energies Article
Efficient Wireless Charger Deployment for Wireless Rechargeable Sensor Networks Jehn-Ruey Jiang * and Ji-Hau Liao Department of Computer Science and Information, National Central University, Taoyuan 32001, Taiwan;
[email protected] * Correspondence:
[email protected]; Tel.: +886-3-422-7151 Academic Editor: Chang Wu Yu Received: 11 June 2016; Accepted: 22 August 2016; Published: 31 August 2016
Abstract: A wireless rechargeable sensor network (WRSN) consists of sensor nodes that can harvest energy emitted from wireless chargers for refilling their batteries so that the WRSN can operate sustainably. This paper assumes wireless chargers are equipped with directional antennas, and are deployed on grid points of a fixed height to propose two heuristic algorithms solving the following wireless charger deployment optimization (WCDO) problem: how to deploy as few as possible chargers to make the WRSN sustainable. Both algorithms model the charging space of chargers as a cone and calculate charging efficiency according power regression expressions complying with the Friis transmission equation. The two algorithms are the greedy cone covering (GCC) algorithm and the adaptive cone covering (ACC) algorithm. The GCC (respectively, ACC) algorithm greedily (respectively, adaptively) generates candidate cones to cover as many as possible sensor nodes. Both algorithms then greedily select the fewest number of candidate cones, each of which corresponds to the deployment of a charger, to have approximate solutions to the WCDO problem. We perform experiments, conduct simulations and do analyses for the algorithms to compare them in terms of the time complexity, the number of chargers deployed, and the execution time. Keywords: wireless rechargeable sensor network (WRSN); energy harvesting; charger deployment; sustainability; greedy algorithms; directional antennas; Friis transmission equation
1. Introduction A wireless rechargeable sensor network (WRSN) [1] consists of many rechargeable sensor nodes and one or more sink nodes. Sensor nodes are powered by batteries and can sense physical phenomena (e.g., temperature, humidity, atmospheric pressure, light intensity, and electromagnetic field strength) and forward the sensed data to the sink node(s) via multi-hop wireless communications. WRSNs are sustainable since they utilize energy harvesting technologies to replenish the batteries of sensor nodes. The replenishing is achieved by equipping every sensor node with an energy harvester to harvest energy from different sources [2,3], such as solar power, thermal power, radio frequency (RF) waves, or air flow, as shown in Figure 1. Many recent studies [4–13] focus on the WRSNs due to their sustainability. Some other studies [14–20] investigate battery-less or battery-free wireless sensor networks (WSNs) in which sensor nodes are powered by capacitors storing energy harvested through wireless power transmission (WPT), especially microwave power transmission (MPT). Without loss of generality, such WSNs are regarded as WRSNs in this paper. Energy harvesting technologies are generally divided into two classes: (1) charger-based energy harvesting and (2) ambient energy harvesting. The former technology deploys specific devices, called chargers, to emit energy, and utilizes harvesters to capture the energy. The latter technology applies ambient energy harvesting devices to capture energy from the ambient environment. Typical devices include solar panels, thermoelectric generators, and ambient RF harvesters to capture energies from TV, Energies 2016, 9, 696; doi:10.3390/en9090696
www.mdpi.com/journal/energies
Energies 2016, 9, 696
2 of 22
Wi-Fi, and 3G/4G stations. It is more difficult for the latter to control the amount of harvested energy Energies 2016, 9, 696 2 of 21 due to the unstable nature of ambient energy sources (ESs). This is problematic if sensor nodes have to TV, Wi‐Fi, and 3G/4G stations. It is more difficult for the latter to control the amount of harvested meet specificenergy due to the unstable nature of ambient energy sources (ESs). This is problematic if sensor nodes energy requirements. Therefore, this paper pays attention to WRSNs using charger-based have to technologies. meet specific energy requirements. Therefore, paper pays attention to WRSNs energy harvesting It practically focuses onthis indoor WRSNs using the RFusing wireless energy charger‐based energy harvesting technologies. It practically focuses on indoor WRSNs using the RF harvesting technology. wireless energy harvesting technology.
Figure 1. A wireless rechargeable sensor network (WRSN) sensor node equipped with an energy wireless rechargeable sensor network (WRSN) sensor node equipped with harvester to harvest energy. RF: radio frequency.
Figure 1. A harvester to harvest energy. RF: radio frequency.
an energy
The deployment of “RF wireless chargers” (or “chargers” for short) is challenging, since every harvester needs to be covered by at least one charger and the charging space of chargers is usually limited. For of example, the effective charging distance is 3–5 m and for the short) effective ischarging angle is since every The deployment “RF wireless chargers” (or “chargers” challenging, 30°–45° between the Powercast TX91501‐3W‐ID charger (Powercast Corp., Pittsburgh, PA, USA) [21] harvester needs to be covered by at least one charger and the charging space of chargers is usually and the Powercast P2110‐EVAL‐02 harvester (Powercast Corp.) [22]. It is also a cost‐consuming task, limited. Foras example, the effective charging distance is 3–5 TX91501‐3W‐ID m and the effective charging angle is wireless chargers are expensive. For example, the Powercast charger currently costs about 1,000 US dollars. This motivated the authors to study the wireless charger deployment 30◦ –45◦ between the Powercast TX91501-3W-ID charger (Powercast Corp., Pittsburgh, PA, USA) [21] optimization (WCDO) problem defined below. Given a set of WRSN sensor nodes equipped with and the Powercast P2110-EVAL-02 harvester (Powercast Corp.) [22]. It is also a cost-consuming task, harvesters, the WCDO addresses how to deploy as few as possible chargers to cover all sensor nodes as wireless chargers aretheir expensive. For example, thethe Powercast TX91501-3W-ID currently for satisfying energy requirements to make WRSN sustainable. To the best charger of our knowledge, no research has addressed the same problem up to now. costs about 1,000 US dollars. This motivated the authors to study the wireless charger deployment This paper proposes two algorithms for solving the WCDO problem. It is assumed that wireless optimization (WCDO) problem defined below. Given a set of WRSN sensor nodes equipped with chargers are equipped with directional antennas whose charging space is modeled as a cone and harvesters, the WCDO addresses how to deploy as few as possible chargers to cover all sensor nodes charging efficiency complies with the Friis free space transmission equation [23]. It is also assumed that chargers are deployed on grid points at a specific height and the sensor nodes equipped with for satisfying their energy requirements to make the WRSN sustainable. To the best of our knowledge, wireless harvesters are distributed beneath the chargers on the floor or object surfaces. The proposed no research has addressed the same problem up to now. algorithms are the greedy cone covering (GCC) algorithm and the adaptive cone covering (ACC) This paper proposes two algorithms for solving the WCDO problem. It is assumed that wireless algorithm. Both algorithms treat chargers as cones. The GCC (respectively, ACC) algorithm greedily chargers are (respectively, adaptively) generates candidate cones that can cover as many as possible sensor nodes. equipped with directional antennas whose charging space is modeled as a cone and The two algorithms then greedily select the fewest number of candidate cones, each of which charging efficiency complies with the Friis free space transmission equation [23]. It is also assumed corresponds to the deployment of a charger, to derive approximate solutions to the WCDO problem. that chargersWe are deployed on experiments grid points at a specific height We andalso theconduct sensor nodes equipped with perform extensive to justify our assumptions. comprehensive simulations and do analyses for the algorithms to compare them in terms of the time complexity, the wireless harvesters are distributed beneath the chargers on the floor or object surfaces. The proposed number of chargers deployed, and the execution time. algorithms are the greedy cone covering (GCC) algorithm and the adaptive cone covering (ACC) The remainder of the paper is organized as follows: some related studies are reviewed in Section algorithm. Both algorithms treat chargersand as the cones. Thedefinition. GCC (respectively, ACC) algorithm greedily 2. Section 3 presents the modeling problem In Section 4, the two proposed (respectively,algorithms to solve the problem are described. Section 5 shows algorithm time complexity analyses adaptively) generates candidate cones that can cover as many as possible sensor and grid point separation analyses for deploying chargers. Experiments about the charging efficiency nodes. The two algorithms then greedily select the fewest number of candidate cones, each of which and results of algorithm simulations are described in Section 6. Finally, Section 7 concludes the paper.
corresponds to the deployment of a charger, to derive approximate solutions to the WCDO problem. We perform 2. Related Work extensive experiments to justify our assumptions. We also conduct comprehensive simulations and In this section, we review the research results [3–8,13–17] most related to our work. Kim et al. [3] do analyses for the algorithms to compare them in terms of the time complexity, the investigated various ambient energy‐harvesting technologies using different power sources, such as number of chargers deployed, and the execution time. The remainder of the paper is organized as follows: some related studies are reviewed in Section 2. Section 3 presents the modeling and the problem definition. In Section 4, the two proposed algorithms to solve the problem are described. Section 5 shows algorithm time complexity analyses and grid point separation analyses for deploying chargers. Experiments about the charging efficiency and results of algorithm simulations are described in Section 6. Finally, Section 7 concludes the paper. 2. Related Work In this section, we review the research results [3–8,13–17] most related to our work. Kim et al. [3] investigated various ambient energy-harvesting technologies using different power sources, such
Energies 2016, 9, 696
3 of 22
as solar power, thermal power, piezoelectric power, and wireless RF power, in the aspect of their applicability for self-sustainable wireless sensor platforms. They stated that although ambient RF energy has low power density values of 0.2 nW/cm2 –1 µW/cm2 and low radio frequency-direct current (RF-DC) conversion values of 10%–30%, a larger amount of total available power can be harvested by utilizing high gain antennas. Some antennas designed by the authors were shown [3]. For example, a folded dual-bad antenna at 915 MHz and 2.45 GHz was designed to harvest RF power from cellular and WiFi sources with power densities about 1 µW/cm2 . Moreover, a folded dipole antenna was designed to harvest RF power from a two-way radio at the UHF band (464 and 468 MHz). They also implemented a prototype of an embedded microcontroller-enabled sensor platform successfully powered by an ambient UHF digital TV signal (512–566 MHz) from a broadcasting antenna that is 6.3 km away. Fu et al. [4] proposed a moving and stopping plan for mobile chargers to guarantee that the time to charge all given stationary sensor nodes is minimized. An optimal solution using linear programming for the plan is proposed; however, the computational complexity of the solution is very high. Therefore, a heuristic solution discretizing the charging power on a two-dimension space was proposed to reduce the time complexity. The heuristic solution has been proved to have a bounded approximation ratio. He et al. [5] proposed two types of energy provisioning problems for WRSNs: point provisioning and path provisioning. The former addresses how to use the least number of chargers to ensure that a static sensor node placed in any position of the network will receive a sufficient recharge rate to ensure WRSN sustainability. The latter addresses the same problem for mobile sensor nodes under the assumption that sensor nodes can harvest excess energy in power-rich areas and store the energy for later use in power-deficient areas. The authors utilized the triangular deployment of chargers to solve the two problems. They also proposed a modified Friis free space transmission equation adjusted for short distance transmission for evaluating the deployment. The original Friis equation and the modified version are shown below in Equations (1) and (2), respectively: Pr = Gt Gr
Pr =
Gt Gr η Lp
λ 4πD
2 Pt
λ 4π ( D + β)
(1)
2 Pt
(2)
In Equations (1) and (2), Pt and Gt denote the power and the antenna gain of the transmitter (i.e., charger), respectively, and Pr and Gr denote the power and the antenna gain of the receiver (i.e., harvester), respectively. Furthermore, λ is the wave length, and D is the distance between the antennas of the charger and the harvester. η is the rectifier efficiency, Lp is polarization loss, and β is a parameter to adjust the original Friis equation for short distance transmission. Dai et al. [6] stated that if the constraints of sensor node speed and battery capacity are considered, the sustainable operation of sensor nodes may not be guaranteed. It proposed the concept of quality of energy provisioning (QoEP) to characterize the expected portion of time that a sensor node can sustain operation by taking into account node speed and battery capacity. The upper and lower bounds of QoEP in 1D and 2D sensor node mobility cases are derived. Xu et al. [7] considered the scheduling problem of arranging multiple mobile chargers to collaboratively charge sensor nodes with the goal of minimizing the total traveling distance of all chargers. This can be formulated as finding multiple travelling salesperson problem (TSP) closed tours covering all sensor nodes such that the total length of the tours is minimized. The authors proposed a 2-approximation algorithm to solve the problem under the assumption that the energy consumption rates of all sensor nodes are fixed. Liang et al. [8] considered a scheduling problem similar to that addresses by Xu et al. in [6]. The problem is how to use the minimum number of mobile chargers to charge sensor nodes so that none of sensor nodes runs out of energy, subject to the energy capacity of mobile chargers. The authors designed a 5-approximation algorithm to find multiple closed tours. Afterwards, by bounding the
