) needs to have a connection to at leastthree anchor nodes in order to be localized. The lack ofenough anchors nodes leads to the failure of the unknown nodelocalization. To avoid this limit, In the proposed BlockLocsecure localization scheme, an unknown node can serve bythe already localized nodes.This permits the localization of the node even without theavailability of fixed anchor nodes. Although it can exist severallocalized nodes, the unknown node prioritizes the closestones. For this purpose, the unknown node first collects positions (1-hop neighbor are the nodes that are onehop far from the unknown node). If it does not get at leastthree neighbor’s responses, it makes a new round and contact2-hop neighbor nodes and so on, the number of neighbor hopswill be incremented at each new round until receiving at leastthree responses. More precisely, when a node A wishes to belocalized, first it sends a discover-message to its neighborswith hopcount value equal to 1. The hopcount field permitsto decide the number of hops the message is traveling in thenetwork. Each neighbor node receiving this message and thatis already localized, responds by sending a message containingits identity . By receiving this response message the node A,rst extracts the corresponding neighbor position from theBlockchain. Then, the node A computes the distance to thisneighbor node using the RSSI method. The RSSI methodis used as the neighbor is a one-hop neighbor. If the nodeA receives at least three responses from three neighbors, itestimates its positions using the Triangulation method [31].Otherwise, the node A starts a new round and sends a new discover-message with hopcount value equal to 2. This processof incrementing the hopcount value is repeated until the nodeA receives at least three responses from already localizednodes. When the responder node is a one-hop neighbor to thenode A the RSSI method is used; However, when the node isnot directly connected to the node A, the RSSI method cannotbe used and so the DV-Hop method is used. The idea of theDV-Hop message is to first compute the average hop-distance,then the estimated distance will be equal to the average hopdistance multiplied by the number of hops [28].As it can be noticed, in our proposed localization method,the nodes’ positions are not sent in the network. Instead, theidentities of nodes are sent along with their positions areextracted from the Blockchain. This characteristic permits toavoid the eavesdropping attack and preserve the privacy of theusers. V. P
ERFORMANCE E VALUATION
This section presents the performance of the proposedsecure localization scheme and the impact of the securityimprovement on the accuracy of the localization schemeunder the presence of malicious nodes. In the BlockLocproposed method, the security mechanism is based on theuse of the Blockchain. To highlight the impact of this se-curity mechanism, we have implemented two versions of thelocalization method. One version without any security scheme(called HDLoc for Hybrid DV-Hop Localization, which is apreviously published improvement of the DV-Hop algorithm)and the second version with BlockLoc security mechanism(called SecHDLoc for Secure Hybrid DV-Hop Localization).Furthermore, we have considered malicious nodes, and wetested the impact of these nodes on the performance of thelocalization scheme. The malicious behavior that we take intoconsideration is the modification of nodes’ positions. Moreprecisely, the malicious node sends a modified value of itsposition. In our simulation, we introduce an error value of50% which means that:
Malicious position = . x real position , (2) A. Simulation Model
In our simulation, we have used a wireless sensor networkconsisting of a fixed number of sensor nodes being 100. Thesenodes were randomly deployed in an area of 100 x m . Weassume that all the nodes in the network have the same charac-teristics. We also assume symmetric links among neighboringnodes, i.e., if node A i can receive a packet transmitted by A j ,then vice versa is also true. We used Matlab for implementingour simulations. The communication range between nodesis fixed to 30 m . During these simulations, the number of malicious nodes vary in (10%, 20%, 30%, 40%, 50%). Foreach simulation scenario, we repeat the experiment ten timeswith new randomly generated nodes locations. B. Simulation Results
Figure 2 shows the impact of the increase of the number ofmalicious nodes on the localization accuracy. More precisely,we have considered two scenarios. In the first scenario, theanchor rate is equal to 20%, and in the second scenario, theanchor rate is equal to 50%. In a real scenario, the anchorsare simply IoT devices with a fixed GPS location.In Figure 2a, where the number of anchor rate is 20%, wenotice that the secure version of the localization algorithm(SecHDLoc) is slightly affected by the number of maliciousnodes. This is due to the fact that malicious nodes are detectedand eliminated from the localization process. However, as itcan be seen in Figure 2a, the basic version of the algorithm(HDLoc) is sharply is affected by the number of maliciousnode and the localization error reached almost 16 m when thenumber of malicious nodes is 50%, whereas it is only 4 m inthe secure version.The increase of the anchor rate, Figure 2b, improves theaccuracy of the secure version of the algorithm, however, itdecreases the accuracy of the insecure version of the algorithm.This can be explained by the fact that the increase of anchorrate also increase malicious anchor rate (as malicious nodeare taken randomly and can be an anchor rate), and the errorintroduced by an anchor node has more impact than the errorintroduced by other nodes.VI. C ONCLUSION
In this paper, we have proposed a secure localization schemebased on Blockchain. The proposed algorithm takes advan-tage of the distributed and the decentralized characteristic ofBlockchain to provide a trustful framework of informationsharing between nodes. Thanks to Blockchain, a maliciousnode cannot inject fake data to the localization mechanismas all data need to be verified and checked before adding itto the Blockchain. The performance evaluation demonstratesthe improvements of the security mechanisms on the accuracyof the localization algorithm under the presence of maliciousnodes. More precisely, the introduced security mechanismsminimize the localization error to the 1 / EFERENCES[1] D. Evans, “The internet of things how the next evolution of the internetis changing everything (april 2011),”
White Paper by Cisco InternetBusiness Solutions Group (IBSG) , 2012.
