Communication Antenas for UAVs

Transcription

1 Communication Antenas fo UAVs Pedo Miguel Matins Maques, Militay Academy/IST Abstact Ove the past decades, paticulaly the last, the use of unmanned aeial platfoms known as Unmanned Ai Vehicle (UAV) by the militay has played a key ole in the conduct of opeations, whethe they be at a militay o civilian level. The Potuguese Ai Foce simila to what happens with othe NATO countepats is developing a poject called "Poject fo Reseach and Technology in Unmanned Aeial Vehicles" (PIVANT) with the aim of poviding the Potuguese Ai Foce with the ability of exploitation of these unmanned aeial vehicles, having aleady developed some UAV platfoms unning econnaissance and suveillance missions ove continental and maitime Potuguese teitoy. The need to communicate o impove communication with these platfoms has led to the poposal of the pesent dissetation, whose goal is to design an antenna that enables impoved efficiency and scope of the downlink between the UAV and the gound station. In this egad an Electonically Steeable Paasitic Aay antenna adiato (ESPAR) will be designed and simulated. This antenna allows the vaiation of the diection of maximum gain in azimuth, which will give the opeato the possibility to choose the intended diection of adiation. The antenna will be designed and simulated using the CST MWS pogam, to opeate in the band of.33 GHz. Afte this stage, the antenna will be constucted and implemented in the UAV to cay out tests. This task is pefomed by using the equipment povided by the Potuguese Ai Foce. Keywods: ESPAR, antenna, communication, gain diection, UAV I. INTRODUCTION In cuent opeations theates (OT), infomation plays a key ole in the conduction of militay actions, having supeioity the foce that moe effectively and apidly acquies this infomation, peventing the opponent to eact in a timely manne. One of the technological tools that has inceased the effectiveness of militay opeations is the Unmanned Ai Vehicle (UAV) []. In this egad the use of UAVs in the opeations theate is a tool with inceasingly impotance in the conduct of militay opeations, specially suveillance and econnaissance missions, since it offes timely infomation in any envionment without the use of conventional foces. The Potuguese Amy and the Potuguese Ai Foce ae developing platfoms fo unning such missions. The Amy with the acquisition of suveillance obots unde the ROVIM poject and the Ai Foce with the development of UAVs unde the PIVANT poject. This is the ationale and backgound fo the pesent dissetation which includes theoetical analysis, design and simulation of a diectional plana antenna opeating in the band.33ghz, with the configuation of an Electonically Steeable Paasitic Aay Radiato (ESPAR) antenna. The antenna will be placed in the UAV wing, not only to ensue the connection of data, including video, but also to impove and incease the UAV communication distance with the gound station. The plana antenna ESPAR (P-ESPAR) should allow azimuthal vaiation of the adiation patten of the main lobe. The antenna has to comply with the opeating equiements, namely bandwidth, gain and SWR (standing wave atio) ensuing UAV continuous communication with the gound station. In addition to these opeational featues, the antenna must have educed dimensions and be obust. The simulation of the antenna adiation paametes is pefomed using an appopiate simulation pogam CST Micowave Studio (CST MWS). The design o layout fo futue constuction will be done in 3D CAD Solidwoks design pogam. II. STATE OF THE ART OF THE UAV ANTENNAS A. Communication, Communication means and Fequencies The communication between the gound station (GS) and the UAV is based on two tansmission links, one upwads and the othe downwads, called "up-link" and "down-link". The "up-link" is include the command and contol tansmission fom opeatos to UAVs, including flight tajectoies (stoed on autopilot contolle o Autonomous flight contol system (AFCS), payload and infomation of the cuent position of the gound station when equied. In the descending link, coesponding to the "down-link", is ensued tansmission of data collected as images and videos, the cuent position of the UAV and othe aicaft opeating infomation (fuel, etc.) to the gound station, usually by a diective antenna. In this sense, the maintenance of communications is extemely impotant in the opeations pefomed by UAVs. Shannon [] deepened Nyquist esults and poved that the maximum capacity C of a channel [bits / s] is given by the following equation: C LBg SNR () In which LB is the bandwidth in Hz and ( SNR ) is the signal to noise atio in db. Though the above equation it can be seen that the channel capacity is diectly popotional to the bandwidth, which in tun depends on the fequency, so fo high fequencies the tansmission capacity will be highe, indispensable component fo the UAV to collect and shae infomation with the gound station. In addition, the use of these fequencies equies line of sight between the eceiving and tansmitting antennas [][3]. B. Antenna types used by UAVs The omnidiectional antennas povide a adiation patten in the shape of a 36 donut [4]. This type of antennas used in UAVs ae geneally constituted only by a vetical monopole with a length coesponding to quate of the opeating wavelength ¼. This antenna is vetically polaized and equies a eceiving antenna with simila polaization. The monopole antenna adiation diagam is omnidiectional and the eceived powe deceases apidly with distance [5]. In figue () (a) we may obseve that this type of antenna is nomally placed at the bottom font of the UAV. Consequently to obtain highe gains and moe diective adiation pattens, antenna aay ae used. The individual fields oiginated by the antennas intefee with each othe, ceating maximum and null adiation in specific diections [6] [7](figue ).