Energies 2016, 9, 696
4 of 22
total distance of each tour, the authors designed another approximation algorithm invoking the 5-approximation algorithm to provide approximate solutions to the problem. Ye et al. [9] introduced a new metric, the charging utility, to measure the charging quality of the charger with the considerations of both the fairness of charging sensor nodes and the number of sensor nodes charged by mobile chargers traveling along closed tours. It is assumed that each mobile charger can only travel a limited length per tour, due to its limited energy capacity. The authors formulated an optimization problem of scheduling a mobile charger to charge sensor nodes, with an objective of maximizing the charging utility, subject to the total traveling distance of the mobile charger per tour and the deadline of each sensor node to be charged. An approximation algorithm with a guaranteed approximation ratio was proposed to solve the problem. Maehara et al. [14] proposed a WPT scheme called multi-point carrier shift diversity (MPCSD), in which the center frequency of the carrier of each power transmitting point (or transmitter) is slightly shifted with a predefined amount to create artificial fading. Suppose there are N different transmitters, the carrier frequency f n of the nth transmitter is defined in Equation (3), shown below: fn = fc −
B (2n − 1) B + 2 2N
(3)
In Equation (3), f c is the center frequency, B is the channel bandwidth, and n = 1, . . . , N. The proposed scheme is shown to be capable of mitigating power attenuation caused by the path loss effect and the standing-wave effect created by multipath and interference between multiple power transmitters so that the coverage of energy supply field is improved. Maehara et al. [15] investigated the energy transmission coverage of the MPCSD scheme through theoretical discussions and practical experiments. It is shown that the MPCSD scheme can achieve 100% of coverage in an indoor environment for the sensor node duty factor of 1/100 (i.e., the sensor node is in the transmitting mode for the duration of 10 ms per second). Imoto et al. [16] conducted experiments in an IEEE 802.11g-based wireless local area network (WLAN) in which an ES transmits microwave power to a sensor station (SS) using the carrier sense multiple access with collision avoidance (CSMA/CA) protocol to transmit data to an access point (AP). They derived the following two requirements for coexistence of MPT and the WLAN wireless data transmission (WDT). First, MPT and WDT should not proceed simultaneously. Therefore, the ES should share the timing information of MPT with the SS so that the SS attempts to transmit data frames only when the ES does not transmit microwave power. Second, the ES should not transmit microwave power to the SS when the AP broadcasts beacon frames so that the SS is not disassociated from the AP. This can also prevent data throughput degradation due to WLAN automatic data rate selection. Yamashita et al. [17] assumed the same environment as assumed by Imoto et al. [16] and designed and implemented a scheduling scheme to solve the disassociation problem and the throughput degradation problem caused by the coexistence of MPT and WDT. In the scheduling scheme, the SS goes to the active mode to transmit data towards and to receive delivery traffic indication map (DTIM) beacons from the AP periodically. It goes to the power-saving (sleep) mode for most of the time to save energy. As long as the SS is in the sleep mode, the ES can transmit microwave power. On the basis of link budget analysis, the authors also clarified the maximum distance between the EC and the SS to conclude that the scheduling scheme can be applied to a closed space or room. Yoshida et al. [18] designed an aerospace WSN that allow the coexistence of high-power MPT and WDT. The aerospace WSN is used in the aircraft application of the reusable vehicle test (RVT) or space capsule for health and safety monitoring inside the RVT. The WSN is based on the time-division operation and the frequency-division operation. The former operation uses a control function to schedule MPT and WDT so that MPT is not activated when WDT is activated. The latter operation applies different frequency bands for WPT and WDT to avoid the interference between them. Through a practical experiment lasting 120 min, it is shown that the aerospace WSN can achieve sustainable balance between harvested power and consumed power.
Energies 2016, 9, 696
5 of 22
Energies 2016, 9, 696 3. Modeling and Problem Definition
5 of 21
3. Modeling and Problem Definition Sensor nodes in the WRSN are assumed to be distributed in a cuboid with the length L, width Energies 2016, 9, 696 5 of 21 W and height Sensor nodes in the WRSN are assumed to be distributed in a cuboid with the length L, width H; they have different energy requirements and can be located on the floor plane or W and height H; they have different energy requirements and can be located on the floor plane or object surfaces. The energy requirements and locations are summed to be known in advance and fixed. 3. Modeling and Problem Definition object surfaces. The energy requirements and locations are summed to be known in advance and Wireless chargers equipped with directional antennas are assumed to be deployed at grid points on the Sensor nodes in the WRSN are assumed to be distributed in a cuboid with the length L, width fixed. Wireless chargers equipped with directional antennas are assumed to be deployed at grid grid plane or the deployment plane that is with height H and parallel to the floor plane. Moreover, the points on the grid plane or the deployment plane that is with height H and parallel to the floor plane. W and height H; they have different energy requirements and can be located on the floor plane or grid points aresurfaces. with the separation S; that is, the grid areto of distance S. Please Moreover, the grid points are with the separation S; that is, the nearest grid points are of the distance object The energy requirements and nearest locations are points summed be the known in advance and refer to S. Please refer to Figure 2 for the illustration of the distribution of sensor nodes and the deployment Wireless chargers equipped with directional are assumed to be deployed at grid Figure 2fixed. for the illustration of the distribution of sensorantennas nodes and the deployment of wireless chargers. of wireless chargers. Note that we assume no more than Q, Q ≥ 1, chargers can be deployed at a same points on the grid plane or the deployment plane that is with height H and parallel to the floor plane. Note that we assume no more than Q, Q ≥ 1, chargers can be deployed at a same grid point. grid point. Moreover, the grid points are with the separation S; that is, the nearest grid points are of the distance S. Please refer to Figure 2 for the illustration of the distribution of sensor nodes and the deployment of wireless chargers. Note that we assume no more than Q, Q ≥ 1, chargers can be deployed at a same grid point.
Figure 2. The schematic view of the WRSN deployment cuboid.
Figure 2. The schematic view of the WRSN deployment cuboid. The effective charging space of wireless chargers is modeled as a cone, called a charger cone. Note that the two terms “charger” and “cone” are used interchangeably later in this paper. The The effective charging space of wireless chargers is modeled as a cone, called a charger modeling is based on the fact that direction antennas usually have specific beamwidth due to vertical Figure 2. The schematic view of the WRSN deployment cuboid. cone. Note that the two terms “charger” and usedof interchangeably later in this paper. and horizontal polarization. For example, the “cone” direction are antenna the Powercast TX91501‐3W‐ID charger has 60° beamwidth in both the horizontal direction and the vertical direction. As shown in The modeling is based on the fact that direction antennas usually have specific beamwidth due The effective charging space of wireless chargers is modeled as a cone, called a charger cone. whose direction is Figure 3, every charger cone is characterized by an apex X, a normal vector to vertical polarization. For example, the direction later antenna the The Powercast Note and that horizontal the two terms “charger” and “cone” are used interchangeably in this ofpaper. parallel to the symmetrical axis of the cone, an effective charging distance (i.e., the cone generatrix ◦ TX91501-3W-ID charger has 60 beamwidth in both the horizontal direction and the vertical direction. modeling is based on the fact that direction antennas usually have specific beamwidth due to vertical length) D, and an effective charging angle (i.e., the opening angle) A. When a sensor node is within * and For example, direction antenna of apex the Powercast TX91501‐3W‐ID the charger cone of a charger, we assume the sensor node can be charged effectively by the charger; As shown inhorizontal Figure 3,polarization. every charger cone is the characterized by an X, a normal vector N whose charger has 60° beamwidth in both the horizontal direction and the vertical direction. As shown in node cannot be charged effectively. that the charging point E is distance regarded as an the cone direction isotherwise, parallelthe to sensor the symmetrical axis of the cone, anNote effective (i.e., whose direction is Figure 3, every charger cone is characterized by an apex X, a normal vector extreme point within the charger cone, where the distance between E and X is D and the angle generatrix length) D, and an effective charging angle (i.e., the opening angle) A. When a sensor node parallel to the symmetrical axis of the cone, an effective charging distance (i.e., the cone generatrix between the normal vector and the vector going from X to E is A/2.
is withinlength) D, and an effective charging angle (i.e., the opening angle) A. When a sensor node is within the charger cone of a charger, we assume the sensor node can be charged effectively by the charger;the charger cone of a charger, we assume the sensor node can be charged effectively by the charger; otherwise, the sensor node cannot be charged effectively. Note that the point E is regarded otherwise, the within sensor node cannot be charged effectively. Note that the point E is X regarded as an as an extreme point the charger cone, where the distance between E and is D and the angle * charger cone, where the distance between E and X is D and the angle extreme point within the between the normal vector and the vector going from X to E is A/2.
between the normal vector N and the vector going from X to E is A/2. A
A
Figure 3. A charger cone and its parameters.
Figure 3. A charger cone and its parameters.
Figure 3. A charger cone and its parameters.
Energies 2016, 9, 696
6 of 22
The famous Friis transmission equation [23], shown in Equation (1), is used to model the energy received by the harvester. By the equation, we can see that the received energy Pr is inversely proportional to D2 , that is, Pr = αD−2 , where α is a constant and D is the distance between the charger and the harvester. This paper utilizes a more general empirical model Pr = αD −β , where β > 0, to estimate the energy received by the harvester. Later in Section 5, we will show how to apply the power regression analysis to derive appropriate α and β values for the model to fit experimental data accurately. Energies 2016, 9, 696 With all the assumptions mentioned above, the WCDO problem deals with how 6 of 21 to deploy the minimum number of wireless chargers to cover all sensor nodes and fulfil their energy demands. The famous Friis transmission equation [23], shown in Equation (1), is used to model the energy To be morereceived precise, a set ofBy sensor nodeswe distributed inthe a cuboid a set is ofinversely grid points, the by given the harvester. the equation, can see that received and energy 2, that is, WCDO problem addresses how to deploy as few as possible chargers at grid points to charge sensor proportional to D α , where α is a constant and D is the distance between the charger and the harvester. This paper utilizes more general that empirical model α , where β > of 0, to nodes to make them sustainable under the a assumption the effective charging space chargers is estimate the energy received by the harvester. Later in Section 5, we will show how to apply the a cone. A special case of the WCDO problem in which every charger’s charging space is a half sphere power regression analysis to derive appropriate α and β values for the model to fit experimental (instead of data accurately. a cone) and every sensor node needs to be covered by only one charger can be solved by the reduction toWith all the assumptions mentioned above, the WCDO problem deals with how to deploy the the NP-hard set covering (SC) problem [24]. It is believed the WCDO problem is also minimum number of wireless chargers to cover all sensor nodes and fulfil their energy demands. To NP-hard. Nevertheless, the NP-hardness of the WCDO problem has not yet proven. be more precise, given a set of sensor nodes distributed in a cuboid and a set of grid points, the WCDO problem addresses how to deploy as few as possible chargers at grid points to charge sensor nodes 4. Proposed Algorithms to make them sustainable under the assumption that the effective charging space of chargers is a cone. A special case of the WCDO problem in which every charger’s charging space is a half sphere Two algorithms, namely the GCC algorithm and the ACC algorithm, are proposed to solve the (instead of a cone) and every sensor node needs to be covered by only one charger can be solved by WCDO problem efficiently. The algorithms have different time complexity and achieve different the reduction to the NP‐hard set covering (SC) problem [24]. It is believed the WCDO problem is also approximate solutions to the WCDO problem. The details of the two algorithms are described below. NP‐hard. Nevertheless, the NP‐hardness of the WCDO problem has not yet proven.
