Energy‐efficient localisation of sensor nodes in WSNs using single beacon node
2020; Institution of Engineering and Technology; Volume: 14; Issue: 9 Linguagem: Inglês
10.1049/iet-com.2019.1298
ISSN1751-8636
AutoresPrateek Raj Gautam, Sunil Kumar, Akshay Verma, Arvind Kumar,
Tópico(s)Underwater Vehicles and Communication Systems
ResumoIET CommunicationsVolume 14, Issue 9 p. 1459-1466 Research ArticleFree Access Energy-efficient localisation of sensor nodes in WSNs using single beacon node Prateek Raj Gautam, Corresponding Author Prateek Raj Gautam prateekrajgautam@gmail.com orcid.org/0000-0002-2889-4275 Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this authorSunil Kumar, Sunil Kumar Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this authorAkshay Verma, Akshay Verma Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this authorArvind Kumar, Arvind Kumar Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this author Prateek Raj Gautam, Corresponding Author Prateek Raj Gautam prateekrajgautam@gmail.com orcid.org/0000-0002-2889-4275 Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this authorSunil Kumar, Sunil Kumar Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this authorAkshay Verma, Akshay Verma Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this authorArvind Kumar, Arvind Kumar Electronics and Communication Engineering Department, Motilal Nehru National Institute of Technology Allahabad, Prayagraj, IndiaSearch for more papers by this author First published: 01 June 2020 https://doi.org/10.1049/iet-com.2019.1298Citations: 1AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinkedInRedditWechat Abstract Identifying the location of randomly deployed sensor nodes in wireless sensor networks (WSNs) is a major challenge because sensor nodes have limited energy and computational capacity. This study presents a low complexity, scalable, and energy-efficient localisation scheme capable of localising dense as well as sparse deployment of nodes using a single beacon node (SBN). The nodes act in passive listening mode to receive information transmitted by the SBN to determine their locations. The energy consumption is reduced at the nodes as well as at the SBN. The SBN has an omni-directional antenna and a directional antenna (DA). DA can rotate to transmit beacons in different directions in the -plane. The time required and energy consumption for localisation are reduced at the node side as well as at the SBN side. The expressions for energy consumption and computational complexity are deducted analytically. The proposed scheme is verified with MATLAB simulation. The simulation results show a significant improvement in localisation accuracy and reduction in energy consumption. Nomenclature transmit power from the single beacon node (SBN) or base station (BS) radial distance from anchor node current communication range of anchor node size of increment in communication range of anchor node maximum communication range of anchor node 1 Introduction A wireless sensor network (WSN) is a subclass of the wireless ad hoc network. The WSNs consist of small and low-cost devices called the sensor node, which aims to collect various physical phenomena. These sensor nodes have limited battery, computational, and storage capacity, and often employs IPv6 over low-power wireless personal area networks to send and receive data with low power over short distances using a 2.5 GHz frequency band [1]. WSNs have touched lives in many areas such as health-care [2], biomedical, industrial, and household applications [3] because of their ease of deployment and low-cost [4]. WSNs are used for internet of things, gas-leakage and fire detection, monitoring, health-care, and indoor localisation. WSNs use hundreds of randomly deployed sensor nodes to gather information on various physical phenomena such as temperature, rain, and humidity. This information is transmitted to the base station (BS) for processing and control operations. However, the information at the BS is useful only if the location of the sensor node is known. The location of sensor nodes is often used to optimising routing protocols [5]. Thereby, the energy consumption is reduced. Hence, localisation is a crucial task in WSNs, and each node must know its location so that the information conveyed by it can be used to take proper action. Furthermore, as per the energy constraints in WSNs, the localisation process must be energy efficient. This study presents a low complexity, energy-efficient localisation process. The localisation process does not require any hardware modification at the node side. Thus, the cost and size of the node remain unchanged. The single beacon node (SBN) or BS does not need to be placed at a certain height above the -plane, and any wide beamwidth directional antenna (DA) can be employed at the SBN without affecting the accuracy of localisation. Localisation is performed in polar coordinates. So, the node knows its location . The location can be converted into Cartesian coordinates. Two types of beacons are transmitted from SBN, one for radial distance (R) estimation and another to estimate angle . The nodes act in passive listening mode, so its energy consumption is reduced in comparison with existing localisation processes. The final location depends on one distance estimated via received signal strength indicator (RSSI). Unlike other range-based localisation schemes where location estimation consists of three or more RSSI estimated distances, hence, the RSSI error does not accumulate. 