2 Cuently the antennas aays moe used in UAVs ae linea aays of dipoles and plana antennas aays. This type of antenna povides a bette SNR, geate diectivity, highe gain and may also have the ability to diect the antenna adiation fo a given desied diection [3]. Fo aeodynamic easons these antennas should be implemented on wings of UAVs, educing ai esistance geneated duing flight due to dag effects, impoving the pefomance of the UAV []. thei ease of constuction, easy adaptation to vaious types of plana o non-plana sufaces and the possibility of suitable adiation pattens []. The need to develop antennas with little weight and size to be easily installed in the UAV wing as well as aeodynamic consideations, makes this type of antenna suitable to pefom communication between UAV and gound station. a) b) c) Figue - Monopole ¼ implemented in UAV; b) UAV model using a linea aay of antennas embedded in the wing stuctue [8] c) Oveview of the antenna embedded in the wing UAV [8] Antennas have been following the tend towads miniatuization and integation with the intended application. In UAVs communication these factos ae of exteme impotance, hence the eason fo the geat applicability of plana antennas in UAVs, which pesent cetain advantages ove the conventional ones, namely: Reduced volume and lighte stuctues; easy to adapt to diffeent sufaces fo mounting, if the substate is flexible; simple and economical to manufactue using pinted cicuit poduction technology; do not distub the aeodynamics of UAVs; easy integation in cicuits; suitable fo use in plana aays. The disadvantages compaed to conventional antennas, ae naow bandwidth, losses in the conducto and dielectic, possible suface waves excitation, educing the efficiency and limited powe. Within the set of plana antennas, the pesent UAVs communication antennas designated by Switched Paasitic Aay Antennas Radiato (SPAR) ae inteesting to achieve communication given the capacity to vay its adiation diagam. Duing the last decade, smat antennas has been studied due to the advantages that they offe in the impovement of wieless communication [4]. These antennas ae called SPAR and have the ability to contol the adiation patten and adjust the main lobe of adiation to a cetain specified diection, and the null of adiation to the intefeing signals. This pocess can be pefomed though eactance contol using vaicap diodes and is denoted by beamfoming [6]. The ESPAR antenna uses mutual coupling to excite thei paasitic elements. Each paasitic antenna element is associated with a vaiable eactance poduced by two vaicap diodes. The vaiation of the value of eactance is eflected in the vaiation of the maximum of antenna adiation, which can thus be diected to the eceive and diect the adiation nulls fo unwanted signals as aleady indicated, which maximizes the signal to noise atio (SNR) of the system [4][9]. III. PLANAR ANTENNAS Plana antennas began to be developed in the 97s mainly fo special applications such as ada, satellite communications, navigation systems, etc []. Cuently, these have wide applicability in communication, because of A. Stuctue and basic chaacteistics of Plana Antennas Conventional plana antennas geneally consist of a metal patch o adiating element typically having a thickness t (whee is the wavelength in fee space) placed on a substate of thickness h, usually with.3 h.5. The substate is placed on a metallic plane, as shown in Fig (). The adiating element and the mass-plane ae made of a vey thin laye of coppe and the dielectic substate that makes the sepaation of the metal plates geneally has a dielectic constant ( ) between., the value of ( ) depending on the [] type of substate. The plana antenna patch can have vaious geometies such as squae, ectangula, cicula, tiangula, etc. The geomety and dimensions of the patch itself detemines the antenna adiation chaacteistics [] []. Figue - Rectangula patch antenna geomety B. Method of tansmission line The analysis of plana antennas is complex due to pesence of non homogeneous dielectics, non homogeneous bounday conditions, excitations divesity and numeous configuations of patch elements. Theefoe, one esots to models o analytical methods to enable a bette undestanding of the opeating pinciples and chaacteistics of the antenna. The full-wave methods geneally ae based on integal equations of the Sommefeld type in the spectal domain and solutions of Maxwell's equations in the time domain. These methods have bette accuacy when compaed with analytical models, but its analysis is complex []. In the analytical method of the tansmission line, the patch element is the esonant element of the tansmission line whee the chaacteistic impedance ( Z ) and popagation c phase constant ( ) ae detemined by the paametes of the antenna dimensions and substate []. The effective dielectic constant ( eff ) vaies with the change of the patch width (W ), howeve this depends on the opeating fequency ( f ), demonstated in (8). At low fequencies it appoaches the dielectic substate constant value ( ). The effective dielectic constant value can be calculated by the following equation (). accoding to the following condition W h []:

3 h eff () W Once the width (W ) and length ( L ) patch ae calculated, popagated electomagnetic fields along the adiating element pesent a discontinuity in the edges of the patch, and, diffacted fields (finging fields) ae geneated at thei ends, as shown in figue (3 a and b). The amount of the finging fields is a function of the dimensions of the patch, the thickness (h) and substate dielectic constant ( ) []. a) b) Figue 3 - a) Plana antenna with ectangula geomety epesentation of finging fields excited by a micostip line b) Side view of electomagnetic fields popagated in substate[] The behavio explained above, causes an electical addition L to the dimensions of the patch on both sides in the diection of the xy plane (main plane E) as shown in figue (4) CN m. Equation (5) does not account fo the edge effect. By taking this into account equation (6) should be modified to []: c fc q (6) L L eff eff Being ( q ) the edge effect facto, calculated as follows[]: fc q (7) f The patch width (W ) is elated to the esonance fequency ( f ) and dielectic constant ( ) and it is calculated by the following equation[]: W f f Neglecting edge effects the patch length is given by the following equation []: L L f eff The electical addition ( L ) oiginated by edge effects at the patch edges (slits), can be epesented by an equivalent admittance ( Y ) constituted by a conductance ( G ) in paallel with a susceptance (B) at each end of the patch[] as shown in figue (5). (8) (9) () () Figue 4 - Physical and affective length of the ectangula patch plana antenna[] This incease in the dimensions of the patch may be epesented by two slots of width ( L ) sepaated by a distance equal to the patch lenght. The length ( L ) can be appoximated accoding to [] expession: W eff.3.64 L.4 h h W eff.58.8 h Thus, the effective length of the patch is given by []: L L L eff Fo the dominant mode TM the esonance fequency ( f ) in the plana patch antenna is a function of lenght ( L ) and is given by [] the expession: C f L L Whee ( C ) is the speed of light in fee space, ( ) the 7 magnetic vacuum pemeability whose value is 4 and ) the pemittivity of vacuum with a value of ( (3) (4) Figue 5 - Equivalent cicuit of tansmission line[] Consideing patch slots numbeed and as shown in figue (4.4), the equivalent admittance of the slot is given by[] : Y G jb () The values of the conductance ( G ) and susceptance ( B ) in slot may be appoximated by []: W G ( kh) 4 Consideing the () condition W B.636ln( kh) h () Consideing the patch slots numbeed and as shown in figue (4), in case the slot is identical to slot, the equivalent admittance is given by []: Y Y, G G, B B The adiation powe ( P ad ) is given by []: P ad kw sin cos V 3 sin d cos (5) (4)

4 and k it is the popagation constant and V he. The electic field fo vey thin thicknesses ( kh ) can be expessed by equation (5) []: kw jk sin cos Ve E j sin (5) cos The conductance may also be witten fom equation (6) []: I G (6) whee I is given by[]: kw sin cos 3 I sin d (7) cos The input esonance admittance can be given by the following expession[]: Yin G (8) if the input admittance ( Y in ) is eal without the coupling effects, it can be expessed as []: Zin Rin (9) Y in G Consideing the coupling effects, the input esonance esistance ( R ) is given by []: in R in () ( G G ) whee G is the mutual conductance between the two slits, given by []: kw sin cos 3 G J ( klsin )sin d cos J is the st Bessel s function of the st kind and zeo ode. The chaacteistic impedance Z in the micostip line can c be given by the following expession[]: 6 8h W W ln, 4 eff W h h () Zc W, W W h eff ln.444 (3) h h consideing Zc 5, the width of the micostip line ( W ) can be calculated by solving the equations () o (3) in ode to W. The input esonance esistance fo a point ( y ) can be appoximated by the expession []: Rin( y y) cos y ( G G ) L Rin( y ) cos y L Solving this last expession in ode to ( y )it is possible to detemine the location of the inset feed point, which coesponds to the distance fom the slot patch to the inset feed point in ode to adapt the impedance of the antenna ( (4) () Zc 5)[]. In (6) the inset feed technique fo impedance matching is shown. Figue 6 -Plana antenna ectangula geomety adapted to the inset feed technique[] C. Excitation Methods The settings that can be used to excite the antenna can be classified into two goups. A diect contact goup and anothe one by an enegy coupling o indiect [8]. The diect way is usually by micostip line o coaxial cable, also known as pobe method and the indiect one by apetue coupling o poximity coupling [8] [3]. The choice of excitation method is a detemining facto fo the adjustment of the antenna. [8] Each of these methods is shown in the figues (7) to (9). The excitation though the