For a WSRN with n sensor nodes and with chargers being deployed on some of m grid points, 4. Proposed Algorithms the set of sensor nodes is denoted by NS j k= {s1 , s2 ,j . . .k, sn } and the set of grid points is denoted by
GS = { g1 , g2 , Two algorithms, namely the GCC algorithm and the ACC algorithm, are proposed to solve the . . . , gm }, where m = ( SL + 1) × ( W S + 1), S is the separation of grid points, and WCDO problem efficiently. The algorithms have different time complexity and achieve different L, W and Happroximate solutions to the WCDO problem. The details of the two algorithms are described below. are the length, width, and height of the sensor node deployment cuboid, respectively. Note that we assume each sensor node si in NS = {s1 , s2 , . . . , sn } has energy requirement ei in For a WSRN with n sensor nodes and with chargers being deployed on some of m grid points, the set of sensor nodes is denoted by of energy , , …requirements: , and the set coverage of grid points is denoted by ES = {e1 , e2 , . . . , en }. There are two forms requirement and power , , … , , where 1 1 , S is the separation of grid points, and L, W and requirement. For the coverage requirement, ei is a positive integer 1, 2, . . . , indicating sensor node si needs theH are the length, width, and height of the sensor node deployment cuboid, respectively. Note that coverage of ei chargers to fulfill its energy demand. For the power requirement, ei is of we assume each sensor node in , ,…, has energy requirement in the unit mW,, indicating si needs to capture ei mW requirements: of power to coverage make itself sustainable. The coverage . There are two forms of energy requirement and power ,…, requirement is actually about extreme cases. Specifically, a node’s coverage requirement can be requirement. For the coverage requirement, is a positive integer 1, 2,…, indicating sensor node derived needs the coverage of chargers to fulfill its energy demand. For the power requirement, is of by calculating the ratio of the node’s power requirement over the power harvested by the extreme unit mW, point E, as the indicated in indicating Figure 3. needs to capture mW of power to make itself sustainable. The coverage requirement is actually about extreme cases. Specifically, a node’s coverage requirement Both algorithms have two phases. The first one is the cone generation phase to generate candidate can be derived by calculating the ratio of the node’s power requirement over the power harvested by cones to bethe extreme point E, as indicated in Figure 3. selected; the second one is the cone selection phase to select some candidate cones for Both algorithms have two phases. The first one is the cone generation phase to generate candidate deploying chargers. In the first phase, the algorithms generate cones by taking every grid point as the cones to be selected; the second one is the cone selection phase to select some candidate cones for apex of cones. As shown in Figure 4, for every grid point, the algorithms first generate a half sphere deploying chargers. In the first phase, the algorithms generate cones by taking every grid point as (HS) centered at the point and calculate the sensor set (SS) of sensor nodes covered by HS (i.e., within the apex of cones. As shown in Figure 4, for every grid point, the algorithms first generate a half HS). Afterwards, the first algorithm generates candidate cones on a node-pair-by-node-pair basis (i.e., sphere (HS) centered at the point and calculate the sensor set (SS) of sensor nodes covered by HS (i.e., within HS). Afterwards, the first algorithm generates candidate cones on a node‐pair‐by‐node‐pair for every pair of nodes in SS), while the second algorithm, on a node-by-node basis (i.e., for every basis (i.e., for every pair of nodes in SS), while the second algorithm, on a node‐by‐node basis (i.e., node in SS). For generating better candidate cones, the algorithms greedily adjust cones so that every for every node in SS). For generating better candidate cones, the algorithms greedily adjust cones so single conethat every single cone can cover as many as sensor nodes or can charge sensor nodes more efficiently. can cover as many as sensor nodes or can charge sensor nodes more efficiently.
Figure 4. The illustration of the lower half sphere HS centered at with radius D.
Figure 4. The illustration of the lower half sphere HS centered at g with radius D.
Energies 2016, 9, 696
7 of 22
Energies 2016, 9, 696 In the second phase, both algorithms first unmark all sensor nodes. Note that a sensor 7 of 21 node is unmarked if its energy requirement is not satisfied; otherwise, it will be marked. The algorithms then In the second phase, both algorithms first unmark all sensor nodes. Note that a sensor node is run iteration by iteration, and greedily select the qualified cone covering most unmarked sensor nodes, unmarked if its energy requirement is not satisfied; otherwise, it will be marked. The algorithms then where a cone is said to be qualified if it takes a grid point g as its apex and g is totally taken by no more run iteration by iteration, and greedily select the qualified cone covering most unmarked sensor than Q selected cones as their apex. A selected cone corresponds to a deployed charger. For every nodes, where a cone is said to be qualified if it takes a grid point as its apex and is totally taken newly selected cone, the algorithms marks every sensor node whose energy requirement is satisfied by no more than Q selected cones as their apex. A selected cone corresponds to a deployed charger. by the cone. And the algorithms continue until all sensor nodes are marked. In this way, the WCDO For every newly selected cone, the algorithms marks every sensor node whose energy requirement problem can be solved near-optimally. is satisfied by the cone. And the algorithms continue until all sensor nodes are marked. In this way, * The pseudo code of the GCC algorithm is shown in Figure 5. In the code, u x stands for a vector the WCDO problem can be solved near‐optimally. * going from grid point g to sensor node s ax , and Cone(u x ) stands for a cone which takes g as the The pseudo code of the GCC algorithm is shown in Figure 5. In the code, stands for a vector * apex, takes u x as the symmetrical axis (or the , and Cone( direction), and) stands for a cone which takes has an effective charging distance D and going from grid point to sensor node as the * an apex, takes effective charging angle A. Also note that Cone u x stands for the number of unmarked sensor as the symmetrical axis (or the direction), and has an effective charging distance D and * an effective charging angle A. Also note that | * * | stands for the number of unmarked sensor nodes in NS covered by Cone (u x ). Moreover, ψ u x , uy stands for the angle between the two vectors nodes in NS covered by Cone ( ). Moreover, ψ , stands for the angle between the two vectors * * u x and uy . . and
Figure 5. The greedy cone covering (GCC) algorithm. Figure 5. The greedy cone covering (GCC) algorithm.
In the GCC algorithm, for every grid point , a lower half sphere HS is obtained with as the In the GCC algorithm, for every grid point g, a lower half sphere HS is obtained with g as the center and D as the radius. If HS covers only one sensor node , then the algorithm generates one center and D as the radius. If HS covers only one node s a1 , then the algorithm generates one candidate cone taking the vector going from to sensor as the direction. However, if HS covers k, k > candidate cone taking the vector going from g to s as the direction. However, covers k, k > 1, a1 1, sensor nodes , , … , , then the algorithm generates k unit vectors , , …if, HS going from * * * sensor a2 , . . . , s ak , then the algorithm generates k unit vectors u1 , u2 , . . . , uk going from g to , , …s a, 1 , s. Afterwards, the GCC algorithm compares the opening angle A with the angle B (i.e., to nodes s a1 ψ , s a2 , . . . ), between any two vectors s ak . Afterwards, the GCC algorithm compares the opening angle A with the angle and B (i.e., and associated with the two distinct sensor nodes . There are three cases for the comparison: B = A, B > A, and B A, and B < A. The GCC algorithm performs the aforementioned angle comparison by projecting every vector The GCC algorithm performs the aforementioned angle comparison by projecting every vector The GCC algorithm performs the aforementioned angle comparison. As shown in Figure 6, the by projecting every vector and every cone onto the surface of a unit sphere centered at grid point and every cone onto the surface of a unit sphere centered at grid point and every cone onto the surface of a unit sphere centered at grid point g.. As shown in Figure 6, the As shown in Figure 6, the projection of a vector is a point, while the projection of a cone is a circle of diameter d. Referring to projection of a vector is a point, while the projection of a cone is a circle of diameter d. Referring to projection of a vector is a point, while the projection and of a cone is a circle of Referring to Figure 7a–c, for any pair of distinct sensor nodes , where, 1 , diameter , let od.x and o y be the Figure 7a–c, for any pair of distinct sensor nodes and , where, 1 , , let o x and o y be the Figure 7a–c, for any pair of distinct sensor nodes s ax and s ay , where, 1 ≤ x, y ≤ , and let d k, let ox and oy be and on the unit sphere O centered at grid point xy be the projection points of * * and on the unit sphere O centered at grid point , and let d xy be the projection points of gs gs the projection points of and on the unit sphere O centered at grid point g, and let d be the ax ay xy xy and Euclidean distance between the two points. There are three cases of the relationship between d Euclidean distance between the two points. There are three cases of the relationship between d xy and Euclidean distance between the two points. There are three cases of the relationship between dxy and d, which correspond to the three relationship cases of A and B. The three cases are elaborated below. d, which correspond to the three relationship cases of A and B. The three cases are elaborated below. d, which correspond to the three relationship cases of A and B. The three cases are elaborated below. Case (a) dxy = d. This case corresponds to the case of B = A. For such a case, one candidate cone is Case (a) d Case (a) dxy = d. This case corresponds to the case of B = A. For such a case, one candidate cone is xy = d. This case corresponds to the case of B = A. For such a case, one candidate cone is generated. Note that o x and o y should be on the diameter of the projection circle of the cone. Case (b) generated. Note that o generated. Note that oxx and o and oyy should be on the diameter of the projection circle of the cone. Case (b) should be on the diameter of the projection circle of the cone. Case (b) > d. This case corresponds to the case of B > A. For such a case, two candidate cones taking * dand > d. This case corresponds to the case of B > A. For such a case, two candidate cones taking gs x and as directions are generated. Case (c) A. For such a case, two candidate cones taking * and as directions are generated. Case (c) 1, sensor nodes *
*
*
algorithm u1 , ,u2 ,,.…. ., , uk going going from from g to The ACC ACC algorithm to s a1 , ,s a2 , ,.…. ., , s ak. . The algorithm algorithm generates generates kk unit unit vectors vectors constructs cones onon a node by node basis. basis. In practice, the ACCthe algorithm initially constructs constructs candidate candidate cones a node by node In practice, ACC algorithm initially * aconstructs a candidate cone candidate cone Cone(u x ) toCone( cover )sensor node s for 1 ≤ x ≤ k. Besides covering node s , the ACC, to cover sensor node for 1 ≤ x ≤ k. Besides covering node ax ax * * u x )Cone( algorithm adaptively and greedily adjusts u x so that Cone( can cover as many as possible other so that ) can cover as many as possible the ACC algorithm adaptively and greedily adjusts * * * sensor nodes. This is achieved by adjusting the direction of u by composing u x and x y for every other sensor nodes. This is achieved by adjusting the direction of by composing uand for *
vector uy , where, where 1 ≤ y ≤ k and x ≠ y (Figure 9). The direction adjustment (Lines 12–14) proceeds 1 ≤ y ≤ k and x 6= y (Figure 9). The direction adjustment (Lines 12–14) proceeds every vector on a node-by-node basis. Certainly, the direction adjustment should guarantee that the number of on a node‐by‐node basis. Certainly, the direction adjustment should guarantee that the number of * * sensor byby Cone( u x ) is) increasing andand that that the sensor node snode by Cone(uby sensor nodes nodes covered covered Cone( is increasing the sensor is still covered a x is still covered x) * whatever the adjusted u x is. is. Cone( ) whatever the adjusted
Figure 8. The adaptive cone covering (ACC) algorithm. Figure 8. The adaptive cone covering (ACC) algorithm.