1.1 Related work The localisation of nodes can be performed by triangulation, after estimating the distance of the node from three or more anchor nodes. RSSI from the anchor node [6] and a path loss model is used to estimate the distance. Global positioning system (GPS)-enabled nodes can use signals from GPS satellites to localise each node. Furthermore, the time of arrival (TOA), angle of arrival (AOA), and time difference of arrival are other commonly used techniques for the localisation of nodes [7-9]. Recently, some mathematical models and learning algorithms are proposed to improve the accuracy. In [10], a multiplicative distance-correction factor (MDCF) is used to counteract the inaccuracy of RSSI distance estimation. Multidimensional scaling with cooperative localisation over the small world and game theory has been used to improve localisation accuracy in [11, 12]. However, each technique has its limitations, such as TOA required synchronisation between the transmitter and receiver, RSSI faces the problem with multi-path fading and shadowing mostly in indoor conditions that affect the accuracy. GPS signal fails in indoor conditions and it is not preferred as it increases the size of the node and cost of the network. AOA requires additional hardware (antenna array and digital compass) and increases the distance between the transmitter and receiver degrading its accuracy [13]. Game theory and MDCF increase computational complexity and energy consumption. Therefore, an accurate and energy-efficient localisation scheme with low-computational complexity is required. Furthermore, no hardware modification at the node side is desired, and the number of beacon nodes (BNs) required should be small. 1.1.1 Ripple localisation algorithm (RLA) [14] In RLA, BNs vary the transmit power to change the communication radius, starting from in uniform steps of up to (maximum communication range). The beacon also contains these values, including the current communication radius () and the coordinate of the BN. The nodes in the region will receive multiple beacons with different power levels, and any node can estimate its distance from the BN using the values from the first beacon received as . Furthermore, multiple BNs are deployed in the region. Any dumb nodes after receiving beacons from three or more BNs can solve these distance equations to estimate its location. Distance estimation is susceptible to RSSI losses. Hence, the accuracy is affected as the final estimation is based on multiple distances, each having some deviation. Furthermore, it requires multiple BNs that increase the cost. 1.1.2 Localisation using mobile anchors with DAs (DIR) [15] DIR uses mobile anchors (BNs) with GPS and four DAs each pointing 90 away from the -plane. The anchors are moved on the -axis and -axis. The anchor nodes transmit beacons with their current location using their DAs. Dumb nodes receive these beacons and select a median of all the received 's and 's as their coordinates. This scheme is energy efficient and has low-computational complexity. However, it requires two or more anchor nodes each having four DAs, and these anchors must be moved across the region in two orthogonal directions. Furthermore, the accuracy depends on the accuracy of GPS (while moving anchor nodes), beamwidth, and orientation of DA. The schematic of the DIR is shown in Fig. 1. Fig. 1Open in figure viewerPowerPoint Schematic of DIR [15] 1.1.3 Localisation using beacons from rotating directional antenna (LBRD) [16] In LBRD, BS employs one DA placed at height h above the origin. The schematic of the LBRD is shown in Fig. 2. The DA has a small beamwidth of having a conical beam pattern. The direction to the DA can be changed to point any region on the -plane, the DA can be rotated to a full around the -axis () and azimuthal angle can be varied to point a region close to the BS, and it can be increased to point at a region far from the BS. The DA transmits beacons containing values of its height [h] and current direction [] and the nodes falling under the beamwidth will receive multiple beacons with different values of , each node calculates the average of the beacon values () received by them and estimate their location as radial distance , and . The energy consumption at the BS is high as the number of transmissions required from the BS is large, and the required height of the antenna at the BS is large (typically 13 m for a 100 m100 m region). Fig. 2Open in figure viewerPowerPoint Schematic of LBRD [16] The remainder of the paper is organised as follows. Section 2 describes the energy-efficient localisation of sensor nodes in WSNs using single beacon node (EE–LSB) process and the required hardware modifications at SBN. Simulation parameters, results, and discussion are presented in Section 3. Finally, Section 4 concludes the paper. 