micostip line is made though a conductive line with a width much smalle than the patch as shown in figue (3 a). The conductive line is pinted on the same antenna substate plane. This fom of excitation has the following popeties [8] []: it is simple to constuct; easy input impedance adaptation by the inset feed technique; as the micostip line is on the same stuctue of the patch it causes discontinuities in the micostip line that gives ise to deceasing spuious adiation and efficiency of the antenna. It also has, when compaed to othe excitation methods including coupling, a lowe bandwidth. The pobe excitation consist in an intenal connecto (coaxial) connected to the patch though the substate to the gound plane (extenal connecto), usually by a soldeing point. []. This method is epesented in Figue (7). Figue 7 - Plana antenna fed by pobe Its main featues ae [] [] the powe supply is isolated fom the est of the antenna deceasing the spuious adiation, which esults in geate efficiency; fo substates with highe thickness it becomes moe difficult to apply; has a smalle bandwidth compaed to othe methods; the pobe method also allows a simple and simila way as the pevious method to pefom the antenna impedance matching, by placing the antenna excitation point at any location of the patch.

5 This type of excitation uses two layes of substate, placing the patch on top of the stuctue and the micostip line positioned between the two dielectic []. Usually the lowe substate possesses a highe dielectic constant ( ) than the substate placed on top. The antennas fed by this method have a geate bandwidth than the antennas fed by the afoementioned methods given the inceased thickness of the substate. Howeve, the powe supply is not fully isolated, thee is still spuious adiation, but to a lesse extent than with the pevious methods. Figue 8 - Plana antenna fed by Apetue-coupled The apetue coupling technique is simila to poximity coupling technique, it consists of two sepaated substates though a gound plane. At the bottom of the substate positioned below, thee is a micostip line whose enegy is coupled to the patch though a slot in the gound plane sepaating the two substates, this slot is usually centeed below the patch. The bottom substate usually has a dielectic constant ( ) highe than uppe substate in ode to educe unwanted adiation.the gound plane between substates isolates the powe to the patch minimizing intefeence fom spuious adiation []. In (9) this kind of excitation method is epesented: Figue 9 - Plana antenna fed by Apetue-coupled Subsequently the antenna is dimensioned theoetically with the above mathematical expessions using the tansmission line method. This phase is impotant to analyze the antenna behavio with the change of its dimensions. The antenna will be simulated and optimized by CST Micowave Studio pogam. The CST MWS pogam is an electomagnetic simulato based on full wave methods (EM) with high pefomance. It integates simulation, visualization, modeling and automation in an easy envionment used to solve 3D EM poblems accuately. Afte the theoetical dimensions wee calculated, the plana antenna was dawn using the MWS CST pogam and was an optimization tool designated by Optimaze Tansient paamete, which makes multiple scans of the paametes that constitute the dimensions of the patch in ode to impove the fit and pefomance o the antenna, was used to obtain a low value of the standing wave atiomodule fo.33 GHz. Within the optimized paametes it was found that the most impotant paamete fo the antenna behavio is the patch length (L), as can be obseved in figue (). Figue - Gaph of change of S a function of fequency fo diffeent db values of L vesus fequency As can be seen, the vaiation of the paamete L dastically changes the behavio of the antenna. Note that fo highe values of L, the esonance fequency deceases consideably. This is in accodance with equation (9) whee L is invesely popotional to the esonance fequency. In figue () is shown the dawing in CST MWS, of the optimized plana antenna with only one element: IV. PROPOSAL PLANAR ANTENNA ESPAR Afte the theoetical explanation of plana antennas and defined the stuctue and method of excitation, becomes necessay to conduct the study and simulation of the poposed antenna. This chapte aims to study, simulate and develop the plana antenna ESPAR called P-ESPAR. Opeational equisites of plana ESPAR antenna that must meet the descending UAV communication conditions: fequency of esonance (.33 GHz); band width geate than o equal (8.7 MHz) and stationay wave atio (< - db). Figue - Repesentation of antenna in CST MWS with (W=87.4. L=74,W=.3, y=6., x=4, Lf=6) all dimensions in mm. A. The plana antenna element configuation (patch) The configuation of the plana antenna, began with the analysis of vaious dielectic substates in ode to choose the most suitable fo the poject. The substate chosen fo the poject was the RT Duoid (587), with (.33., tan.5 a. and a thickness h.575mm ).