Energies 2016, 9, 696
10 of 22
Figure 8. The adaptive cone covering (ACC) algorithm.
Figure 9. Illustration of candidate cone direction adjustment by composing two vectors. Figure 9. Illustration of candidate cone direction adjustment by composing two vectors.
In the GCC and the ACC algorithms, US stands for the set of unmarked sensor nodes with unsatisfied energy requirement. For most cases, the two algorithm stop and return CS* and US, with US is empty. For such cases, the cones in CS* can be used to deploy chargers to satisfy the energy requirements of all sensor nodes. On the contrary, when the algorithms terminate and return a non-empty set of US, they should re-run by setting NS = US, and the re-running should proceed until US is empty. The re-running is to satisfy the pending energy requirements of unmarked nodes. When US is empty, the requirements of all sensor nodes are satisfied and all the cones ever reported in CS* by running and re-running the algorithms are the final selected cones for deploying chargers. However, the re-running algorithms will report an error if it finds that no available grid point exists, where an available grid point is a grid point that no more than Q selected cones take as their apex. For such an error condition, we may need to decrease S, the separation of grid points, to increase the total number of grid points for more possible locations for deploying chargers to satisfy sensor nodes’ energy requirements. 5. Analysis 5.1. Algorithm Time Complexity Analysis In this subsection, we analyze the time complexity of the two proposed algorithms. In the GCC algorithm, for every grid point g, a half sphere HS centered at g of radius D is formed (Lines 3–4). If S covers k sensor nodes, then candidate cones are generated on a node-pair-by-node-pair basis (Lines 9–17). There are up to 4C2k candidate cones that can be generated, where k ≤ n. Since there are m grid points, the number of candidate cones generated is then 4mC2k ≤ mn2 = O(mn2 ) and the time complexity to generate the candidate cones is also O(mn2 ). On each iteration in the repeat-until loop (Lines 18–24), the GCC algorithm first selects the cone c from CS that covers most unmarked nodes in NS (Line 19). Since there are at most mn2 cones in CS, and the algorithm should check if a cone can cover all nodes, the selection can be done in O(mn3 ) time complexity. The GCC algorithm then checks if every unmarked node covered by cone c can be marked after updating its energy requirement (Lines 21–23). The checking takes O(n) time complexity. A cone can be selected only once in the repeat-until loop, so there are at most O(mn2 ) iterations in the loop. In summary, there are O(mn2 ) iterations in the repeat-until loop, and each iteration’s time complexity is O(mn3 ) (Line 19). Thus, the time complexity of the loop is O(m2 n5 ). Since the repeat-until loop has the highest time complexity in the algorithm, we thus have that the time complexity of the GCC algorithm is O (m2 n5 ). The ACC algorithm forms a half sphere HS centered at g of radius D for every grid point g (Lines 3–4). If HS covers k sensor nodes, then k candidate cones will be generated, k ≤ n (Line 8). Every candidate cone is tested for possible direction adjustment for k times (Lines 9–15), so there are k2 = O(n2 ) adjustment tests. Since there are m grid points, the ACC algorithm forms m half spheres. Let the half spheres cover k1 , k2 , . . . , k m sensor nodes, respectively. Thus, the total number of candidate cones is ∑im=1 k i ≤ mn = O(mn) and the total number of adjustment tests is of the time complexity O(mn3 ). On each iteration in the repeat-until loop (Lines 16–22), the ACC algorithm first selects the cone c from CS that covers the most unmarked nodes in NS (Line 17). Since there are at most mn cones
Energies 2016, 9, 696
11 of 22
in CS, and the algorithm should check if a cone can cover all nodes, the selection (Line 17) can be done in O(mn2 ) time complexity. The ACC algorithm then checks if every unmarked node covered by cone c can be marked after updating its energy requirement (Lines 19–21). The checking takes O(n) time complexity. A cone can be selected only once in the repeat-until loop, so there are at most O(mn) iterations in the loop. Thus, the time complexity of the repeat-until loop is O(m2 n3 ). Since the repeat-until loop has the highest time complexity in the ACC algorithm, the time complexity of the ACC algorithm is O(m2 n3 ). 5.2. Grid Point Separation Analysis Energies 2016, 9, 696 In this subsection,
11 of 21 we analyze the relationship of S, D and H, which are the grid point separation, the effective charging distance and high of the deployment plane, respectively. For an arbitrary sensor charging distance). The intersection of O and the charger deployment plan (i.e., the grid plane) P is a node located at position v = (X, Y, Z), we can construct a sphere O with radius D (i.e., the charging circle C with radius R distance). The intersection of O and the centered at u (Figure 10). For any point u’ within the area of charger deployment plan (i.e., the grid plane) P is a circle q . So, if there is at least one grid point within C, the sensor node located at v can C, we have Energies 2016, 9, x ′ 0 2 11 of 21 C with radius R = D2 − ( H − Z ) centered at u (Figure 10). For any point u within the area of C, be covered by a charger deployed at . When Z = 0, v is on the floor plane and the circle C has the 0 we have vu charging distance). The intersection of O and the charger deployment plan (i.e., the grid plane) P is a ≤ D. So, if there . As shown in Figure 11, if is at least one grid point g within C,, then there may be no grid point the sensor node located at v can smallest radius √ √2 circle C with radius R centered at u (Figure 10). For any point u’ within the area of be covered by a charger at g. When Z = 0, v is on the floor plane and the circle C has the √ ′ deployed within the area of circle C. On the contrary, if 2 √ , there is at least one grid point √2 = within C, the sensor node located at v can C, we have . So, if there is at least one grid point 2 − H 2 . As shown in Figure 11, if S > smallest radius r = D 2 r, then there may be no grid point be covered by a charger deployed at . When Z = 0, v is on the floor plane and the circle C has the p √ ensures that every sensor node within the area of circle C. In summary, In summary, S ≤ 2 2 within the area of circle C. On if S ≤ 2 r= p2(√2 D , then there may be no grid point − H 2 ), there is at least one grid point . As shown in Figure 11, if smallest radius √ the contrary, is covered by at least one charger deployed at one of the grid points. within the area of circle C. On the contrary, if = 2 2( D 2 − , there is at least one grid point within the area of circle C. In summary, In summary,√2 S≤ H 2 ) ensures that every sensor node ensures that every sensor node within the area of circle C. In summary, In summary, S ≤ 2 is covered by at least one charger deployed at one of the grid points. is covered by at least one charger deployed at one of the grid points.
Figure 10. Illustration of the intersection of the grid plane with an upper half sphere of radius D (the Figure 10. Illustration of the intersection of the grid plane with an upper half sphere of radius D (the charging distance) centered at a sensor node. Figure 10. Illustration of the intersection of the grid plane with an upper half sphere of radius D (the
charging distance) centered at a sensor node. charging distance) centered at a sensor node.
Figure 11. Illustration of no grid point within the area of circle C with radius r.
Below we show that if , then there is at least two grid points within the area √ of circle C centered at u with radius r. implies √2 , which in turns implies there is at least one grid point within C by the fact just mentioned in the last paragraph. Let C’ be the circle centered at with radius r and let and b be the intersection of C and C’ (Figure 12). Since C and Figure 11. Illustration of no grid point within the area of circle C with radius r. Figure 11. Illustration of no grid point ∠within the area of circle C with radius r. C’ have the same radius, we have that the angel associated with the minor arc should be larger than 120°; otherwise, will not be within C. For neighboring grid points , ’ and ′′ that Below we show that if , then there is at least two grid points within the area √ √ ∠ ′ ′′ is 90°. We have that either are apart from each other with separation S, their associated angle Below we′ or show that if S ≤ r = D2 − H 2implies , then there is at least two grid points within the area of circle C centered at u with radius r. √ . Therefore, there are at least two grid √2 , which in turns implies there is at ′′ is within the sector associated with the minor arc of circle C centered at u within C by the fact just mentioned in the last paragraph. Let C’ be the circle with radius r. S ≤ r implies S ≤ 2 r, which in turns implies there is at least points within the area of C. That is to say, there are at least two chargers covering a sensor node. In least one grid point 0 summary, S ≤ C by the fact ensures mentioned that every sensor node is paragraph. covered by at Let least C two one grid point g within in the last bechargers the circle centered centered at deployed at two of the grid points. with radius r and let just and b be the intersection of C and C’ (Figure 12). Since C and 0 0
at g with radius r and let a and b be the intersection and C (Figure 12). Since C and C should be have the C’ have the same radius, we have that the angel ∠ of Cassociated with the minor arc ◦ a b same radius, we have that the angel ∠agb associated with the minor arc ab should be, larger larger than 120°; otherwise, will not be within C. For neighboring grid points ’ andthan′′ 120 that ; 0 00 g' otherwise, g will not be within C. For neighboring that are apart from each r grid points g, g and g are apart from each other with separation S, their associated angle ∠ ′ ′′ is 90°. We have that either Cʹ 0Sgg00 is 90◦ . We other with separation S, their associated angle ∠ g have that either g0 or g00 is within the ′ or ′′ is within the sector associated with the minor arc . Therefore, there are at least two grid g b sector associated with the minor arc ab.C Therefore, there are at least two grid points within the area of u points within the area of C. That is to say, there are at least two chargers covering a sensor node. In S
r
summary, S ≤ ensures that every g''sensor node is covered by at least two chargers deployed at two of the grid points. b Figure 12. Illustration of two circles both of the radius r intersecting at two points a and b. a
least one grid point within C by the fact just mentioned in the last paragraph. Let C’ be the circle centered at with radius r and let and b be the intersection of C and C’ (Figure 12). Since C and C’ have the same radius, we have that the angel ∠ associated with the minor arc should be larger than 120°; otherwise, will not be within C. For neighboring grid points , ’ and ′′ that are apart from each other with separation S, their associated angle ∠ ′ ′′ is 90°. We have that either Energies 2016, 9, 696 12 of 22 ′ or ′′ is within the sector associated with the minor arc . Therefore, there are at least two grid points within the area of C. That is to say, there are at least two chargers covering a sensor node. In p 2 − H2 ) C. That is toS say, two chargers covering sensoris node. In summary, S ≤two ( Dchargers summary, ≤ there are at least ensures that every sensor a node covered by at least ensures that every sensor node is covered by at least two chargers deployed at two of the grid points. deployed at two of the grid points. a
g'
r Cʹ
S
g C
u S
r
g'' b
Figure 12. Illustration of two circles both of the radius r intersecting at two points a and b. Figure 12. Illustration of two circles both of the radius r intersecting at two points a and b.
For even smaller separation, we can use the “dies per wafer” equation in [25] to estimate √ 2 approximately the number of grid points within circle C with radius r = D − H 2 . The equation is shown below: π × (Wafer Ridus)2 2π × (Wafer Ridus) Dies per wafer = − p (4) Die area 2 × (Die area) If we take the grid square as the die area and the wafer radius as r, the radius of circle C, then we have the following equation for NGS , the number of grid squares in circle C: NGS =
2π × r π × r2 − √ S2 S 2
(5)
One grid square has four grid points. However, adjacent grid square share grid points. When we count the number of grid points, a grid square is associated with two non-redundant grid points. Therefore, we have the following equation for NGP , the number of grid points in circle C: 2π × r π × r2 − √ NGP = 2 × NGS = 2 × (6) S2 S 2 √ In summary, for grid separation S much smaller than r = D2 − H 2 , each sensor node is covered by
at least 2 ×
π×r 2 S2
−
2π√×r S 2
chargers approximately.
