Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) Security Implementation and Verification in Smart Buildings 1st Walid Miloud Dahmane 2nd Ouchani Samir 3rd Hafida Bouarfa saad dahlab university LINEACT, École d’Ingénieur CESI saad dahlab university Blida, Algeria France Blida, Algeria walid.miloud.dahmane@gmail.com souchani@cesi.fr hafidabouarfa@hotmail.com Abstract—The homes are dangerous environments like outside humidity, noise, light, etc). Nowadays different numerical since it contains risks affect on the life of the inhabitant (humid- models are available to describe the vapor balance of transient ity, temperature, noise, light, etc.), especially with the increase water in a room and predict indoor humidity. A typical of the attention on smart homes and buildings in the previous few years where studies focused on the IoT domain exclude room moisture balance includes water vapour production by partially these risks. Smart homes/buildings are equipped with moisture sources (humans, plants,. . . ), convective water vapour IoT objects that capture the conflicting changes in a controlled transfer with ventilation air, and water vapour exchange with manner and introduce actions that stop or declare the existing the building fabric and furniture.The water vapour exchange threats. A mechanism that guarantees to the inhabitant a stable between room air and surrounding materials (walls and fur- and comfortable life is more than mandatory. In this context, we propose a global approach that defines the architecture of a smart niture) is governed by three physical processes: the transfer home/building by formalizing the main nodes (sensors, actuator, of water vapour between the air and the material surface, the server, etc.) and the technologies that bind them. Further, we moisture transfer within the material and the moisture storage define the characteristics and the functioning of nodes by a formal within the material. The existing models mainly differ in the representation in the form of state machines, the applicable way this last part of the moisture balance is described [17]. In norms to build a secure environment, and further the security measures that must respect them in order to guarantee a general, sensors communicate directly with the home gateway protected environment. We finished our study by experimentation and feed the system information with regards to the obtained with Uppaal, a verification and validation tool, to ensure the environment measures, for example light intensity inside a accuracy of the system operations that showed a satisfactory particular room, temperature inside and outside the home and results. motion sensing to name a few [29]. Index Terms—Smart Home, Smart Building, Home risks, IoT, MQTT Protocol, Formal verification, Simulation, Uppaal. In this paper, we propose a smart living framework by modeling the different components needed for an indoor environment and developing a trustworthy architecture that I. I NTRODUCTION ensure the well functioning correctness of such system, and For a better living quality, the smart spaces paradigm aims also its configuration and control. First, we rely on the existing at constructing advanced service infrastructures that follow limitations and the requirements for a home that can affect the ubiquitous computing approaches where smart objects the inhabitant like humidity which causes corrosion coating are executed on a variety of digital devices and services are of the wall and household furniture, the appearance of molds constructed as interaction of agents in a communication envi- and bacteria, the temperature also has to be regulated in the ronment [19]. Recent advances in intelligent computer systems home according to the outside climate, loud noise especially and communications have created the necessary conditions for at night, the handicapped can not open the doors of the the networking of a wide variety of heterogeneous devices. room, natural and artificial phenomena such as the earthquake This led to the integration of short-range mobile transceivers and fire that threatens the life of the human. The proposed into everyday life objects and has enabled new forms of solutions consider all indoor issues, implement sensors for communication between objects and even between people and each measure, collect data in real time and make reactions objects. The concept of smart devices, i.e. the inclusion of to prevent risks. software, identifiers and networking to devices