2 Energy-efficient localisation of sensor nodes in WSNs using single beacon node (EE–LSB) Any node can determine its location relative to the SBN if the distance from the SBN and angle at which it lies can be estimated. The proposed scheme is capable of determining these two components using one SBN. This section describes the assumptions taken, the hardware requirements of the SBN followed by the proposed scheme (EE–LSB) and its energy consumption. The notations used in this paper are listed in the Nomenclature section. 2.1 Assumptions The SBN and all nodes lie on the same -plane (). The energy transmitted outside the beamwidth of DA of SBN is zero. The nodes start data sensing, routing, and transmitting after the completion of the localisation process. 2.2 Hardware requirements of the BN The SBN needs two transmitting antennas, one omni-DA, and one DA. The omni-DA can vary its transmit power and DA is mounted on a mechanical setup capable of changing its direction in steps of , . The omni-DA transmits beacons with information to facilitate radial distance estimation (R) from the SBN. The DA transmits beacons that contain direction information. 2.3 Process description In EE–LSB, the SBN transmits two types of beacons, one from an omni-DA for radial distance estimation similar to RLA [14] and another from the DA for direction estimation. The SBN uses omni-DA to transmit a number of beacons each with different transmit power (), such that the communication range is . Depending on the distance of any node from the SBN, it can receive one or more beacons. As shown in Fig. 3 when transmitting with corresponding to , if the radial distance of any node from the SBN is less than , it will receive beacons transmitted with a communication range greater than or equal to (1) (2)The beacon signal contains calculated by (2). The nodes can store the smallest value of -estimated (received in beacon) directly as their radial distance. The maximum error in radial estimation is given by (3)Furthermore, this RSSI estimation is susceptible to fading and attenuation, so tolerance of to may be modelled. The total number of beacons transmitted with an omni-DA is . Fig. 3Open in figure viewerPowerPoint Beacons from omni-DA of SBN The SBN uses one DA with beam-width () to transmit beacons containing the current direction of the DA. The beacon contains only one parameter . The direction of DA is changed with steps of and a new beacon is transmitted containing current direction . So, new is given by (4) Now, as shown in Fig. 4, any node n lying at an angle will receive all the beacons transmitted by DA in the range . Fig. 4Open in figure viewerPowerPoint Nodes receiving beacons from a rotating DA. Nodes in green colour are able to receive beacons because they lie within the beam-width of DA, a different set of nodes will receive beacons when the direction is changed The beacons received by any node can be represented by matrix (5), where and it accounts for deviation between and the direction of DA . The node can estimate its angle as (6) (5) (6)Maximum error in can be given as (7)The total number of beacons transmitted from DA is . Now, using (2) and (6), node location is estimated as (8) is the estimated location in polar coordinates and it may be converted to (x,y) as and . Furthermore, the SBN may broadcast its GPS coordinates and nodes can convert their relative locations into GPS locations as described in [16]. EE–LSB may be divided into two parts, functions of the SBN (or BS) and functions of any node as shown in Algorithms 1 and 2 (see Figs. 5 and 6). The flowchart of the process at the SBN and dumb nodes is shown in Figs. 7 and 8, respectively. If the SBN is not at the BS, then the BS transmits a start localisation message to the SBN. The BS stops transmitting until the SBN completes localisation and acknowledges it by broadcasting the end localisation message. Fig. 5Open in figure viewerPowerPoint Algorithm 1: functions of the SBN or BS Fig. 6Open in figure viewerPowerPoint Algorithm 2: functions of any dumb node Fig. 7Open in figure viewerPowerPoint Flowchart of the process at SBN (or BS) Fig. 8Open in figure viewerPowerPoint Flowchart of the process at dumb nodes 2.4 Format of the beacon signal from the SBN The SBN transmits three types of beacons, each of 17 bits, namely types 1–3, as shown in Fig. 9. Types 1 and 3 are transmitted using the omni-DA of the SBN, whereas type 2 is