6. Experiments and Simulations 6.1. Experiments In this subsection, we show results of extensive experiments in indoor environments for modeling the RF charging efficiency (i.e., the received energy or received power). The devices used in the experiments are Powercast TX91501-3W-ID wireless charger [21] and P2110-EVAL-02 power harvester [22], as shown in Figure 13. The former is a wireless charger with a patch directional antenna with 8 dBi gain and 60◦ beamwidth in both horizontal and vertical directions. It is a 915 MHz transmitter with 3 Watts equivalent isotropically radiated power (EIRP). The latter is a power harvester with a dipole omni-directional antenna with 1 dBi gain. It can harvest energy from RF waves, provide intermittent/pulsed power output up to 4.21 V, and connect to rechargeable batteries.
in the experiments are Powercast TX91501‐3W‐ID wireless charger [21] and P2110‐EVAL‐02 power in the experiments are Powercast TX91501-3W-ID wireless charger6.1. [21]Experiments and P2110-EVAL-02 power harvester [22], as shown in Figure 13. The former is a wireless charger with a patch directional harvester [22], as shown in Figure 13. The former is a wireless charger with a patch directional In this subsection, we show results of extensive exp antenna with 8 dBi gain and 60° beamwidth in both horizontal and vertical directions. It is a 915 MHz antenna with 8 dBi gain and 60° beamwidth in both horizontal and vertical directions. It is a 915efficiency MHz modeling the RF charging (i.e., the received energy with equivalent 3 Watts equivalent isotropically radiated power (EIRP). The latter is a power transmittertransmitter with 3 Watts isotropically radiated power (EIRP). The latter is a power in the experiments are Powercast TX91501-3W-ID wireless c harvester with Energies 9, 696a dipole omni‐directional antenna with 1 dBi gain. It can harvest energy from 13 ofRF 22 harvester with a 2016, dipole omni-directional antenna with 1 dBi gain.harvester It can harvest RF [22], asenergy shownfrom in Figure 13. The former is a wir waves, provide intermittent/pulsed power output up to 4.21 V, and connect to rechargeable batteries. waves, provide intermittent/pulsed power output up to 4.21 V, and antenna connect to rechargeable with 8 dBi gain batteries. and 60° beamwidth in both horizonta
transmitter with 3 Watts equivalent isotropically radiated harvester with a dipole omni-directional antenna with 1 dB waves, provide intermittent/pulsed power output up to 4.21 V
(a)
(a)
(b)
(b)
Figure 13. (a) Powercast TX91501‐3W‐ID wireless charger; and (b) P2110‐EVAL‐02 power harvester. 13. (a)TX91501-3W-ID Powercast TX91501-3W-ID wireless and (b) P2110-EVAL-02 power harvester. Figure 13. (a)Figure Powercast wireless charger; andcharger; (b) P2110-EVAL-02 power harvester.
As shown in Figure 14, a wireless charger and a power are harvester are set up. The distance D (a)Ddistance As shown As in Figure a wireless and a power set up.are The shown 14, in Figure 14, acharger wireless charger andharvester a power harvester setdistance up. The D between the two devices is set to be 0.5 m, 1.0 m, …, or 4.5 m for experiments. As shown in Figure 15, between the two devices is set to be 0.5 m, 1.0 m, …, or 4.5 m for experiments. As shown in Figure 15, between the two devices is set to be 0.5 m, 1.0 m, . . . , or 4.5 m for experiments. As shown in Figure 15, 13. angles, (a) Powercast TX91501-3W-ID wireless charger; and the orientation of the charger’s patch antenna is specified Figure by two the horizontal angle the orientation of the charger’s patch antenna is specified by two angles, thethehorizontal the orientation of the charger’s patch antenna is specified by two angles, horizontalangle angle (azimuth) (azimuth) ф and the vertical angle (elevation) θ. Either angle is set to be 0°, 15°, …, or 90°. Figure 16 ◦ , 15 ◦ , . …, (azimuth) ф and and the the vertical vertical angle angle (elevation) (elevation)θ. θ.Either Eitherangle angleisisset settotobe be0As 0°, 15°, or 90°. . . ,in orFigure 90◦ . Figure Figure 16 illustrates shown 14, a16 wireless charger and a powe illustrates the distance and the angle settings. In Figure 16, the charger is assumed to be at the origin illustrates the distance and the angle settings. In Figure 16, the charger is assumed to be at the origin the distance and the angle settings. In Figure 16, the charger is assumed to be at is the andm, the between the two devices setorigin to be 0.5 1.0 m, …, or 4.5 m f and the harvester is assume to be on the x‐axis and be D apart from the charger. Moreover, the and the harvester the the harvesteris isassume assumetotobebeon onthe thex-axis x-axisand andbe beDD apart apart from fromorientation the charger. charger. Moreover, the patch charger’s the of Moreover, the charger’s antenna is specified charger’s antenna direction is indicated by vector with the horizontal angle ф and the vertical ⃑ with charger’s antenna direction is indicated by vector* 𝑁 the horizontal angleand ф and the vertical (azimuth) the vertical vertical angleθ(elevation) antenna direction is indicated by vector N with the horizontal angle ф and the angle relative θ. Either angle angle θ relative to the x‐axis. angle θ relative to the x-axis. illustrates the distance and the angle settings. In Figure 16, th to theEnergies 2016, 9, 696 x-axis. 13 of 21 and the harvester is assume to be on the x-axis and be D a ⃑ with t charger’s antenna direction is indicated by vector 𝑁 Energies 2016, 9, 696 13 of 21 angle θ relative to the x-axis.
Figure 14. The experimental setting of a wireless charger and a power harvester.
Figure 14. The experimental setting of a wireless charger and a power harvester. Figure 14. The experimental setting of a wireless charger and a power harvester.
(a) (a)
(b)(b)
(c)
(d)
Figure 15. Experimental settings for (a,b) different horizontal angles (azimuths); and (c,d) different (c) (d) 15.vertical angles (elevations). Experimental settings for (a,b) different horizontal angles (azimuths); and (c,d)
Figure different Figure 15. Experimental settings for (a,b) different horizontal angles (azimuths); and (c,d) different vertical angles (elevations). vertical angles (elevations).
antenna with 8 dBi gain and 60° beamwidth in both h
harvester transmitter with a dipolewith omni-directional antenna with 1 dBi gain.r 3 Watts equivalent isotropically waves, provide intermittent/pulsed power output up to 4.21 V, and c harvester with a dipole omni-directional antenna w waves, provide intermittent/pulsed power output up (d)
(c)
Figure 15. Experimental settings for (a,b) different horizontal angles (azimuths); and (c,d) different Energies 2016, 9, 696 14 of 22 vertical angles (elevations).
(a)
Figure 13. (a) Powercast TX91501-3W-ID wireless charger; and (b) P21
(a)
As shown in Figure 14, a wireless charger and a power harve between the two devices is set to be 0.5 m, 1.0 m, …, or 4.5 m for expe Figure 13. (a) Powercast TX91501-3W-ID wireless cha the orientation of the charger’s patch antenna is specified by tw Figure 16. The illustration of distance D, horizontal angle ф, and vertical angle θ. (azimuth) vertical θ. Eithercharger angle is set Asthe shown inangle Figure 14, a wireless andto Figure 16. The illustration of distance D, horizontal angle ф, and and vertical angle θ. (elevation) illustratesbetween the distance angle settings. theand twothe devices is set to In beFigure 0.5 m,16, 1.0the m,charg …, o By a multimeter, we measure the voltage (V) and current (I) of the output of the harvester 10 and the harvester is assume to be on the x-axis and be D apart fr the charger’s patch antenna is s By a multimeter, we measure the voltage (V) and current (I) ofthe the orientation output of theofharvester 10 times ⃑ with the hori charger’s antenna direction is indicated by vector 𝑁 times for every distance D = 0.5 m, 1.0 m, …, and 4.5 m, every horizontal angle ф = 0°, 15°, …, and ◦ ◦ ◦ (azimuth) ф and the vertical angle (elevation) θ. Eith for every distance D = 0.5 m, 1.0 m, . . . , and 4.5 m, every horizontal angle = 0 , 15 , . . . , and 90 , angle θ relative to the x-axis. 90°, and every vertical angle θ = 0°, 15°, …, and 90°. We average the measured data and then apply
distance and the angle and every vertical angle θ = 0◦ , 15◦ , . . . , and 90◦ . We average theillustrates measuredthe data and then apply the settings. In Figu the average data to derive the received power (i.e., received energy). Table 1 (respectively, Table 2) and the harvester is assume to be on the x-axis an average data to derive the received power (i.e., received energy). Table 1 (respectively, Table 2) shows shows the experimental results of the received power in the unit of mW for different distances and ⃑ charger’s antenna direction is indicated by vector 𝑁 the experimental results of the received power in the unit of mW for different distances and horizontal horizontal (respectively, vertical) angles with the vertical (respectively, horizontal) angle is 0. Note angleangle θ relative to the x-axis. (respectively, vertical) angles with the vertical (respectively, horizontal) is 0. Note that 0.01 V and that 0.01 V and 0.01 mA are the smallest measurable voltage and current, respectively. Therefore, the 0.01 mA are the smallest measurable voltage and current, respectively. Therefore, the voltage values voltage values smaller than 0.01 V or the current value smaller than 0.01 mA are not measurable and smaller than 0.01 V or the current value smaller than 0.01 mA are not measurable and such cases lead such cases lead to “not available” (i.e., “‐”) symbols shown in the Tables. to “not available” (i.e., “-”) symbols shown in the Tables. Energies 2016, 9, x
14 of 21
Table 1. Received power (mW) for different distances and horizontal angles.
Table 1. Received power (mW) for different distances and horizontal angles.
D\ф 0.5 m 0.5 m 1.0 m 1.0 m 1.5 m 1.5 m 2.0 m 2.0 m 2.5 m 2.5 m 3.0 m 3.0 m 3.5 m 3.5 m 4.0 m 4.0 m 4.5 m
4.5 m
0◦ 0°
15° 30◦ 17.63 16.38 17.63 16.38 13.89 6.3 5.72 6.3 5.72 3.96 1.93 1.26 1.93 1.26 1 1.39 1.04 1.39 1.04 0.8 0.84 0.74 0.84 0.74 0.62 0.47 0.31 0.47 0.31 0.18 0.28 0.15 0.28 0.15 0.21 0.11 0.21 0.11 0.14 ‐
0.14
15◦
-
-
30° 45◦ 13.89 9.89 3.96 2.29 1 0.72 0.8 0.57 0.62 0.34 0.18 ‐ ‐ ‐
-
◦ 6045°
75◦60°
90◦ 75°
9.89 3.96 2.29 1.14 0.72 0.42 0.57 0.25 0.34 ‐ ‐ ‐ ‐
3.96 2.15 1.14 0.58 0.42 0.25 ‐ ‐ ‐ ‐ ‐
2.15 0.37 0.58 ‐ ‐ ‐ ‐ ‐ ‐ ‐
-
-
-
90° 0.37 ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐
Table 2. Received power (mW) for different distances and vertical angles.
Table 2. Received power (mW) for different distances and vertical angles.