typically not The proposed framework is a web service based solution computerized, led to the “Internet of Things” (IoT) [7]. The where sensitive nodes are indoor planted and their measures main feature of this technology is the integration of heteroge- change in real time. The architecture proposed for the frame- neous and action elements (actuators) in a distributed system work considers different classes of nodes. A database node which performs different actions based on the information containing the collected data by sensors, a server node that gathered by the sensors combined with the requirements of ensures the communication and the reliability between nodes, the particular application [25]. and reacts when necessary by sending the appropriate control The inside environment has several factors that can affect it commands; the actuator node executes the received commands or the life of inhabitants or both at the same time (temperature, from the server and/or external actors who can extract or edit home data. The architecture uses MQTT protocol [28] same family of LPWANs. The goal is to know the number of to ensure a reliable communication between the predefined gateways needed to cover the city (inexpensive or not), and internal nodes. Further, the architecture implements a precise to know the benefits in return after deployment. They experi- constraints and requirements for the communication and dur- mented two tests, the first installs LoRaTM network in a 19- ing executing actions. Otherwise, the nodes do not respecting history building to measure temperature and humidity, using certain conditions are considered as unacceptable nodes. Fi- one single gateway and 32 nodes. The second estimates the nally we ensure the functional correctness of the nodes and number of gateways required to cover the city of Padova.They their secure communication by simulation and verification in placed a gateway with no antenna gain at the top of two history Uppaal tool [3]. The results show that the proposed framework buildings to assess the ‘worst case’ coverage of the topology is a deadlock free, secure, and respecting the indoor living since LoRaTM technology allows to cover a cell of about 2 requirements. km of radius. With simple calculations they concluded that The remainder of this paper is organized as follows. Section to cover Padova city that has about 100 square kilometers, it II presents the related work and compares it with the proposed needs 30 gateways. At the present, LoRaTM has an acceptable framework detailed in Section III. Then, the implementation coverage in worst cases, but the number of gateway ports with the experimental results are shown in Section IV. Finally, is limited and does not satisfy progressive evolution of IoT Section V concludes the paper and provides hints on the future technology. works. A. Zanella [32] apply the principles of smart cities for Padova city to collect environmental data. The architecture is II. R ELATED W ORK composed of constrained IoT sensors, a database server which In literature, we review the existing work related to IoT use technologies CoAP3 , 6LoWPAN4 , unconstrained devices modeling, functional analysis, network architectures, and ap- that use traditional technologies like HTML. The interconnec- plication in real life with concrete cases. tion between users and sensors is made by an intermediary Ouchani [22] proposes a security analysis framework for gateway and HTTP-CoAP proxy-grown that plays the role IoT that covers the probability and costs of actions, formalizes of translator between the two sides. During a week of tests, IoT, analyzes the correctness and measures their security level the results show how do people react with different situations by relying on the probabilistic model checking PRISM. To and phenomena, for example benzene consumption at the end ensure the functional correctness of an IoT-based system, of weeks. This architecture allows the compatibility between Ouchani develops five steps: defines the IoT components, for- constrained and unconstrained devices through a cross proxy. malizes the architecture in a process algebra expression. Then, In general, the constrained physical and link layer technologies it expresses the IoT requirements in PCTL and transforms the are characterized by a low energy consumption, the transfer IoT model into the PRISM input language. Finally, PRISM rate and data processing in constrained devices is relatively checks how much a requirement