transmitted using the DA of the SBN. The first bit of the received beacon packet is used by nodes to determine the type of beacon. Fig. 9Open in figure viewerPowerPoint Format of the beacon. Types 1 and 3 are transmitted from the omni-DA, and type 2 is transmitted from DA The first bit of the beacon is '1' for types 1 and 3 and '0' for type 2 beacon. Type 1 message is used by nodes to determine their radial distance from the SBN, and type 2 beacon is used to determine their angular position. Type 3 is transmitted once after all the transmissions required for localisation have been completed from the SBN. It is transmitted with a maximum power to ensure the reception from all the nodes in the region. The end localisation message (type 3) is a type 1 message with 16 zeros. The nodes start the estimation of location after receiving a type 3 message. The beacon message is updated each time before transmission with current values of and . The SBN may broadcast its GPS coordinates so that nodes can convert their relative locations to GPS coordinates as in [16]. 2.5 Energy consumption of EE–LSB Energy consumption (E) of any node can be given as a sum of energy consumed by a node while transmitting (), receiving (), sample sensing date (), listening radio channel (), and sleeping (). These parameters depend on the number of beacons transmitted (), received (), and time duration for localisation (). Table 1 and 9-14 are derived from [17] for Mica2 mote and CC1000 transceiver Table 1. Energy consumption parameters [17] Parameter Notation Value sample rate, packets/s r 1/300 preamble length, byte 271 one byte duration, s , receive 1 byte current, mA 15 transmit 1 byte current, mA 20 sample sensing duration, s 1.1 sample sensing current, mA 20 voltage, V V 3 power consumed in listening radio for preamble, J 17.3 sample listening duration, s preamble sample interval, s sleep current, mA 0.030 (9) (10) (11) (12) (13) (14) Here, time duration of one beacon is , , where and are the time spend while receiving and transmitting beacons. assumed for simplicity assuming no sample is sensed during the localisation process (). So, the energy consumption for the localisation () can be expressed as a function of , , and (15) 2.5.1 Energy consumption at the SBN SBN needs to transmit directional beacons and radial beacons. So, the total number of beacons transmitted from the SBN can be given as (16) and represent the number of beacons transmitted from the DA and omni-DA to estimate angle and radius, respectively. The total energy consumed at the SBN () can be estimated by (17)Mechanical energy is required to change the direction of DA for 360 rotation. Assuming sequential transmission, is the time required to transmit beacons. Preamble sample interval () is the minimum time interval to transmit one beacon. So, total localisation time is given as (18) 2.5.2 Energy consumption at dumb node The energy consumption at the receiving dumb node can be calculated using (20) (19) (20) 2.6 Energy-efficiency of localisation The energy-efficiency of localisation () in the WSN can be defined as in [18] (21)Here represents the mean square localisation error (MSE). The energy consumed at a node in the localisation process is given as (22) 2.7 Computational complexity of EE–LSB The sensors in the WSN have limited computational power. However, increasing the number of computations adversely affects energy consumption. So, to select any localisation scheme, an analysis of computational complexity should be considered. In the proposed scheme, any node has to solve only 2-6, and (8) to localise itself. The scheme can be classified as complexity class . By using the uniform cost model [19], the worst-case computational complexity at the SBN is estimated as (23)The complexity at node side is (24)The computational complexity is linear and independent of the number of nodes in the network. 3 Simulation results and discussions An area of is considered for the WSN. The nodes are deployed randomly in the region. The BS is assumed to be located at the centre (it may be assumed at the corner, but the farthest node should be in the radio coverage range or the transmit power needs to be increased). The deviation in radial estimation due to RSSI losses is ignored for simplicity. The simulation parameters are shown in Table 2. Table 2. Simulation parameters and values Parameter Symbols Value area, A total nodes n 100 base-station location BS O (0,0) beam-width of antenna step starting position of ripple step, m 4 starting ripple radius, m max ripple radius, m packet length, byte 17 beacon transmission distance (DIR), m 10 beamwidth of DA (LBRD) height of DA (LBRD), m h 13 Fig. 10 shows the actual radial distance of the nodes from the SBN and their error in radial estimation. The errors lie within the range . Fig. 10Open in figure viewerPowerPoint Actual radial distance of the node and the error in their radial estimation Fig. 11 