D\θ 0.5 m D\θ 1.0 m 1.5 m 0.5 m 2.0 m 1.0 m 2.5 m 1.5 m 3.0 m 2.0 m 3.5 m 2.5 m 4.0 m 3.0 m 4.5 m 3.5 m
0° 17.63 6.3 17.631.93 6.3 1.39 1.930.84 1.390.47 0.840.28 0.470.21 0.280.14
0◦
15◦
15° 13.52 ◦ 5.23 30 1.52 9.07 0.95 2.65 0.61 1.02 0.42 0.87 0.21 0.76 0.140.36 0.12 0.19
30° 9.07 ◦ 45 2.65 1.02 5.55 0.87 1.26 0.76 0.56 0.36 0.34 0.19 0.28 ‐ 0.15 ‐ -
45° 5.55 ◦ 60 1.26 0.56 2.78 0.34 0.79 0.28 0.29 0.15 0.18 -‐ -‐ -‐
60° 2.78 ◦ 750.79 0.29 0.92 0.18 0.41 - ‐ - ‐ - ‐ - ‐ - ‐
75° 0.92 0.41 0.26 ‐ - ‐ - ‐ - ‐ - ‐ - ‐ - ‐
90◦
90° 0.26 ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐
13.52 5.23 1.52 0.95 0.61 0.42 0.21 4.0 m 0.21 0.14 By experimental data in Table 1, Table 2 and the specification [13] stating that the charger has a 4.5 m 0.14 0.12 beamwidth of 60° in both horizontal and vertical directions, we model the charging space of a wireless charger as a cone with the opening angle of 60° and the generatrix length of D. We perform By experimental data in Table 1, Table 2 and the specification [13] stating that the charger has power regression on the data to obtain an exponential function taking D as the independent variable ◦ in both horizontal and vertical directions, we model the charging space of a beamwidth of 60 (i.e., x) and P as the dependent variable (i.e., y) for every vertical and horizontal angle of 0°, 15° and a wireless charger as a cone with the opening angle of 60◦ and the generatrix length of D. We perform 30°. For example, Figure 17 shows the functions derived by performing power regression analysis on experimental data of received energy (or power) for horizontal angel ф = 0°, 15° and 30°, and vertical power regression on the data to obtain an exponential function taking D as the independent variable angle θ = 0°. Moreover, Figure 18 shows the every functions derived performing angle power (i.e., x) and P as the dependent variable (i.e., y) for vertical andby horizontal ofregression 0◦ , 15◦ analysis on experimental data of received power for horizontal angel ф =0° and vertical angle θ = 0°, 15° and 30°. We can see that the functions, namely y = 5.019x−2.217, y = 3.8881x−2.225, y = 2.6394x−1.821, y = 1.3522x−1.939, y = 0.7134x−2.01, fit well with the experimental data, as they all have coefficient of determination (i.e., R2) [26] values larger than 0.958, which is very close to the maximum value 1. They also comply with the Friis equation, since the function exponents are approximately‐2. Therefore, we may well estimate the power received by the harvester with the functions. In practice,
Commented
4.5 m
0.14
equation for NGP‐ , the number‐ of grid points in circle C: 0.12 Therefore, ‐ we have the following ‐ ‐
π 𝑟2 2π 𝑟 (a) NGP = 2 × NGS = 2 × ( 2 − 12 of)21 Energies 2016, 9, 696 By experimental data in Table 1, Table 2 and the specification [13] stating that the charger has a 𝑆 √2 𝑆 Figure 13. (a) Powercast TX91501-3W-ID wirele beamwidth of 60° in both horizontal and vertical directions, we model the charging space 2 −a 2 , each sensor nod (a) for grid separation S much smaller 𝐻 √𝐷of For even smaller separation, In wesummary, can use the “dies per wafer” equation in than [25] 𝑟to=estimate 2 wireless charger as a cone with the opening angle of 60° and the generatrix length of D. We perform π𝑟 2π 𝑟 Energies 2016,approximately 9, 696 15 of 22 2 2 the number of grid r =As − 13. 𝐻 .(a) The equation isa wireless charge √𝐷 by atpoints least 2within × ( circle −C with ) radius chargers approximately. shown in Figure 14,TX91501-3W-ID Figure Powercast wirel 𝑆√2 𝑆2 power regression on the data to obtain an exponential function taking D as the independent variable shown below: between the two devices is set to be 0.5 m, 1.0 m (i.e., x) and P as the dependent variable (i.e., y) for every vertical and horizontal angle of 0°, 15° and 2 6. Experiments andRidus) Simulations π (Wafer 2π(Wafer Ridus) As shown of in the Figure 14, a wireless charge the orientation charger’s patch antenn and 30◦ . For example, FigureDies 17 per shows the wafer = functions derived − by performing power regression (4) 30°. For example, Figure 17 shows the functions derived by performing power regression analysis on Die area area) √2(Die ◦ ◦ ◦ between two devices 0.5 m, 1.0 θ m (azimuth) ф= and angle analysis on experimental data of received energy (or power) for horizontal angelthe 0 ,the 15 vertical and is 30set , to be(elevation) experimental data of received energy (or power) for horizontal angel ф = 0°, 15° and 30°, and vertical 6.1. Experiments ◦ illustrates the distance and the angle settings. In the orientation of the charger’s patch antenn If we take the grid square as the die area and the wafer radius as r, the radius of circle C, then and vertical angle θ = 0 . Figure Moreover, Figurethe 18 functions shows thederived functions by performing power angle θ = 0°. Moreover, 18 shows by derived performing power regression In this we show and results ofharvester experiments ◦ and we have the following equation forof NGS , thesubsection, number of grid squares in circle C:extensive the is assume toinbeindoor on theenviro x-ax (azimuth) ф =0 and the vertical angle (elevation) regression analysis on experimental data received power for horizontal angel vertical analysis on experimental data of received power for horizontal angel ф =0° and vertical angle θ = 0°, (a) (b) modeling the RF charging efficiency (i.e., the received energy or received power). The dI −2.217 −2.225 illustrates the and the angle settings. charger’s antenna direction is, indicated by vec π 𝑟 2 namely 2π 𝑟 y = 5.019x angle θ = 0◦ , 15◦ and 30◦ . We can see that the functions, , y =distance 3.8881x −2.217 −2.225, y = 2.6394x −1.821, y = 15° and 30°. We can see that the functions, namely y = 5.019x , y = 3.8881x (5) N GS = − in the experiments are Powercast TX91501-3W-ID wireless charger [21] and P2110-EVA −1.821 , y = 1.3522x −1.939(a) − 2.01 , fit well and the harvester is all assume to be on the x-a angle θ relative the x-axis. 𝑆2 𝑆√2 Figure TX91501-3W-ID wireless charger; and (b) P2110-EVAL-02 power y1.3522x = 2.6394x , yPowercast = 0.7134x with the experimental data, asto they have −1.939, y = −2.01, 13. 0.7134x fit well with the experimental as 13. they have of harvester. harvester [22], as shown indata, Figure Theall former iscoefficient a wireless charger with a patch 2 antenna direction is indicated by ve coefficient of determination (i.e., R four ) [26] values larger than which issquare veryinclose thepoints. maximum One grid square has grid points. However, adjacent gridcharger’s share grid When 2) [26] values larger than 0.958, which is very close to the maximum value 1. determination (i.e., R antenna with 8 dBi gain0.958, and 60° beamwidth bothto horizontal and vertical directions. It is As shown inFriis Figure wireless charger and atwo power harvester arex-axis. set up. The distance D angle θ relative to the we count the number ofthe grid points, a14, grida with square associated with non-redundant grid points. value They also comply with equation, since the function exponents are approximately-2. transmitter 3 isWatts equivalent isotropically radiated power (EIRP). The latter They 1.also comply with the Friis equation, since the function exponents are approximately‐2. between the two devices is set to be 0.5 m, 1.0 m, …, or 4.5 m for experiments. As shown in Figure 15 Therefore, we have the following equation for N GP , the number of grid points in circle C: Therefore, we may well estimate the power received by the harvester with the functions. In practice, harvester with a dipole omni-directional antenna with 1 dBi gain. It Therefore, we may well estimate the power received by the harvester with the functions. In practice, can harvest energ orientation of the charger’s patch antenna is 𝑟power specified byup two angles, the horizontal angle waves, intermittent/pulsed to 4.21 V, and connect to rechargeab π 𝑟2 2π we interpolate thethe values of the functions to provide estimate the received power foroutput given parameters, like we interpolate the values of the functions to estimate the received power for given parameters, like NGP = 2 × NGS = 2 × ( 2 − ) (6) √2 (azimuth) ф and the vertical angle (elevation) θ. Either angle is set to be 0°, 15°, …, or 90°. Figure 16 𝑆 𝑆 distance D, horizontal angle , and vertical angle θ. distance D, horizontal angle ф, and vertical angle θ. illustrates the distance and the angle settings. In Figure 16, the charger is assumed to be at the origin In summary, for grid separation S much smaller than 𝑟 = √𝐷 2 − 𝐻 2 , each sensor node is covered and the πharvester 𝑟2 2π 𝑟is assume to be on the x-axis and be D apart from the charger. Moreover, the by at least 2 × ( 2 − ) chargers approximately. 𝑆√2direction is indicated by vector 𝑁 ⃑ with the horizontal angle ф and the vertica charger’s 𝑆antenna angle θ relative to the x-axis. 6. Experiments and Simulations
6.1. Experiments
In this subsection, we show results of extensive experiments in indoor environments for modeling the RF charging efficiency (i.e., the received energy or received power). The devices used (b) in the experiments are Powercast TX91501-3W-ID wireless (a) charger [21] and P2110-EVAL-02 power harvester [22], as shown in FigureFigure 13. The is a wireless charger with charger; a patch and directional 13. former (a) Powercast TX91501-3W-ID wireless (b) P2110-EVAL-02 power ha antenna with 8 dBi gain and 60° beamwidth in both horizontal and vertical directions. It is a 915 MHz transmitter with 3 Watts equivalent radiated is a harvester power As isotropically shown in Figure 14, a power wireless(EIRP). charger andlatter a power are set up. The The harvester with a dipole omni-directional withis1set dBi from RF between theantenna two devices to gain. be 0.5Itm,can 1.0harvest m, …, orenergy 4.5 m for experiments. As shown i Figure 17. The functions derived by performing power regression analysis on experimental data (ED) waves, provide intermittent/pulsed power output up to 4.21 V, and connect to rechargeable batteries. the orientation of the charger’s patch antenna is specified by two angles, the horiz Figure 17. The functions derived by performing power regression analysis on experimental data (ED) of the power received (PR) for horizontal angel ф = 0°, 15° and 30°, and vertical angle θ = 0°. ◦ ◦ ◦ ◦ (azimuth) ф and the vertical angle (elevation) θ. Either angle is set to be 0°, 15°, …, or 90 of the power received (PR) for horizontal angel = 0 , 15 and 30 , and vertical angle θ = 0 . Energies 2016, 9, 696 illustrates the distance and the angle settings. In Figure 16, the15 of 21 charger is assumed to be a and the harvester is assume to be on the x-axis and be D apart from the charger. Mo ⃑ with the horizontal angle ф and charger’s antenna direction is indicated by vector 𝑁 angle θ relative to the x-axis.
(a)
(b)
Figure 13. (a) Powercast TX91501-3W-ID wireless charger; and (b) P2110-EVAL-02 power harvester.
As shown in Figure 14, a wireless charger and a power harvester are set up. The distance D between the two devices is set to be 0.5 m, 1.0 m, …, or 4.5 m for experiments. As shown in Figure 15, Figure 18. orientation The functions derived by performing power analysis on ED the of the PR for angle of the charger’s patch antenna is regression specified two angles, Figurethe 18. The functions derived by performing power regressionbyanalysis on ED of horizontal the PR for horizontal angel ф = 0°, and vertical angle θ = 0°, 15° and 30°. ◦ ◦ ◦ ◦ (azimuth) ф and the vertical angle (elevation) θ. Either angle is set to be 0°, 15°, …, or 90°. Figure 16 horizontal angel = 0 , and vertical angle θ = 0 , 15 and 30 . illustrates the distance and the angle settings. In Figure 16, the charger is assumed to be at the origin 6.2. Simulations and the harvester is assume to be on the x-axis and be D apart from the charger. Moreover, the 6.2. Simulations ⃑ with the horizontal angle ф and the vertical charger’s antenna direction is indicated by vector 𝑁 This subsection shows the simulation results of the proposed GCC algorithm and ACC This subsection shows the x-axis. simulation results of the proposed GCC algorithm and ACC algorithm. angle θ relative to the
algorithm. The simulator is implemented in C++ language, and the simulation parameters are shown The simulator is implemented in C++ language, and the simulation parameters are shown in Table 3. in Table 3. The sensors are distributed randomly in a 20 m × 15 m × 2.3 m cuboid. The chargers can The sensors are distributed randomly in a 20 m × 15 m × 2.3 m cuboid. The chargers can be deployed at be deployed at grid points on the grid plane at the height of 2.3 m, and the grid point separation is grid points on the grid plane at the height of 2.3 m, and the grid point separation is 1.8 m. The charging 1.8 m. The charging distance is 3 m, and the charging cone opening angle is 30°. It is noted that 30 distance is 3 m, and the charging cone opening angle is 30◦ . It is noted that 30 experiments were experiments were executed per simulation case for averaging the simulation results. The parameter executed per simulation case for averaging the simulation results. The parameter settings are shown settings are shown in Table 3. in Table 3. We first perform simulations for three scenarios of coverage requirements: all sensor nodes require 1‐, 2‐, or 3‐coveage of chargers. Figure 19 shows the comparisons of GCC and ACC algorithms in terms of the number of the candidate cones generated. The simulation results are all the same for the 1‐, 2‐ and 3‐coverage requirement scenarios, so only one diagram is shown in Figure 19. We can observe that GCC generates much more candidate cones than ACC. This is because GCC has O(mn2) of candidate cones, while ACC has O(mn) of candidate cones, where n is the number of sensor nodes and m is the number of grid points.