is ensured on the IoT model. low, but the dependence on unconstrained ones increase in However, the proposed framework involves a large amount cost. of data and messages which make the probabilistic model Based on the reviewed literature, we found few works that checking expensive in terms of time and memory. detail well the components of an indoor environment and their Moreno-Salinas [13] proposes a method that detects the formal semantics, and less of them discussing a trustworthy optimal position of sensors to receive information from several communication between components. The proposed contribu- targets. To find the perfect place, they rely on FIM1 to measure tion considers these issues and we believe it is easy to extend the amount of information that a random variable (sensor) and deploy a more secure smart building/home system. carries about a parameter that is sometimes unknown (target). III. F RAMEWORK After several progressive tests, they use two separated tests, the first tries to find the optimal position for a sensor that receives Figure 1 illustrates the steps to how construct a secure smart data from a target transmitter with a known placement. This building/home system and analyze it. The system’s architec- first test considers one sensor and one target, eight sensors and ture is composed from a set of nodes, security constraints and one target, four sensors and two targets, and five sensors and management mechanism, and the communication protocols. three targets. The second one finds the optimal positions of The nodes are active/passive objects to collect the needed sensors with unknowns positions experimenting five sensors environment measures. The communication protocols ensure and two targets, then five sensors and three targets. However, how well the connection between nodes is established and FIM showed significant results for a small amount of objects the measured data are packed and encrypted. The security but the cost of time computing is expensive when the target management mechanism reinforces the architecture in order is unknown in a known area. to create a protected system. It develops a set of security rules Centenaro [11] studies a start topology of LPWANs2 in including the authentication and identification of nodes, the smart cities where the used network LoRaTM belongs to the control access, and how to keep the availability of services. 1 Fisher Information Matrix. 3 Constrained Application Protocol 2 LowPower Wide Area Networks. 4 IPv6 Low power Wireless Personal Area Networks The analysis step enables the verification of the accuracy of (the evaluation of dynamic attributes) to another one Sj . The the implemented architecture with respect to the security rules. following lists the set of possible actions. Finally, the results show the different scenarios, traces, or • turnOn/turnOff to turn on/off the smart object [1]. errors that might affect the security and the well functioning • send/receive to send/receive data to/from another IoT of the architecture in order to decide or not its deployment. node [33]. • collectData to collect the received information [33]. Reinforcement, • applyAction apply an action after getting command Security management • Control Application • Authentication [33]. • Availability Architecture • Smart home • Smart building Results • encrypt/decrypt to encrypt/decrypt a message. • Scenarios • Traces • authenticate grants the possibility to send data. Object • Errors • Attributes • Behavior Formulation, Transformation • Correct • Secure We define in Definition 1 a smart node that can be a sensor, • Type • Applicability Composition actuator, broker, database, server, or smartphone. Analysis • Verification • Simulation Outputs Definition 1 (Smart node). A smart object SoT ∈ SoT is a Communication Formulation • Protocols • Data tuple hID, Att, Σ, Bi where: Establishment 1) ID is a finite set identifiers idi ∈ ID{Oi , i ∈ N where Fig. 1: A Security and Analysis Framework for Smart Homes and id∅ ∈ id is an empty object. Buildings. 