shows the actual angle of all the nodes and the error in angle estimation. It is observed that the angle error lies within the range of . Thus, by reducing the accuracy can be improved. Fig. 11Open in figure viewerPowerPoint Actual angle distance of the node and the error in their angle estimation Fig. 12 shows the line representing the distance and direction between the actual location of the node and its estimated location for all the nodes on the -plane. Fig. 12Open in figure viewerPowerPoint Error distance between the actual node location and the estimated location 3.1 Scalability of EE–LSB To test the scalability of the proposed scheme, the number of nodes is increased to 200 and 1000 as shown in Fig. 13. The deployment density is also varied (the number of nodes in each quadrant is changed) keeping the area fixed. The nodes are deployed sequentially starting from quadrant 1. Fig. 14 shows the localisation error with the proposed scheme for node deployment of Fig. 13. The maximum and mean error in both cases is and 2.1 m, respectively. It shows that changing the number of nodes or node density does not affect the accuracy of localisation. Fig. 13Open in figure viewerPowerPoint Variable density node deployment with , the number of nodes is different in each quadrant Fig. 14Open in figure viewerPowerPoint Error distance for node deployment of Fig. 13 Now the area of deployment is increased from to and keeping . To increase the area, the farthest node must lie within the communication range of the SBN. So, the communication range of the SBN is assumed as , 106.1, and 141.5 m, respectively. The range of error remains the same . However, the numerical value of increases as the radial distance from the SBN is increased. Hence, the localisation error is increased on increasing the area. As shown in Fig. 15, on increasing the area to four times from to the maximum error is doubled. However, the error can be reduced by reducing the step size from to . The accuracy can be improved by reducing the rotation step size . The figure shows that by reducing the step size to half (from to ) the error is reduced (max error ). Fig. 15Open in figure viewerPowerPoint Cumulative distribution function (CDF) of error distance on increasing the area of localisation and decreasing 3.2 Performance evaluation This section presents a comparison of the proposed scheme with RLA [14], DIR [15], and LBRD [16]. EE–LSB uses a mechanism similar to RLA for radial distance estimation and it uses a rotatable DA similar to LBRD for angle estimation. The parameters used to simulate RLA and EE–LSB are given in Table 2. The irregularities in radial (distance) estimation (due to RSSI losses) in both cases are ignored. Fig. 16 shows the error distance of each node, BNs in RLA show zero error. The maximum error in EE–LSB is and the average error is 2.8 m. Fig. 16Open in figure viewerPowerPoint Comparison of error distance Fig. 17 shows a comparison between the CDF of error distance of two schemes. It shows that the EE–LSB performs better than RLA and similar to LBRD. Fig. 17Open in figure viewerPowerPoint CDF of error distance Table 3 shows a comparison between various parameters obtained from the simulation. It shows that the accuracy of localisation is better in EE–LSB and the energy consumed for localisation is lower in the case of EE–LSB. Hence, EE–LSB performs better in terms of accuracy and energy consumption. Table 3. Performance evaluation of RLA, DIR, LBRD, and EE–LSB RLA DIR LBRD EE–LSB number of dumb nodes 90 100 100 100 number of BN 10 2 1 1 number of DA at each BN 0 4 1 1 number of omni-DA at each BN 1 0 0 1 total beacons transmitted by BN 200 80 360 56 localisation time , s 20 8 36 5.6 total beacons received by dumb nodes 8148 400 376 1379 average beacons received by dumb nodes 90.53 4 3.76 13.79 energy consumed in transmitting by all BN, J 0.51 0.004 0.019 0.003 energy consumed in receiving by all dumb nodes, J 0.68 0.218 0.954 0.17 total energy consumed by all nodes including BNs, J 1.19 0.222 0.972 0.172 maximum localisation error, m 51.31 0.68 10.33 4.71 average localisation error, m 23.53 0.68 2.53 1.93 MSE 767.47 0.47 10.55 4.67 energy-efficiency of localisation [18] 0.11 963.18 9.75 124.32 The location of each node in EE–LSB uses one RSSI estimated distance similar to RLA. However, in EE–LSB, accuracy is better than RLA because the locations of the nodes depend on three or more RSSI estimated distances from different anchor nodes. Thus, the RSSI error accumulates in RLA. The EE–LSB requires single DA at the SBN as compared to four DAs used by each BN of DIR. Furthermore, one 360 rotation of the DA is required in the EE–LSB. In DIR, BNs need to move across the region. A DA is required in both LBRD and EE–LSB. One axis rotation of the DA is required in the EE–LSB. However, the two-axis rotation of the DA is required in LBRD. Furthermore, the required height of the DA is very high in the case