This subsection shows the simulation results of the proposed GCC algorithm and ACC algorithm. The simulator is implemented in C++ language, and the simulation parameters are shown in Table 3. The sensors are distributed randomly in a 20 m × 15 m × 2.3 m cuboid. The chargers can be deployed at grid points on the grid plane at the height of 2.3 m, and the grid point separation is 1.8 m. The charging distance is 3 m, and the charging cone opening angle is 30°. It is noted that 30 Energies 2016, 9, 696 16 of 22 experiments were executed per simulation case for averaging the simulation results. The parameter settings are shown in Table 3. We first simulations forfor three scenarios of coverage requirements: all sensor nodes require We first perform perform simulations three scenarios of coverage requirements: all sensor nodes 1-, 2-, or 3-coveage of chargers. Figure 19 shows the comparisons of GCC and ACC algorithms in terms require 1‐, 2‐, or 3‐coveage of chargers. Figure 19 shows the comparisons of GCC and ACC algorithms of the number of the candidate cones generated. The simulation results are all the same for the 1-, 2in terms of the number of the candidate cones generated. The simulation results are all the same for and 3-coverage requirement scenarios, so only one diagram is shown in Figure 19. We can observe that the 1‐, 2‐ and 3‐coverage requirement scenarios, so only one diagram is shown in Figure 19. We can 2) GCC generates much more candidate cones than ACC. This is because GCC has O(mn2 ) of candidate observe that GCC generates much more candidate cones than ACC. This is because GCC has O(mn cones, while ACC has O(mn) of candidate cones, where n is the number of sensor nodes and m is the of candidate cones, while ACC has O(mn) of candidate cones, where n is the number of sensor nodes number of grid points. and m is the number of grid points. Table 3. Simulation settings. Table 3. Simulation settings.
Parameter ItemItem Parameter Sensor deployment cuboid 20 m (L) × 15 m (W) × 2.3 m (H) Sensor deployment cuboid 20 m (L) × 15 m (W) × 2.3 m (H) 50, 100, 150, 200, 250 Number of sensors Number of sensors 50, 100, 150, 200, 250 Effective charging distance Effective charging distance 33 m (D) m (D) Opening angle 30° (A) ◦ (A) Opening angle 30 Height of grid points 2.3 m (H) Height of grid points 2.3 m (H) 1.8 m (S) GridGrid point separation point separation 1.8 m (S) Coverage requirement of sensor nodes 1-, 2-, or 3-coverage Coverage requirement of sensor nodes 1‐, 2‐, or 3‐coverage Energy requirement of sensor nodes 0.18, 0.54, 0.9 mW Energy requirement of sensor nodes 0.18, 0.54, 0.9 mW Times of simulation per case 30 times/case Times of simulation per case 30 times/case
Figure 19. The comparisons of GCC and ACC in terms of the number of generated candidate cones. Figure 19. The comparisons of GCC and ACC in terms of the number of generated candidate cones.
Figure 20 shows the comparisons of GCC and ACC in terms of the number of the chargers deployed for the 1-, 2- and 3-coverage requirement scenarios. By Figure 20, we can observe that GCC outperforms ACC in terms of the number of chargers required. This is because GCC greedily adjusts charger antenna directions to the extreme for some cases to cover more sensor nodes by one candidate cone, and GCC generates more candidate cones to be selected. On the contrary, ACC adaptively adjusts antenna directions to cover sensor nodes with smaller horizontal angels and vertical angels, and ACC generates fewer candidate cones. Figure 21 shows the comparisons of GCC and ACC in terms of execution time for scenarios of different coverage requirements. We can observe that ACC runs much faster than GCC, especially for the scenario of high energy consumption. This is because ACC has O(m2 n3 ) time complexity, while GCC has O(m2 n5 ) time complexity. However, the time complexity of the big O notation is just an upper bound. Some experimental cases may have sensor nodes whose coverage requirements are met very early. For such cases, the algorithm execution may not be too long. This accounts for why the execution times of algorithms do not grow according to the functions of m2 n3 or m2 n5 . We now consider the form of power requirements, in which every sensor node present its power in the unit of mW for compensating its power consumption. For setting simulation parameters, we derive the sensor node power consumption from the specification of TI eZ430 (Texas Instruments Inc., Dallas, TX, USA) [27], which a common off-the-shelf sensor node device. The working voltage of eZ430
Energies 2016, 9, 696 Energies 2016, 9, 696
16 of 21 17 of 22
Figure 20 shows the comparisons of GCC and ACC in terms of the number of the chargers is 1.8 V and the working currents are 270 µA, 0.1 µA, 11.1 mA and 16.6 mA for processor processing, deployed for the 1‐, 2‐ and 3‐coverage requirement scenarios. By Figure 20, we can observe that GCC sleep mode, transmission mode, and reception mode, respectively. According to Reference [28], the outperforms ACC in terms of the number of chargers required. This is because GCC greedily adjusts sensor node can use the energy most efficiently when the ratios for the processor processing, sleep charger antenna directions to the extreme for some cases to cover more sensor nodes by one candidate mode, transmission mode,more and reception 0.47%, 98.77%, and 0.28%, respectively. cone, and GCC generates candidate mode cones are to be selected. On 0.48% the contrary, ACC adaptively According to the ratios, the current is about 0.1 mA for the eZ430-based sensor node to use the energy adjusts antenna directions to cover sensor nodes with smaller horizontal angels and vertical angels, most efficiently, so the power consumption is thus about 0.1 mA × 1.8 V = l.8 mW for such ratios. and ACC generates fewer candidate cones.
(a)
(b)
(c) Figure 20. The comparisons of GCC and ACC in terms of the number of deployed chargers for (a) Figure 20. The comparisons of GCC and ACC in terms of the number of deployed chargers for: 1‐coverage; (b) 2‐coverage; and (c) 3‐coverage requirement. (a) 1-coverage; (b) 2-coverage; and (c) 3-coverage requirement.
Figure 21 shows the comparisons of GCC and ACC in terms of execution time for scenarios of We perform simulations for three scenarios of energy requirements, namely the low, medium, different coverage requirements. We can observe that ACC runs much faster than GCC, especially and high power requirement scenarios. For the low power requirement scenario, it is assumed 80% 2n3) time complexity, while for the scenario of high energy consumption. This is because ACC has O(m of sensor nodes are of 1.8 mW power requirement, 10%, 5.4 (=1.8 × 3) mW and 10%, 9.0 (=1.8 ×an 5) GCC has O(m2n5) time complexity. However, the time complexity of the big O notation is just mW. For the medium energy requirement scenario, it is assumed 10% of sensor nodes are of 1.8 mW upper bound. Some experimental cases may have sensor nodes whose coverage requirements are met very early. For such cases, the algorithm execution may not be too long. This accounts for why the execution times of algorithms do not grow according to the functions of m2n3 or m2n5.
Energies 2016, 9, 696
18 of 22
power requirement, 80%, 5.4 mW and 10%, 9.0 mW. And for the high power requirement scenario, it is assumed 10% of sensor nodes are of 1.8 mW power requirement, 10%, 5.4 mW and 80%, 9.0 mW. In the candidate cone generation phase, the GCC and the ACC algorithms generate the same number of candidate cones no matter what energy requirements (coverage requirements or power requirements) are considered. Please refer to Figure 19 for the comparisons of GCC and ACC in terms Energies 2016, 9, x 17 of 21 of the number of candidate cones.
(a)
(b)
(c) Figure 21. The comparisons of GCC and ACC in terms of the execution time for: (a) 1‐coverage; (b) 2‐ Figure 21. The comparisons of GCC and ACC in terms of the execution time for: (a) 1-coverage; covearge; and (c) 3‐coverage requirements. (b) 2-covearge; and (c) 3-coverage requirements.
We now consider the form of power requirements, in which every sensor node present its power Figure 22 shows the comparisons of GCC and ACC in terms of the number of the chargers in the unit of mW for compensating its power consumption. For setting simulation parameters, we deployed for different power requirement scenarios. We can observe that ACC outperforms GCC derive the sensor node power consumption from the specification of TI eZ430 (Texas Instruments for almost all scenarios. The superiority of ACC is most significant in the high power requirement Inc., Dallas, TX, USA) [27], which a common off‐the‐shelf sensor node device. The working voltage scenario, while is least significant in the low power requirement scenario. This is because GCC greedily of eZ430 is 1.8 V and the working currents are 270 μA, 0.1 μA, 11.1 mA and 16.6 mA for processor adjust directions of cones to the extreme for some sensor node-pair cases that nodes can cover by one processing, sleep mode, mode, reception mode, respectively. According cone. Consequently, sometransmission sensors are covered byand chargers with larger horizontal and vertical angles.to Reference [28], the sensor node can use the energy most efficiently when the ratios for the processor processing, sleep mode, transmission mode, and reception mode are 0.47%, 98.77%, 0.48% and 0.28%, respectively. According to the ratios, the current is about 0.1 mA for the eZ430‐based sensor node to use the energy most efficiently, so the power consumption is thus about 0.1 mA × 1.8 V = l.8 mW for such ratios.
Comme
of the number of candidate cones. Figure 22 shows the comparisons of GCC and ACC in terms of the number of the chargers deployed for different power requirement scenarios. We can observe that ACC outperforms GCC for almost all scenarios. The superiority of ACC is most significant in the high power requirement scenario, Energies 2016,while 9, 696 is least significant in the low power requirement scenario. This is because 19GCC of 22 greedily adjust directions of cones to the extreme for some sensor node‐pair cases that nodes can cover by one cone. Consequently, some sensors are covered by chargers with larger horizontal and Hence, those sensor nodes may need to be covered by more cones (i.e., chargers) to make their power vertical angles. Hence, those sensor nodes may need to be covered by more cones (i.e., chargers) to requirements satisfied. On the contrary, ACC greedily and adaptively adjust the direction of every make their power requirements satisfied. On the contrary, ACC greedily and adaptively adjust the cone to cover manycone as possible sensor composition. Consequently, most sensors are direction of as every to cover as nodes many by as vector possible sensor nodes by vector composition. covered by chargers with smaller horizontal and vertical angles. Hence, sensor nodes may need to be Consequently, most sensors are covered by chargers with smaller horizontal and vertical angles. covered by fewer cones (i.e., chargers) to make their power requirements satisfied. The explanation Hence, sensor nodes may need to be covered by fewer cones (i.e., chargers) to make their power above also accounts for the reason why the superiority of ACC is more significant for scenarios of requirements satisfied. The explanation above also accounts for the reason why the superiority of higher power requirements. ACC is more significant for scenarios of higher power requirements.
(a)
(b)
(c) Figure 22. The comparisons of GCC and ACC in terms of the number of deployed chargers for (a) Figure 22. The comparisons of GCC and ACC in terms of the number of deployed chargers for: (a) low; low; (b) medium; and (c) high power requirement. (b) medium; and (c) high power requirement.