2) Att : ID → 2T is a function that assigns for each object a sequence of attributes. A. Smart object 3) Σ is the set of possible actions for an objects, A smart object (SoT ∈ SoT) is identified by a set of dynamic 4) Beh : ID → B returns the expression that precises and static attributes (T). The dynamic attributes are classified the behavior of an object in the dominant case where : into two categories: data (di of type real) and flags (fi of B ::= Start.actions +g actions.End where actions = type Boolean). In the following, we cite the most used static α|α.actions such as α ∈ Σ and +g is a deterministic attributes that describe the physicality and the technicality of choice with respect to a guard g. an SoT. Example 1 (Smart object). Based on Definition 1, the se- • The identifier (id ∈ ID): is the unique reference to SoT, mantics of a general sensor is the state machine depicted in our case id is IPv6 [10]. in Figure 2 where states s0 , s1 , s2 , s3 stand respectively • The connectivity ( COn ∈ T) describes when devices have for Is_On, detection, declaration, Is_Of f . The attributes extensions to connect to each other [6]. values specifying a state change regarding the executed action. • The battery life ( BLi ∈ T): represents the longevity of The actions α1 , α2 , α3 , α4 , and α5 represent respectively a battery [15]. turn_on, detect , send,turn_off, and initialize.The dynamic • Powered by electricity ( PEl ∈ T): when SoT can be attributes (d and f) of a sensor are: d1 evaluates the energy, plugged with an electricity line. d2 measures other properties (smoke, noise, temperature,...), • Data security ( DSe ∈ T) informs about the ability to f1 : detection, f2 : availability, f3 : alerte_msg. Each state is encrypt informations stored or sent [8]. presented by the following predicates where M ax_V al is the • Small size ( SSi ∈ T): describes the volume of SoT. maximum for the measure related to the smart object. • High product quality ( HPr ∈ T) indicates the possibility 1) Js0 K = (d1 > 0)∧(d2 < M ax_V al)∧(f1 )∧(f2 )∧(¬f3 ) to increase the service life and to reduce the cost of 2) Js1 K = (d1 > 0) ∧ (d2 >= M ax_V al) ∧ (f1 ) ∧ (f2 ) ∧ maintenance. (¬f3 ) • Constrained device ( CDe ∈ T) describes if a cheaper 3) Js2 K = (d1 > 0)∧(d2 >= M ax_V al)∧(f1 )∧(f2 )∧(f3 ) device can cover a specific space [24]. 4) Js3 K = (d1 = 0) ∧ (d2 = 0) ∧ (¬f1 ) ∧ (¬f2 ) ∧ (¬f3 ) • Price ( PRi ∈ T) helps in the budget management [4]. • Service availability ( SAv ∈ T) to check if the device works continuously or not [14]. s1 • Minimum error ( MEr ∈ T) increases the quality of α2 α3 service [18]. • Easy to maintain ( EMa ∈ T) is to reduce time, effort and the cost of maintenance. start s0 α4 s2 α5 • Required a low connection rate ( RLo ∈ T): to stay α4 connected in the worst case [12]. α1 α4 • Interoperability of nodes ( INo ∈ T) defines the technolo- s3 gies supported by the node [31]. The behavior of an object is the effect of the executed Fig. 2: The state machine of a sensor. actions (Σ) that allows it to transfer from its current state Si B. Smart environment We define a smart environment sEnv as a structured physi- cal infrastructure, building or home, that carries smart nodes. sEnv is composed of at least two smart rooms/locations dis- jointed by separators like walls, doors, and windows. To collect information and sensitive data, smart nodes are connected with a precise architecture mechanism that helps them to communicate easily through a dedicated protocols. Definition 2 (Smart Environment). A smart environment sEnv is a tuple of hE, L, SoT, pl, dli, where: Fig. 5: The architecture of Smart Home 1) E is the environment name/id, 2) L ={R1 , ..., Ri , ..., Rn |i, n ∈ N } is the set of loca- tions/rooms (Ri ) composing E, access the internet connection. The third level is IP objects 3) SoT = { SoT1 , ..., SoTm |m ∈ N} is the set of smart nodes use wireless technology like Wi-Fi, Bluetooth, 4G... in E, The fourth level has processing devices like router, firewall 4) PL = { pl1 , ..., pln |n ∈ N} is the set of physical structure and switch, they are used to make an interconnection between that defines E, smart home objects and they are like a point between the 5) DL = { dl1 , ..., dln |n ∈ N} is the set of logical architecture outdoor and indoor smart home. that connects SoT. The fifth level is the set of APIs and devices outside smart home that can access the smart home interior objects. Figures 3 and Figure 4 show respectively an abstraction of the physical structure of E and the logical architecture between nodes in E. D. Communication In this part we will present some protocols that can be E used in the proposed framework that deals with architecture as the one showed in Figure 6. Herein we present the adopted protocols by the framework. Ri Rj Fig. 3: The physical structure of E. Ri Rj
