OTA MEASUREMENTS OF SMALL RADIO TERMINALS BY A SPHEROIDAL COUPLER

Transcription

1 EUROPEAN COOPERATION IN THE FIELD OF SCIENTIFIC AND TECHNICAL RESEARCH EURO-COST SOURCE: Tokyo Institute of Technology, Japan COST1 TD(1)16 Bologna, Italy 1/Nov/3-5 OTA MEASUREMENTS OF SMALL RADIO TERMINALS BY A SPHEROIDAL COUPLER Jun-ichi Takada 1, Tasuku Teshiogi, Takashi Kawamua, Aya Yamamoto, Tou Sakuma, Yasuhiko Nago 1 Gaduate School of Science and Engineeing, Tokyo Institute of Technology Box S6-4, -1-1, O-okayama, Meguo-ku, Tokyo , JAPAN Phone/Fax: takada@ide.titech.ac.jp R&D Cente, Anitsu Copoation 5-1-1, Onna, Atsugi-shi, Kanagawa, Japan

2 OTA Measuements of Small Radio Teminals by a Spheoidal Couple Jun-ichi Takada*, Tasuku Teshiogi**, Takashi Kawamua**, Aya Yamamoto**, Tou Sakuma**, Yasuhiko Nago** * Tokyo Institute of Technology ** R&D Cente, Anitsu Copoation ** 5-1-1, ONNA, ATSUGI-SHI, KANAGAWA, JAPAN Tasuku.Teshiogi@anitsu.com 1. Intoduction Recently much attentions have been paid to OTA (Ove The Ai) measuements fo small adio teminals, such as cellula phones, RFID o wieless LAN devices [1][][3]. Peviously we poposed a spheoidal couple (SC) as a novel OTA measuement tool [4][5]. It povides a low-cost and apid measuement system fo TRP (Total Radiated Powe) and TRS (Total Radiated Sensitivity) of cellula phones based on deteministic manne not on statistical manne such as a evebeation chambe. We also intoduced displacement method as a key technology fo TRP measuement, togethe with development esults of a pototype model. In this pape, we popose an estimation method of EUT s eflection coefficient that impoves accuacy and eliability of the displacement method. The measued TRP of seveal EUTs by the SC ae compaed with othe conventional ones, i.e. 3D-integation methods and a evebeation chambe method. The SC is also useful fo TRS measuement. We will descibe the theoy and the method, and measued esults fo actual UMTS teminals using the SC and a commecial adio teste. Since the displacement method compensates a impedance mismatch of EUT, the measued TRP coesponds to the tansmit powe deliveed to Tx antenna. In the cuent adio egulations, license tests of adio equipments ae conducted by the tansmit powe, thus the TRP measued by the SC is well matched to the adio egulations. On the othe hand, we ae often equied to measue TRP of EUT including impedance mismatch. Fo this pupose, we ae developing a new method based on anothe opeational mode of the SC, namely multiple eflection suppession mode. Basic studies about this method will be also intoduced.. TRP Measuements.1 Displacement method In a spheoidal couple, lage multiple eflections occu between Tx and Rx antennas. To measue TRP (Total Radiated Powe) of EUT (Equipment Unde Test) in such multiple eflection envionment, peviously we poposed displacement method as shown in Fig.1. We assume a lossless spheoidal couple in which the Tx and Rx antennas ae placed aound the two foci, F1 and F. The powe P o is supplied to the Tx antenna. The system is consideed to be a two-pot cicuit in which the fist pot is the Tx input pot and the second is the Rx output pot whee a matched load is connected. In the steady state, because some pat of the input powe P o is eflected and the esidual powe becomes the eceived powe P L at the second pot, P L is expessed as: P L P 1 S 11 (1) o whee is the adiation efficiency of the Tx antenna. Consequently, P o denotes the TRP when the Tx antenna is pefectly matched. By displacing the positions of the Tx and Rx antennas fom the

3 focal points along the z-axis, we can vay the coupling between them and thus, we can find the optimal positions whee S11 becomes sufficiently small i.e. the pefect coupling is ealized. Usually, the Tx and Rx antennas ae displaced symmetically by z fom the focal points in the inside o outside diections. At the optimal position, P L becomes P () L P o This means that eceived powe expess the TRP. Figues shows a simulation esult to veify the effectiveness of the displacement method. The simulation model assumes a spheoid whee a = 3 mm and e =.5, in which the Tx and Rx antennas ae half-wavelength dipoles esonant at 3 GHz. Both dipoles ae aanged co-linealy along the z-axis. z expesses the length of symmetical displacement fom the foci, + means outwad and - means inwad. The eflection and tansmission coefficients ae shown when both antennas ae displaced symmetically at 3 GHz. We can see that S11 is less than 15 db and S1 is almost db when z is +7 mm and 5 mm. By applying this method to all fequencies, we can obtain the maximum S1, namely db, ove the equied fequency band. Futhemoe, a SC applying the displacement method povides absolute measuements of antenna adiation efficiency. At the optimal position, o pefectly matched positions, input impedance of Tx antenna is almost the same as that in fee space, while at the pefectly eflected positions, since adiation is suppessed, the input esistance becomes the inne esistance of the antenna. Based on Wheele cap method, we can obtain a adiation efficiency of Tx antenna by measuing input esistances at the pefectly matched and pefectly eflected positions as shown in Fig.3. x ( y) z Multiple eflection P o S 11 z z S 1 Z L P L Magnitude [db] a Fig. 1 Pinciple of displacement method S 1 S z [mm] Fig. Pinciple of displacement method Pefect matching A B Pefect eflection S paametes [db] S 11 S z [mm] Fig. 3 Reflection and tansmission vesus symmetical displacement (.4GHz)