Figure 23 shows the comparisons of GCC and ACC in terms of execution time for scenarios of Figure 23 shows the comparisons of GCC and ACC in terms of execution time for scenarios of different power requirements. We can observe that ACC runs much faster than GCC, especially for different power requirements. We can observe that ACC runs much faster than GCC, especially for the scenario of high energy consumption. This is because ACC has O(m2 n3 ) time complexity, while GCC has O(m2 n5 ) time complexity.
Energies 2016, 9, x
19 of 21
2n3) time complexity, while Energies 2016, 9, 696 20 of 22 the scenario of high energy consumption. This is because ACC has O(m 2 5 GCC has O(m n ) time complexity.
(a)
(b)
(c) Figure 23. The comparisons of GCC and ACC in terms of the execution time for: (a) low; (b) medium; and Figure 23. The comparisons of GCC and ACC in terms of the execution time for: (a) low; (b) medium; (c) high power requirement. and (c) high power requirement.
7. Conclusions 7. Conclusions In this paper, we study the WCDO problem concerning with how to deploy as few as possible In this paper, we study the WCDO problem concerning with how to deploy as few as possible chargers to make a WRSN sustainable. We have proposed two heuristic algorithms, namely the GCC chargers to make a WRSN sustainable. We have proposed two heuristic algorithms, namely the GCC algorithm and the ACC algorithm, to provide approximate solutions to the problem. We consider algorithm and the ACC algorithm, to provide approximate solutions to the problem. We consider two forms of energy requirements: coverage requirement and power requirement. For the coverage two forms of energy requirements: coverage requirement and power requirement. For the coverage requirement scenarios, GCC outperforms ACC in terms of the number of chargers deployed. On the requirement scenarios, GCC outperforms ACC in terms of the number of chargers deployed. On the contrary, for the power requirement scenarios, ACC outperforms GCC in terms of the number of contrary, for the power requirement scenarios, ACC outperforms GCC in terms of the number of chargers deployed. However, for both forms of requirements, ACC has shorter execution time than chargers deployed. However, for both forms of requirements, ACC has shorter execution time than GCC, as ACC has O(m22n33) time complexity and GCC has O(m22n5) time complexity, where n is the GCC, as ACC has O(m n ) time complexity and GCC has O(m n5 ) time complexity, where n is the number of sensor nodes and m is the number of grid points. number of sensor nodes and m is the number of grid points. In the proposed GCC and ACC algorithms, the chargers are restricted to be deployed on grid In the proposed GCC and ACC algorithms, the chargers are restricted to be deployed on grid points of a grid, which may limit the application of the proposed algorithms. Moreover, the charger’s points of a grid, which may limit the application of the proposed algorithms. Moreover, the charger’s charging space is modeled as a cone with a specific generatrix length and a specific opening angle, charging space is modeled as a cone with a specific generatrix length and a specific opening angle, which may not be accurate enough for calculating the energy provision offered by the charger. We which may not be accurate enough for calculating the energy provision offered by the charger. We are are now designing more efficient and flexible algorithms, such as meta‐optimization algorithms, to now designing more efficient and flexible algorithms, such as meta-optimization algorithms, to solve solve the problem without the charger deployment restriction or the cone modeling of the charging the problem without the charger deployment restriction or the cone modeling of the charging space.
Energies 2016, 9, 696
21 of 22
Such enhanced algorithms are suitable to WSNs in a closed space or room like space capsule in which sensor nodes have known and fixed positions and energy requirements. Redundant sensor nodes may be applied to such WSNs to prolong the network life time; however, the weight of those redundant nodes may be a burden on the space capsule. The WRSNs that rely on WPT to power sensor nodes are thus suitable for the environment, to which the enhancements of the GCC and ACC algorithms can be applied. We have noticed a related solution [29] using Voronoi diagrams to help solve the directional coverage problem in WSNs whose sensor nodes are associated with rotatable directional sector-shaped sensing areas. However, the solution is for 2D environments and cannot be applied directly for solving the WCDO problem which considers 3D environments. We thus plan to study how to apply 3D Voronoi diagram algorithms to solve the WCDO problem in the future. Acknowledgments: This work was supported in part by the Ministry of Science and Technology (MOST), Taiwan, under Grant Nos. 104-2221-E-008-017-, 105-2221-E-008-078- and 105-2218-E-008-008-. Author Contributions: Ji-Hau Liao conceived, designed and performed the experiments; Jehn-Ruey Jiang and Ji-Hau Liao designed the algorithms, analyzed the data, and wrote the paper. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2.
3.
4.
5. 6. 7.
8.
9. 10. 11. 12. 13.
Xie, L.; Shi, Y.; Hou, Y.T.; Sherali, H.D. Making sensor networks immortal: An energy-renewal approach with wireless power transfer. IEEE/ACM Trans. Netw. 2012, 20, 1748–1761. [CrossRef] Tan, Y.K.; Panda, S.K. Review of Energy Harvesting Technologies for Sustainable Wireless Sensor Network. In Sustainable Wireless Sensor Networks; Seah, W., Tan, Y.K., Eds.; Tech Publishing: Rijeka, Croatia, 2010; pp. 15–43. Kim, S.; Vyas, R.; Bito, J.; Niotaki, K.; Collado, A.; Georgiadis, A.; Tentzeris, M.M. Ambient RF energy-harvesting technologies for self-sustainable standalone wireless sensor platforms. Proc. IEEE 2014, 102, 1649–1666. [CrossRef] Fu, L.; Cheng, P.; Gu, Y.; Chen, J.; He, T. Minimizing charging delay in wireless rechargeable sensor networks. In Proceedings of the IEEE International Conference on Computer Communications (INFOCOM), Turin, Italy, 14–19 April 2013; pp. 2922–2930. He, S.; Chen, J.; Jiang, F.; Yau, D.; Xing, G.; Sun, Y. Energy provisioning in wireless rechargeable sensor networks. IEEE Trans. Mob. Comput. 2013, 12, 1931–1942. [CrossRef] Dai, H.; Chen, G.; Wang, C.; Wang, S.; Wu, X.; Wu, F. Quality of energy provisioning for wireless power transfer. IEEE Trans. Parallel Distrib. Syst. 2015, 26, 527–537. [CrossRef] Xu, W.; Liang, W.; Lin, X.; Mao, G.; Ren, X. Towards perpetual sensor networks via deploying multiple mobile wireless chargers. In Proceedings of the 43rd IEEE International Conference on Parallel Processing (ICPP), Minneapolis, MN, USA, 9–12 September 2014; pp. 80–89. Liang, W.; Xu, W.; Ren, X.; Jia, X.; Lin, X. Maintaining sensor networks perpetually via wireless recharging mobile vehicles. In Proceedings of the IEEE 39th conference on local computer networks (LCN), Edmonton, AB, Canada, 8–11 September 2014; pp. 270–278. Ye, X.; Liang, W. Charging utility maximization in wireless rechargeable sensor networks. Wirel. Netw. 2016. [CrossRef] Zhao, M.; Li, J.; Yang, Y. A framework of joint mobile energy replenishment and data gathering in wireless rechargeable sensor networks. IEEE Trans. Mob. Comput. 2014, 13, 2689–2705. [CrossRef] Han, G.; Qian, A.; Liu, L.; Jiang, J.; Zhu, C. Impacts of traveling paths on energy provisioning for industrial wireless rechargeable sensor networks. Microprocess. Microsyst. 2015, 39, 1271–1278. [CrossRef] Lin, C.; Wu, G.; Obaidat, M.S.; Yu, C.W. Clustering and splitting charging algorithms for large scaled wireless rechargeable sensor networks. J. Syst. Softw. 2016, 113, 381–394. [CrossRef] Lin, C.; Wang, Z.; Han, D.; Wu, Y.; Yu, C.W.; Wu, G. TADP: Enabling temporal and distantial priority scheduling for on-demand charging architecture in wireless rechargeable sensor Networks. J. Syst. Archit. 2016. [CrossRef]
Energies 2016, 9, 696
14.
15.
16.
17.
18.
19.
20.
21. 22. 23. 24.
25. 26. 27. 28.
29.
22 of 22
Maehara, D.; Tran, G.K.; Sakaguchi, K.; Araki, K.; Furukawa, M. Experiments validating the effectiveness of multi-point wireless energy transmission with carrier shift diversity. IEICE Trans. Commun. 2014, 97, 1928–1937. [CrossRef] Maehara, D.; Tran, G.K.; Sakaguchi, K.; Araki, K. Experimental study on battery-less sensor network activated by multi-point wireless energy transmission. 2016. arXiv:1601.00788. arXiv.org e-Print archive. Available online: https://arxiv.org/abs/1601.00788 (accessed on 15 August 2016). Imoto, N.; Yamashita, S.; Ichihara, T.; Yamamoto, K.; Nishio, T.; Morikura, M.; Shinohara, N. Experimental investigation of co-channel and adjacent channel operations of microwave power and IEEE 802.11g data transmissions. IEICE Trans. Commun. 2014, 97, 1835–1842. [CrossRef] Yamashita, S.; Imoto, N.; Ichihara, T.; Yamamoto, K.; Nishio, T.; Morikura, M.; Shinohara, N. Implementation and feasibility study of co-channel operation system of microwave power transmissions to IEEE 802.11-based batteryless sensor. IEICE Trans. Commun. 2014, 97, 1843–1852. [CrossRef] Yoshida, S.; Hasegawa, N.; Kawasaki, S. The aerospace wireless sensor network system compatible with microwave power transmission by time-and frequency-division operations. Wirel. Power Transf. 2015, 2, 3–14. [CrossRef] Ichihara, T.; Mitani, T.; Shinohara, N. Study and development of an intermittent microwave power transmission system for a ZigBee device. In Proceedings of the IEEE Wireless Power Transfer Conference (WPTC), Jeju, Korea, 8–9 May 2014; pp. 40–43. Correia, R.; De Carvalho, N.B.; Fukuday, G.; Miyaji, A.; Kawasaki, S. Backscatter wireless sensor network with WPT capabilities. In Proceedings of the IEEE MTT-S International Microwave Symposium, Phoenix, AZ, USA, 17–22 May 2015; pp. 1–4. Powercast TX91501–3W-ID Wireless Charger User Manual. Available online: http://www.powercastco. com/PDF/TX91501-manual.pdf (accessed on 30 May 2016). Powercast P2110-EVAL-02 Power Harvester User Manual. Available online: http://www.powercastco.com/ PDF/P2110-EVAL-02-manual.pdf (accessed on 30 May 2016). Shaw, J.A. Radiometry and the Friis transmission equation. Am. J. Phys. 2013, 81, 33–37. [CrossRef] Yang, Q.; Nofsinger, A.; McPeek, J.; Phinney, J.; Knuesel, R. A complete solution to the set covering problem. In Proceedings of the International Conference on Scientific Computing (CSC), Las Vegas, NV, USA, 27–30 July 2015; p. 36. Hennessy, J.L.; Patterson, D.A. Computer Architecture: A Quantitative Approach, 5th ed.; Elsevier: Amsterdam, The Netherlands, 2011. Nagelkerke, N.J. A note on a general definition of the coefficient of determination. Biometrika 1991, 78, 691–692. [CrossRef] TI eZ430-RF2500 Development Tool Guide (Rev. F). Available online: http://www.ti.com/lit/pdf/slau227 (accessed on 30 May 2016). Bougard, B.; Catthoor, F.; Daly, D.C.; Chandrakasan, A.; Dehaene, W. Energy efficiency of the IEEE 802.15.4 standard in dense wireless microsensor networks: modeling and improvement perspectives. Des. Autom. Test Eur. 2005, 1, 196–201. Sung, T.W.; Yang, C.S. Voronoi-based coverage improvement approach for wireless directional sensor networks. J. Netw. Comput. Appl. 2014, 39, 202–213. [CrossRef] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).