Fig. 6: Operation protocol MQTT in architecture. SoT1 SoTm SoT1 SoTm MQTT: It is a machine-to-machine connectivity protocol Fig. 4: A Logical/Digital structure in E. designed as an extremely lightweight publish/subscribe mes- saging transport [2]. The operations of this protocol passes through steps shown in Figure 6, where it is applied on Smart C. Architecture room, and it is the first level represented in the architecture. The architecture is grouped into five main levels depicted 1) A sensor collects information (temperature, fire, humidity, in Figure 5: etc.) then it publishes the data to the broker. The first is the most important because it contains sensors 2) The database subscribes into the Broker that is periodi- that capture the state of smart home periodically then they cally keep track of the retrieved data. report if there is a contradictory case (fire, humidity, high 3) The web server subscribes into the Broker and receives temperature, ...), the analysis devices as the database, web the published sensors data. server and broker save or process the signals of the sensors 4) The web server, including smart applications, presents the then give the actuators the commands to do the necessary appropriate command, and pulls it into the MQTT Broker. actions. 5) The actuators subscribe in the Broker then it receive and The second level is the set of objects referenced by an IP execute the commands. address linked with the router by a network wire; they can 6) The application retrieves or updates the database values. 7) External actors, through web and smart applications, A. MQTT protocol test communicate securely with web server. We test the MQTT protocol via a scenario simulates the case ONVIF: It is used to establish a communication between of fire in smart home, the first scenario steps are presented in the network camera and a point outside the building in order the figure 7, our system function without deadlock. to monitor its status in real time. Http: People authenticated in the web server can access through an API that uses this protocol to view or edit infor- mation about the building. VoIP: Phones equipped with a network card can make calls using this protocol. Ethernet: It is a data link layer protocol in the Open Sys- tems Interconnection (OSI) model that allows objects affiliated with the same LAN to interchange data. E. Security The digital environment always at risk, for this we rely on the security side in our approach to avoid information theft, data interception or disservice. We consider the following five Fig. 7: MQTT Simulation Scenario. concepts in order to stop or decrease threats. • Confidentiality: ensure that each data access only by objects (people, devices) that we define them through en- B. Connection with distant points crypting data with a strong encryption method. Ignoring The distant smart home users use IoT nodes to access smart this principle can cause a destruction of information. home objects via the internet connection. In this point we will • Authentication: Some smart home objects (such as the study two examples, the first is a user that accesses by his server) request objects that want to access it to define its smartphone to the smart home server in order to extract data identification in order to prevent unauthorized access. from the database, and the second is a web API accesses to a • Data Integrity: Man in the middle [20] can intercept the Webcam Home, system operation does not give errors. flow of data between IoT objects, change it then send it back to the receiver. So we use some mechanisms like C. Exceptional cases: hashing [26] (MD5 and SHA-2) and electronic signatures The nature of these tests simulates contradictory cases that [9] to control if the message is changed or no. affect the exchange of messages, in this test we check the • Access control: Smart home objects with their security operation of system with three cases contradictory with the levels allow functions according to a predefined autho- natural operation(Webcam not linked, Firewall prevents web- rization and prevention rules. The architecture supports cam contact and The API does not authenticate the webcam), firewalls [30] at the gateway level that manage the input the resulat was that the test procedure is not finished. and output packets. Further, for security policies we are D. Security rules verification interested in access control mechanisms [16] (RBAC) and adapting the router by an access control list ACL [27]. Uppaal has a language called ’query language’ which allows • Non-repudiation: Since IoT objects always in