6 In figue () it is found that the antenna is fully adapted to the desied fequency band f.334 GHz. As mentioned above, the adaptation level is epesented by the stationay wave atio module, S. db Figue 9 - Gaphical epesentation of S a function of fequency fo db dimensions of the plana antenna single-patch optimized Fo this case it was obtained S 9.94 the citeion poposed initially S db db satisfying. The plana antenna with a single patch has a bandwidth of (MHz 5.4), which complies with equisites. In (3) we can see the adiation chaacteistics in the plane E (phi = ) and H (theta = 9). The antenna has a gain of 5.44 db a beam width at half powe bandwidth (HPBW) (-3 db) of 95.4, in plane E and in the H plane is about 6.. The vaicap diodes intoduce capacitive eactances that vay, as the evese bias voltage is applied to the teminals of the diodes. This phenomenon allows fo the adjustment of mutual coupling between the patches, which esults in a change of the effective length (L) of the antenna, o the effective length of the two paasites patches. This change is manifested in the vaiation of the amplitude and cuent phase of the paasitic elements in elation to the active element which in tun vaies the slope of the main beam [3]. The diodes vaicap BB833 wee chosen fo simulation puposes. Theefoe, analyzing the behavio of the antenna with the addition of vaicap diodes is essential to undestand how the paametes of the antenna vay. In this sense it is pesented a simple equivalent cicuit model that helps to analyze this behavio. The equivalent cicuit model of the plana antenna without diodes may be epesented by an RLC equivalent cicuit in paallel, as illustated in Figue (4)[4]. To detemine the values of L and C, full wave simulation of electomagnetic fields by the CST MWS pogam was used. L and C epesent the magnetic and electic enegy stoed in the esonant cicuit. The capacity of the vaicap diode is epesented as a combination of capabilities in paallel of the two diodes positioned between the patches ( C T ) as shown in FIG. (4) Figue - Equivalent cicuit of plana antenna with vaicap diodes Figue - Pola epesentation of the adiation patten of the antenna single-patch optimized in the plane E and H V. P-ESPAR ANTENNA An antenna aay may be constituted by two o moe patches in ode to impove cetain desied chaacteistics, such as bandwidth, diectivity and gain. These chaacteistics can not be impoved with only a single element. In [8] it is mentioned that the antenna aay s adiated powe can be concentated in a smalle aea, which tanslates into an incease of antenna diectivity. Besides this paamete is the gain and half powe beam width ae also impoved. Howeve, the aay of antennas alone is not able to pefom the specifics of the oiginally poposed antenna, this means that a econfiguable antenna capable of changing the diection of the main beam. The antenna P-ESPAR consists of a cluste of thee patch elements, the active element coesponds to the cental patch, excited by a micostip line simila to the pevious antenna with only one element. The two othe patches, designated as paasitic patches, have dimensions identical to the active element and ae disposed on each side of the cental element. The mutual coupling between the patches is contolled using fou vaicap diodes. The esonant fequency ( f ) without the diode, that is, the unloaded cicuit is given by equation (5) and occus at esonance [4]: f (5) LC As a esult of loading the capacity of the vaicap diodes, the cicuit capacity becomes the combination of C paallel to C T. In this case the esonance fequency ( f c ) of the cicuit is given by [4]: fc (6) L C C Since the value of C T is known and f c can easily be detemined by a simulation of S using CST MWS the value of C can be calculated by [4]: CT C (7) f f c T