Malicious Nodes Rate (%) Lo c a li z a t i on E rr o r ( m ) SecHDLocHDLoc (a) Anchor rate 20%
10 15 20 25 30 35 40 45 50
Malicious Nodes Rate (%) Lo c a li z a t i on E rr o r ( m ) SecHDLocHDLoc (b) Anchor rate 50%
Fig. 2: Localization error vs Malicious nodes rate [2] M. A. Al-Garadi, A. Mohamed, A. Al-Ali, X. Du, and M. Guizani, “Asurvey of machine and deep learning methods for internet of things (iot)security,” arXiv preprint arXiv:1807.11023 , 2018.[3] A. Koubaa, B. Qureshi, M.-F. Sriti, A. Allouch, Y. Javed, M. Alajlan,O. Cheikhrouhou, M. Khalgui, and E. Tovar, “Dronemap planner:A service-oriented cloud-based management system for the internet-of-drones,”
Ad Hoc Networks
IEEE Access ,vol. 6, pp. 13 810–13 824, 2018.[5] F. Yang, J. Li, T. Lei, and S. Wang, “Architecture and key technologiesfor internet of vehicles: a survey,”
Journal of Communications andInformation Networks , vol. 2, no. 2, pp. 1–17, Jun 2017. [Online].Available: https://doi.org/10.1007/s41650-017-0018-6[6] F. Khelifi, A. Bradai, A. Benslimane, P. Rawat, and M. Atri, “A surveyof localization systems in internet of things,”
Mobile Networks andApplications , pp. 1–25, 2018.[7] A. Koubˆaa and M. B. Jamˆaa, “Taxonomy of fundamental conceptsof localization in cyber-physical and sensor networks,”
Wirel. Pers.Commun. , vol. 72, no. 1, pp. 461–507, Sep. 2013. [Online]. Available:http://dx.doi.org/10.1007/s11277-013-1025-z[8] V. Sharma, D. N. K. Jayakody, I. You, R. Kumar, and J. Li, “Secure andefficient context-aware localization of drones in urban scenarios,”
IEEECommunications Magazine , vol. 56, no. 4, pp. 120–128, April 2018.[9] H. Chen, W. Lou, Z. Wang, J. Wu, Z. Wang, and A. Xia, “Securing dv-hop localization against wormhole attacks in wireless sensor networks,”
Pervasive and Mobile Computing , vol. 16, pp. 22–35, 2015.[10] J. Li, D. Wang, and Y. Wang, “Security dv-hop localisation algorithmagainst wormhole attack in wireless sensor network,”
IET WirelessSensor Systems , vol. 8, no. 2, pp. 68–75, 2017.[11] P. Zhang, S. G. Nagarajan, and I. Nevat, “Secure location of things(slot): Mitigating localization spoofing attacks in the internet of things,”
IEEE Internet of Things Journal , vol. 4, no. 6, pp. 2199–2206, 2017.[12] Y. Yuan, L. Huo, Z. Wang, and D. Hogrefe, “Secure apit localizationscheme against sybil attacks in distributed wireless sensor networks,”
IEEE Access , vol. 6, pp. 27 629–27 636, 2018.[13] M. B. Jamaa, A. Koubaa, and Y. Kayani, “Easyloc: Rss-basedlocalization made easy,”
Procedia Computer Science
ThirdInternational Conference on Communications and Networking , March2012, pp. 1–7.[15] T. He, C. Huang, B. M. Blum, J. A. Stankovic, and T. Abdelzaher,“Range-free localization schemes for large scale sensor networks,” in
Proceedings of the 9th annual international conference on Mobilecomputing and networking . ACM, 2003, pp. 81–95.[16] P. Perazzo, F. B. Sorbelli, M. Conti, G. Dini, and C. M. Pinotti, “Dronepath planning for secure positioning and secure position verification,”