contact it to edit rules after the construction of states machines of the is important to check the legitimacy of the sender and objects to test the accuracy of these objects. The language is the receiver. The most able method to realize that is the written according to specific norms and symbols.To verify the electronic certificate [21]. security rules, we express the query language to check these goals Confidentiality, Authentication, Data Integrity, Access IV. E XPERIMENTAL R ESULTS control and Non-repudiation. The verification results show that To test the accuracy of the proposed, we built it within all the security rules are checked and satisfied. the validation and verification tool Uppaal, by integrating the machine states of smart objects and create the smart home V. C ONCLUSION architecture where the smart home objects (composition of The approach suggests to deploy a complete theoretical states machines) react. The logic behind this composition and practical framework that builds secure smart homes and ensures that the proposed framework does not oppose the buildings in order to protect the inhabitants, the environment, requirements. First we ensure through simulation then ver- and to optimize the standard of living for an inhabitant. ification. The simulation is partitioned in four phases, the The proposed formalization considers the characteristics and first tests the operations of MQTT protocol, the second tests the behavior of smart nodes and facilitates the expression the connectivity with a external point, the third tests for of their operations. The flexibility of the architecture makes exceptional cases where IoT devices can not connect to each it applicable on different structures so it is not affected by other and finally we verify the satisfaction of the security rules the number of rooms, doors, style of construction, nature of that we must respect in the proposed system. wall, etc. The framework covers a number of technologies, and fit the compatibility between them. Further we propose [17] Arnold Janssens and Michel De Paepe. Effect of moisture inertia models a set of security rules in order to reinforce the architecture on the predicted indoor humidity in a room. Proceedings of the 26th AIVC Conference, 01 2005. and to check how much it is secure. For the security and [18] Amos Kingatua. Iot system tests :: Checking for failure. the correctness analysis that helps to reduces the error rate [19] Dmitry G. Korzun, Sergey I. Balandin, and Andrei V. Gurtov. De- after deployment, we rely on the simulation and the formal ployment of smart spaces in internet of things: Overview of the design challenges. In Sergey Balandin, Sergey Andreev, and Yevgeni Kouch- verification that showed the strong and the weak points of the eryavy, editors, Internet of Things, Smart Spaces, and Next Generation a defined architecture. The results show that the implemented Networking, pages 48–59, Berlin, Heidelberg, 2013. Springer Berlin architecture is free from deadlocks, simulate the reality, and Heidelberg. [20] Avijit Mallik, Abid Ahsan, Mhia Md. Zaglul Shahadat, and Jia-Chi Tsou. respect the security rules. Man-in-the-middle-attack: Understanding in simple words. 3:77–92, 01 As a future work, we intend to extend the framework to 2019. support smart cites as first step. Then we look to how to [21] Mick O Brien and George Weir. Understanding digital certificates. 06 2019. optimize the architecture features such as minimizing energy [22] Samir Ouchani. Ensuring the functional correctness of iot through consumption, large-scale coverage by limited number of gate- formal modeling and verification. In Model and Data Engineering - ways. Further, from a security perspective we will increase 8th International Conference, MEDI 2018, Lecture Notes in Computer Science, pages 401–417. Springer International Publishing, 2018. the security level by relying on a distributed architecture [23] Luis Sanchez, Luis Muñoz, Jose Antonio Galache, Pablo Sotres, Juan R. ”Blockchain”. Santana, Veronica Gutierrez, Rajiv Ramdhany, Alex Gluhak, Srdjan Krco, Evangelos Theodoridis, and Dennis Pfisterer. Smartsantander: Iot R EFERENCES experimentation over a smart city testbed. Computer Networks, 61:217 – 238, 2014. Special issue on Future Internet Testbeds – Part I. [1] Iot sensors. https://fiware-tutorials.readthedocs.io/en/latest/iot-sensors/, [24] A. Sehgal, V. Perelman, S. Kuryla, and J. Schonwalder. Management 2019. of resource constrained devices in the internet of things. IEEE Commu- [2] Mqtt. http://mqtt.org/, May 2019. nications Magazine, 50(12):144–149, December 2012. [3] Uppaal home. http://www.uppaal.org/, 2019. [25] Giorgos Sfikas, Charilaos Akasiadis, and Evaggelos Spyrou. Creating a [4] M. Aazam and E. Huh. Fog computing micro datacenter based dynamic smart room using an iot approach. 