7 The P ESPAR antenna configuation takes into account vaious citical paametes as the choice of substate, its dimensions, the excitation method, the selection of the vaicap diodes, the capacitive eactance adjustment necessay fo the equied specifications, and the position of coupling of the diodes. This design pocess is depicted in figue (5) appoaches the desied.33ghz with inceasing values. Only afte choosing the coupling position, the antenna optimization P-ESPAR, was completed. Figue (7 a and b) epesent the final design of the poposed antenna seen fom above and side ways, espectively. The images coespond to the design made in the CST MWS pogam. a) b) Figue 5 a) Pocess study and development of P-ESPAR antenna The final stuctue of the antenna consists of an aay of thee identical patches positioned side by side, the space between the patches coesponding to the length of the vaicap diode G =.5 mm. The cental patch is the active element and the side to this paasitic elements. The active element is excited by Wave pot tool of CST Micowave studio pogam aleady used in the antenna with a single element. The antenna substate is RT Duoid 578 having a thickness of h =.575 mm. The pocess of detemining the dimensions of the P- ESPAR antenna elements was conducted similaly to the one fo the sigle patch, using optimization tool of CST MWS pogam designated Optimaze Tansient paamete. As the mutual coupling position (O) is essential in balancing all the paametes of the antenna, mainly the esonance fequency, bandwidth and adaptation, seveal simulations wee pefomed paticulaly of S. It was db veified that fo smalle poximity of the vaicap diodes the level of vaiation of the main paametes was smalle and smalle as the capacitive eactances of the diodes wee changed as shown in figue (6). This facto is impotant fo maintaining the opeational equiements of the antenna. Figue - Gaphical epesentation of O S vaiation fo diffeent values db Afte seveal simulations it was found that the best equilibium was achieved with O = 3 mm. Although the figue (6) takes into account only the capacity C=pF note that the esonance fequency Figue 3 - a)stuctue and dimensions of P-ESPAR antenna in pespective view fom above b) P-ESPAR antenna on the side pespective C. Simulation P-ESPAR Antenna The antenna design shown in figue (7 a and b) was simulated using CST MWS pogam fo the fequency of.33ghz. Following the pocess shown above in figue (6). The values of the capacities of the vaicap diodes wee set afte numeous simulations in ode to find the ideal chaacteistics that fulfill the opeational equiements fo each situation of configuation of the adiation beam. The azimuths chose wee: -5, -, -5, -, -5, 5,, 5, and 5 in elation to its efeence azimuth ( ). The vaiable capacity (C T) coesponding to all the junction capacities of the two vaicap diodes in the left position of the cental patch while (C T) coesponds to the set of the junction capacities of diodes positioned on the ight side of the cental patch, as illustated in figue (7 a). To an identical mutual coupling in each of the paasitic patches it has always been consideed the same value in the two vaiable capacity diodes which constitute (C T) o (C T) to avoid asymmeties in the mutual coupling. The nine simulated adiation settings ae shown in table (). Tabel - Antenna chaacteistics fo each P-ESPAR adiation setting P-ESPAR Antenna Chaacteistics Main Lobe [ ] Ganho f c S LB C T C T [dbi] [GHz] [db] [MHz] [pf] [pf] Analyzing the table it appeas that thee is symmety in the values of the vaicaps fo each adiation beam configuation (beamfoming). Fo example, fo -5

8 diection of the main lobe is C T=.7 pf and C T=.4 pf, wheeas to 5 the values of the vaiable capacitos ae switched. This phenomenon occus due to the existence of symmety in the patches dimensions. In figues (8) and (9) simulations ae pesented simultaneously to S vesus fequency and the adiation db pattens coesponding to the uppe and lowe diection of the main lobe of the antenna. Figue (9) shows that thee ae geat fluctuations in the esonance fequency, fo coesponds to f c=.98 GHz, while that fo the maximum and minimum values achieved in the vaiation of the main lobe (-5 and 5 ) f c=.3 GHz, existing consistency of opeational equiements. a) b) c) d) e) Figue 8 - Gaphical epesentation adiation patten in the plane and to -5, and 5 Now, all the simulations of S and adiation diagams db coesponding to the adiation beam settings achieved by the P-ESPAR antenna will be pesented. f) g) Figue 9 - Gaphical epesentation of S a function of fequency