of LBRD. The simulation results show that the proposed scheme performs similar to LBRD, which is better than RLA in terms of accuracy. However, the energy consumption at the node is reduced, and energy consumption at the SBN (or BS) is reduced significantly. The hardware requirement at the SBN is simple in comparison with LBRD. Furthermore, no specific height of the antenna is required in the EE–LSB as compared to LBRD. 4 Conclusion The proposed localisation scheme is capable of localising the nodes using an SBN. The total number of beacons received by dumb nodes is small, and transmission is not required from them. The nodes act in passive listening mode for localisation. The total number of beacons transmitted from the BN (SBN) is 56. Thus, energy consumption during localisation is reduced at the node side as well as at the BN (or BS). Furthermore, the total time required for the localisation of all the nodes in the region (communication range of BN) is 5.6 s. The maximum localisation error is , the average localisation error is 1.93 m, and the total energy consumption is . Hence, the proposed localisation scheme (EE–LSB) performs better in terms of accuracy and energy consumption. Furthermore, it has low-computational complexity and requires only one BN. In the future, multiple SBNs can be deployed at different locations or movable SBNs may be used to improve accuracy. The scheme may be explored to localise moving node. References 1Wang X.: 'Multicast for 6LowPAN wireless sensor networks', IEEE Sens. J., 2015, 15, (5), pp. 3076– 3083 2El-Qawasma F.A., Elfouly T.M., Ahmed M.H.: 'Minimising number of sensors in wireless sensor networks for structure health monitoring systems', IET Wirel. Sens. Syst., 2019, 9, (2), pp. 94– 101 3Khelifi F., Bradai A., Kaddachi M.L. et al.: 'Design and experimental implementation of monitoring system in wireless sensor networks', IET Wirel. Sens. Syst., 2018, 8, (6), pp. 350– 359 4Yan J., Zhou M., Ding Z.: 'Recent advances in energy-efficient routing protocols for wireless sensor networks: a review', IEEE Access, 2016, 4, pp. 5673– 5686 5Rashid A., Tripathi Y., Prakash A. et al.: 'Load aware energy-balanced data gathering approach in CRSNs', IET Wirel. Sens. Syst., 2019, 9, (3), pp. 143– 150 6Pease S.G., Conway P.P., West A.A.: 'Hybrid ToF and RSSI real-time semantic tracking with an adaptive industrial internet of things architecture', J. Netw. Comput. Appl., 2017, 99, pp. 98– 109 7Chowdhury T.J.S., Elkin C., Devabhaktuni V. et al.: 'Advances on localization techniques for wireless sensor networks: a survey', Comput. Netw., 2016, 110, pp. 284– 305 8Chelouah L., Semchedine F., Bouallouche-Medjkoune L.: 'Localization protocols for mobile wireless sensor networks: a survey', Comput. Electr. Eng., 2018, 71, pp. 733– 751 9Han G., Jiang J., Zhang C. et al.: 'A survey on mobile anchor node assisted localization in wireless sensor networks', IEEE Commun. Surv. Tutor., 2016, 18, (3), pp. 2220– 2243 10Gui L., Yang M., Fang P. et al.: 'Rss-based indoor localisation using MDCF', IET Wirel. Sens. Syst., 2017, 7, (4), pp. 98– 104 11Pandey O.J., Hegde R.M.: 'Cooperative localisation over small world WSN using optimal allocation of heterogeneous nodes', IET Wirel. Sens. Syst., 2018, 8, (4), pp. 162– 169 12Wu M., Xiong N., Tan L.: 'Adaptive range-based target localization using diffusion gauss newton method in industrial environments', IEEE Trans. Ind. Inf., 2019, 15, (11), pp. 5919– 5930 13Shit R.C., Sharma S., Puthal D. et al.: 'Location of things (LoT): a review and taxonomy of sensors localization in IoT infrastructure', IEEE Commun. Surv. Tutor., 2018, 20, (3), pp. 2028– 2061 14Farooq-I-Azam M., Ni Q., Ansari E.A.: 'Intelligent energy efficient localization using variable range beacons in industrial wireless sensor networks', IEEE Trans. Ind. Inf., 2016, 12, (6), pp. 2206– 2216 15Ou C.: 'A localization scheme for wireless sensor networks using mobile anchors with directional antennas', IEEE Sens. J., 2011, 11, (7), pp. 1607– 1616 16Gautam P.R., Kumar S., Verma A. et al.: 'Energy-efficient localization of sensor nodes in WSNs using beacons from rotating directional antenna', IEEE Trans. Ind. Inf., 2019, 15, (11), pp. 5827– 5836 17Polastre J., Hill J., Culler D.: ' Versatile low power media access for wireless sensor networks categories and subject descriptors'. Proc. of the 2nd Int. Conf. on Embedded Networked Sensor Systems, Baltimore, MD, USA, 2004, pp. 95– 107 18Zheng K., Wang H., Li H. et al.: 'Energy-efficient localization and tracking of mobile devices in wireless sensor networks', IEEE Trans. Veh. Technol., 2017, 66, (3), pp. 2714– 2726 19Ausiello G., Crescenzi P., Gambosi G. et al.: ' Complexity and approximation: combinatorial optimization problems and their approximability properties' ( Springer Science & Business Media, Germany, 2012) Citing Literature Volume14, Issue9June 2020Pages 1459-1466 FiguresReferencesRelatedInformation
Referência(s)