05 2016. resource estimation and pricing model for iot. In 2015 IEEE 29th [26] Rajeev Sobti and Geetha Ganesan. Cryptographic hash functions: International Conference on Advanced Information Networking and A review. International Journal of Computer Science Issues, ISSN Applications, pages 687–694, March 2015. (Online): 1694-0814, Vol 9:461 – 479, 03 2012. [5] A. Al-Fuqaha, M. Guizani, M. Mohammadi, M. Aledhari, and [27] Shipra Suman and Aditi Agrawal. Ip traffic management with access M. Ayyash. Internet of things: A survey on enabling technologies, control list using cisco packet tracer. International Journal of Science, protocols, and applications. IEEE Communications Surveys Tutorials, Engineering and Technology Research, 5:1556–1561, 05 2016. 17(4):2347–2376, Fourthquarter 2015. [28] Konglong Tang, Yong Wang, Hao Liu, Yanxiu Sheng, Xi Wang, and [6] S. Andreev, O. Galinina, A. Pyattaev, M. Gerasimenko, T. Tirronen, Zhiqiang Wei. Design and implementation of push notification system J. Torsner, J. Sachs, M. Dohler, and Y. Koucheryavy. Understanding based on the mqtt protocol. In 2013 International Conference on the iot connectivity landscape: a contemporary m2m radio technology Information Science and Computer Applications (ISCA 2013). Atlantis roadmap. IEEE Communications Magazine, 53(9):32–40, Sep. 2015. Press, 2013/10. [7] Luigi Atzori, Antonio Iera, and Giacomo Morabito. The internet of [29] Dhiren Tejani, Ali Al-Kuwari, and Vidyasagar Potdar. Energy conser- things: A survey. Comput. Netw., 54(15):2787–2805, October 2010. vation in a smart home. 05 2011. [8] Sachin Babar, Parikshit Mahalle, Antonietta Stango, Neeli Prasad, and [30] A. Wool. A quantitative study of firewall configuration errors. Computer, Ramjee Prasad. Proposed security model and threat taxonomy for the 37(6):62–67, June 2004. internet of things (iot). In Natarajan Meghanathan, Selma Boumerdassi, [31] G. Xiao, J. Guo, L. D. Xu, and Z. Gong. User interoperability with Nabendu Chaki, and Dhinaharan Nagamalai, editors, Recent Trends in heterogeneous iot devices through transformation. IEEE Transactions Network Security and Applications, pages 420–429, Berlin, Heidelberg, on Industrial Informatics, 10(2):1486–1496, May 2014. 2010. Springer Berlin Heidelberg. [32] A. Zanella, N. Bui, A. Castellani, L. Vangelista, and M. Zorzi. Internet [9] Jean-François Blanchette. The digital signature dilemma le dilemme de of things for smart cities. IEEE Internet of Things Journal, 1(1):22–32, la signature numérique. 2006. Feb 2014. [10] Dr. Lakshmi Devasena C. Ipv6 low power wireless personal area [33] Q. Zhu, R. Wang, Q. Chen, Y. Liu, and W. Qin. Iot gateway: Bridging- network (6lowpan) for networking internet of things (iot) - analyzing wireless sensor networks into internet of things. In 2010 IEEE/IFIP its suitability for iot. 9, 01 2016. International Conference on Embedded and Ubiquitous Computing, [11] M. Centenaro, L. Vangelista, A. Zanella, and M. Zorzi. Long-Range pages 347–352, Dec 2010. Communications in Unlicensed Bands: the Rising Stars in the IoT and Smart City Scenarios. IEEE Wireless Communications, 23, October 2016. [12] Y. Chen and T. Kunz. Performance evaluation of iot protocols under a constrained wireless access network. In 2016 International Conference on Selected Topics in Mobile Wireless Networking (MoWNeT), pages 1–7, April 2016. [13] Antonio M. Pascoal David Moreno-Salinas and Joaquin Aranda. Op- timal sensor placement for multiple target positioning with range-only measurements in two-dimensional scenarios. Sensors, 13(8), August 2013. [14] P. Desai, A. Sheth, and P. Anantharam. Semantic gateway as a service architecture for iot interoperability. In 2015 IEEE International Conference on Mobile Services, pages 313–319, June 2015. [15] Xenofon Fafoutis, Atis Elsts, Antonis Vafeas, George Oikonomou, and Robert Piechocki. On predicting the battery lifetime of iot devices: Experiences from the sphere deployments. In Proceedings of the 7th International Workshop on Real-World Embedded Wireless Systems and Networks, RealWSN’18, pages 7–12, New York, NY, USA, 2018. ACM. [16] David Ferraiolo and D Kuhn. Role-based access controls. 03 2009.