fo db vaious configuations adiation Looking at the chats S db of figues (4) and (9) it is found that the esonance fequency emains vitually unchanged and close to the value of.33 GHz. The bandwidth of all the antenna adiation pattens of vaying configuations of the P-ESPAR antenna meet the 8.7 MHz. Figue (9) is the gaph of S as a function of fequency fo the multiple adiation beam configuations. Obseving the gaph of figue (8) it is found that the cuve which descibes a given beam diection is identical to the cuve that descibes the symmetical diection, this is, thee is symmety in mutual coupling. This is due to the symmety of the patches, cited above. Next, ae pesented the adiation pattens fo each configuation of the adiation of P-ESPAR antenna ae pesent. db Figue 4 - Gaphical epesentation adiation patten in the plane E to a) b) 5 and c) -5) d) e) - f) 5 g) -5

9 a) b) Figue - Repesentation of the font configuation of P-ESPAR antenna c) d) In (3) the layout side of the P-ESPAR antenna as shown. Due to the scale in the figue (3) the gound plane and the patches can not be seen, theefoe detailed views A and B on a scale of 5: ae povided. The detail A is the scale expansion at point A, noting that item 3 indicates the patch and item is the dielectic substate. While the detail B is the scale expansion of point B, whee item indicates the dielectic substate as in the detail and the item the gound plane. Figue - Gaphical epesentation adiation patten in the plane E to a) b) - c) 5 d) -5 Though the adiation pattens epesented by figues () and () it can be veified that the chaacteistic vaiation of the diection of the main lobe of the P-ESPAR antenna was eached. In the diection 5 and -5 coesponding to a maximum gain of 6.7 db to 6.79dB, espectively, wheeas, in the diection 5-5 its gain coesponds to only 3.6 db and 3.7 db. The level of secunday lobes lies on an aveage value of - db. It is to be noted that in pactice the P-ESPAR antenna nulls will not emit adiation in the ea diection due to the placement of the antenna on the aicaft wing. Figue 5 - Repesentation of the side configuation of P-ESPAR antenna with its details A and B In (4) is epesentes the ea layout of P-ESPAR antenna constituted by holes also obseved in figues () and (4). It must be noted that the holes contain a cicle centeed at each cental point of the holes diamete of 3 mm. In this place aound the holes thee is no gound plane in ode to avoid intefeence fom wies with the gound plane. D. Simulation P-ESPAR Antenna This chapte will pesent the P-ESPAR antenna design, which compises the font, side and ea view of the antenna with the espective dimensions. The poject was caied out though the 3D CAD design pogam called Solidwoks. The points that define the antenna in the CST MWS pogam wee impoted into the Solidwoks pogam facilitating the design of the antenna. The dimensions ae pesented on a millimete scale. In addition to the P-ESPAR antenna, an UAV wing pofile was also designed in Solidwoks. In () is shown the layout font of the P-ESPAR antenna. As shown in the figue () thee ae 8 holes of mm diamete. These holes indicate the passage of conductos needed fo invesely polaizing the diode vaicap. The diodes ae placed between the holes in ode to pevent any futue intefeence between the wies, that is, consideing a diode vaicap the bottom hole to this, connects to its cathode, while the uppe hole connects to the anode. Figue 6 - Repesentation of ea configuation of P-ESPAR antenna

10 In Figue (5) an image of the UAV wing with the P- ESPAR antenna is pesented, This figue was made using the SolidWoks softwae. Figue 7 - Photogaphy pefomed by SolidWoks pogam in a full scale epesenting the wing UAV with P-ESPAR antenna placed. VI. CONCLUSIONS Militaily, the use of UAVs is a tool with an inceasing impotance to pefom econnaissance and suveillance missions in dangeous envionments with high isk fo human life. In this context, the development of an antenna that impoves the UAV communication efficiency with the gound station was pefomed. The main advantage of plana antennas is the simplicity of constuction, and insetion in seveal sufaces, and the ability to give adequate