IEEE Transactions on Mobile Computing , no. 1, pp. 1–1, 2017. [17] E. F. Jesus, V. R. Chicarino, C. V. de Albuquerque, and A. A. d. A.Rocha, “A survey of how to use blockchain to secure internet of thingsand the stalker attack,”
Security and Communication Networks , vol.2018, 2018.[18] A. Ouaddah, A. Abou Elkalam, and A. Ait Ouahman, “Fairaccess: a newblockchain-based access control framework for the internet of things,”
Security and Communication Networks , vol. 9, no. 18, pp. 5943–5964,2016.[19] M. T. Hammi, B. Hammi, P. Bellot, and A. Serhrouchni, “Bubbles oftrust: A decentralized blockchain-based authentication system for iot,”
Computers & Security , vol. 78, pp. 126–142, 2018.[20] A. Kchaou, R. Abassi, and S. Guemara, “Toward a distributed trustmanagement scheme for vanet,” in
Proceedings of the 13th InternationalConference on Availability, Reliability and Security . ACM, 2018, p. 53.[21] A. Dorri, S. S. Kanhere, R. Jurdak, and P. Gauravaram, “Blockchain foriot security and privacy: The case study of a smart home,” in . IEEE, 2017, pp. 618–623.[22] O. Cheikhrouhou, “Secure group communication in wireless sensornetworks: a survey,”
Journal of Network and Computer Applications ,vol. 61, pp. 115–132, 2016.[23] O. Cheikhrouhou, A. Koubaa, G. Dini, H. Alzaid, and M. Abid, “Lnt: Alogical neighbor tree secure group communication scheme for wirelesssensor networks,”
Ad Hoc Networks , vol. 10, no. 7, pp. 1419–1444,2012.[24] O. Cheikhrouhou, A. Koubˆaa, G. Dini, and M. Abid, “Riseg:A ring based secure group communication protocol for resource-constrained wireless sensor networks,”
Personal Ubiquitous Comput. ,vol. 15, no. 8, pp. 783–797, Dec. 2011. [Online]. Available:http://dx.doi.org/10.1007/s00779-011-0365-5[25] Y. Xiao, H. Chen, and F. H. Li,
Handbook on sensor networks . WorldScientific, 2010.[26] S. T. Patel and N. H. Mistry, “A review: sybil attack detection techniquesin wsn,” in . IEEE, 2017, pp. 184–188.[27] O. Cheikhrouhou, A. Koubaa, M. Boujelben, and M. Abid, “Alightweight user authentication scheme for wireless sensor networks,”in
ACS/IEEE International Conference on Computer Systems andApplications-AICCSA 2010 . IEEE, 2010, pp. 1–7.[28] O. Cheikhrouhou, G. M Bhatti, and R. Alroobaea, “A hybrid dv-hop algorithm using rssi for localization in large-scale wireless sensornetworks,”
Sensors , vol. 18, no. 5, p. 1469, 2018.[29] N. Baccour, A. Koubˆaa, L. Mottola, M. A. Z´u˜niga, H. Youssef,C. A. Boano, and M. Alves, “Radio link quality estimation inwireless sensor networks: A survey,”
ACM Trans. Sen. Netw. ,vol. 8, no. 4, pp. 34:1–34:33, Sep. 2012. [Online]. Available:http://doi.acm.org/10.1145/2240116.2240123[30] Y. Xu, X. Luo, W. Wang, and W. Zhao, “Efficient dv-hop localizationfor wireless cyber-physical social sensing system: A correntropy-basedneural network learning scheme,”
Sensors , vol. 17, no. 1, p. 135, 2017.[31] R. I. Hartley and P. Sturm, “Triangulation,”