adiation pattens. Howeve, these antennas have low efficiency and naow band width. The tansmission line method was used to obtain a fist appoximation of the dimensions of the patch stuctue, as well as how to match the antenna to the desied impedance. The full-wave methods wee used with the CST MWS simulation pogam. The study began with an analysis of a single plana antenna. The choice of substate is a key paamete in the antenna adiation chaacteistics, that is, factos such as the dielectic constant, the thickness of the substate and the tangent of loss angle should be consideed in the desied chaacteistics. Afte this step begins the development of an aay of patch antennas designated by P-ESPAR antenna. Howeve the antenna aay alone is not able to cay out the specifications of the initially poposed antenna, this means that it is necessay a econfigue the antenna. Antenna configuation of P-ESPAR takes into account vaious citical paametes as the choice of substate, its dimensions, the excitation method, the selection of the vaicap diodes, the capacitive eactance adjustment necessay fo the equied specifications, and the position of the mutual coupling diodes. The vaiation chaacteistics of the adiation patten of the antenna P-ESPAR that meets the opeational equiements such as esonant fequency, and bandwidth ae only achieved by adjusting the vaiable capabilities of vaicap diodes. These values wee adjusted afte seveal simulations in ode to find the ideal chaacteistics that fulfill the objectives fo each configuation state of the adiation beam. The antenna P- ESPAR antenna azimuthal vaiations wee -5, -, -5, -, -5, 5,, 5, and 5 in elation to its efeence azimuth ( ). It was veified that the main lobe of adiation in the positive diection is symmetical to the main lobe of adiation in the negative diection as equied. As a conclusion fo the ealization of this maste's thesis seveal key studies to meet the poposed objectives wee pefomed. This wok involved the use of seveal simulation tools such as the CST MWS pogam, essential fo the design and optimization of the P-ESPAR antenna, as well as fo undestanding the antenna behavio duing its development. The design pogam Solidwoks 3D CAD assisted in poject design with the layout of the P-ESPAR antenna. REFERENCES [] M. Nogueia de Sousa, Uso de Veículos Aéeos Não Tipulados no Sistema Tático de Guea Eletônica (SITAGE), p., 8. [] C. J. Oliveia Ribeio, AS OPERAÇÕES MILITARES NA ERA DA INFORMAÇÃO E DA COMUNICAÇÃO, pp. 9 3, 5. [3] US Depatment of Defense, Unmanned Aicaft Systems Roadmap 5-3, p. 3, 5. [4] J. A. Mogado and J. T. Boges de Sousa, O PROGRAMA DE INVESTIGAÇÃO E TECNOLOGIA EM VEÍCULOS AÉREOS AUTÓNOMOS NÃO-TRIPULADOS DA ACADEMIA DA FORÇA AÉREA, p. 6, 5. [5] J. P. Mogado, Cento de Investigação da Academia da Foça Aéea: Desenvolvimento & Inovação na áea dos Sistemas Aéeos Autónomos Não-Tipulados, Cid. e Def., pp. 6, 5. [6] R. Austin, Unmanned Aicaft Systems.. [7] G. B. Ronconi, T. J. Batista, and V. Meola, THE UTILIZATION OF UNMANNED AERIAL VEHICLES ( UAV ) FOR MILITARY ACTION IN FOREIGN AIRSPACE, p. 44, 4. [8] L. Sun, B. Sun, J. Yuan, W. Tang, and H. Wu, Low Pofile, Quasi-Omnidiectional, Substate Integated Waveguide (SIW) Multi-Hon Antenna, IEEE Antennas Wiel. Popag. Lett., vol. 5, no. c, p., 5. [9] D. T. I. Cente, UAV Requiements and Design Consideation, p. 9,. [] J. Leland and I. Poche III, Futue Amy Bandwidth Needs and Capabilities. RAND Copoation, 4. [] C. A. Balanis, Antenna Theoy Analysis and Design, 3d Editio. 5. [] R. Gag, P. Bhatia, I. Bahl, and A. Ittipiboon, Micostip Antenna Design Handbook. Boston: Atech House, INC. [3] S. Zhang, G. H. Huff, C. Cung, and J. T. Benhad, Thee vaiations of a patten econfiguable micostip paasitic aay, Micow. Opt. Technol. Lett., pp , 5. [4] J. Luthe, MICROSTRIP PATCH ELECTRICALLY STEERABLE PARASITIC ARRAY RADIATORS, Univesity of Cental Floida, 3. Pedo Maques was bon in Guada, Potugal on Setembe 4, 99. In he joined the Potuguese Amy whee he completed the degee in Telecommunications at the Militay Academy, in Lisbon. At same time, he is a Maste Student in the Electical and Compute Engineeing Maste couse at Instituto Supeio Técnico, Lisbon.