4 . Estimation of eflection coefficient of EUT In the displacement method, theoetical basis whethe the maximum eceived powe might always coincide with TRP is not clea, paticulaly fo a SC with limited size. In ode to solve this poblem and impove accuacy of the displacement method, we developed an estimation method of EUT s eflection coefficient in the SC by applying phase otation technique. The pinciple of this technique is depicted in Fig. 4. If we give continuous phase shifts to one of the two input signals by using a vaiable phase shifte (VPS), one is known and the othe is unknown, we have a ippled output signal. Fom the atio of the minimum and the maximum output signals,, we can detemine the amplitude of the unknown signal. Fig. 5 shows a configuation of a TRP measuement system employing phase otation technique. In the spheoidal couple, the EUT and the eceiving antenna ae aanged, and the output signal by the eceiving antenna is connected to a eceive, such as a spectum analyze, though the connecting cable and the VPS. A fixed eflecto poviding a fixed eflection is inseted between the VPS and the eceive. As VPS, we used fou coaxial cable line stetches which ae seially connected. B A C A B C [dbm] [dbm] [db] 5 B A B B A A Fig.4 Pinciple of phase otation technique L 1 [mm] Line length of VPS (mm) We denote the tansmitted powe of EUT by P, the antenna adiation efficiency by, the input eflection coefficient of EUT by EUT, the output eflection coefficient of the spheoidal couple by, the total tansmission coefficient of the cable and the VPS by K, phase shit by and the eflection coefficient of the fixed eflecto seen fom pot B by. The S-matix of the spheoidal couple can be witten as SC S 11 C1 EUT SC1 S C (3) SC S 1 C S C 1 and the S-matices of the cicuit joining the connecting cable and VPS, and the fixed eflecto ae expessed as j Ke S P j (4) Ke and S S 11 1 S 1 S (5) S S 1 S S 1 Since the system in Fig. consists of cascade-connected thee potions, the SC, the connecting cable and the VPS, and the fixed eflecto, we can deive a tansmission coefficient S 1 fo the oveall

5 system. In this case, the pot 1and the pot ae the input pot of the EUT and the output pot of the fixed eflecto. Afte matix analysis, S 1 becomes S 1 K SC 1 S 1 j 1 K e By vaying the phase of the VPS, we obtain a ippled eceived signal, and then a atio of the minimum and maximum eceived amplitudes. By using, we finally obtain the output eflection coefficient of the SC as 1 (7) K 1 Since the SC is a high Q and low loss system, the following elation holds. EUT (8) Theefoe, TRP of the EUT can be obtained by the measued eceived powe P R (EUT) and as PR EUT TRP P C (9) 1 EUT whee C is an unknown constant of the SC, expessing common losses such as the spheoidal eflecto and othes. C can be eliminated by a calibation with a efeence system equipped by a efeence tansmitting antenna and a signal geneato. (6) Cable Fixed Reflecto (Loss:K) A1 B C3 EUT (P,, EUT ) Rx. antenna L 1 SA Receive Spheoidal couple Vaiable phase shifte Fig.5 TRP measuement system using phase otation technique

6 .3 Developed system and expeimental esults We developed an automatic TRP measuement system fo adio teminals at UHF band. The oveall configuation is shown in Fig. 6. The system compute contols the displacement of EUT and the eceiving antenna, the length of the VPS, and data acquisition and pocessing. In Table 1, majo paametes of the system ae shown. The pictue in Fig. 7 shows an inne view of the couple when the top eflecto is emoved whee we can see an EUT and a eceiving sleeve antenna aanged co-linealy. Fig.8 shows a compaison of the measued S11 and S in the SC. The measuements wee caied out fo two diffeent antenna models. One has a matched and the othe has a mismatched (VSWR=3) monopole antenna mounted on a metallic body which has the same size with 1.47GHz EUT as shown in Fig.7. Fig.8 (a) and (b) show eflection coefficients fo a matched and a mismatched antennas, espectively. Fom the measuements, we see that S11 and S agee vey well fo evey displacement positions fo both cases. The esults gives theoetical basis to TRP measuement using a SC as descibed peviously. Fig. 9 shows a diectly measued and estimated s, output eflection coefficients in the SC. Tx. and Rx. antennas ae sleeve antennas which ae co-linealy aanged and displaced to vay The estimated wee obtained fom the measuements using phase otation technique (PRT) fo two diffeent fixed eflectos, one is =-1.55 db, and the othe is =-1. db. Fom this figue, we can say that the estimated s ae vey close each othe in spite of diffeent, and they agee with the diect measuement. Finally we descibe TRP measuement. The EUT is a test tansmitte geneating 1.47GHz CW signal with 1.8 dbm output powe. The powe is deliveed to a monopole antenna mounted on the metallic body as shown in Fig. 7. Fig. 1 shows the TRPs obtained by the displacement method only (solid line) and the eflection compensation method (dotted line). The system configuation is same as Fig.5. In the displacement measuement, the line length of the VPS was set to be the shotest, and the to be zeo, while in the eflection estimation, was set to be db. Fom the figue, we can see that even if at the positions whee pefect coupling ae not achieved, i.e. the eceived powe is not maximum, the almost same TRPs ae obtained by eflection compensation. In Table, the paametes and the measued values ae summaized. EUT Rx. antenna Displacement Moto Displacement Moto VPS Spheoidal Couple SA System Contol & Data Pocessing Fig.6 Automatic TRP measuement system

7 Table 1: Stuctual paametes of the SC system Diamete of majo axis (a) 1 mm Diamete of mino axis (b) 194 mm Eccenticity (e).41 Spheoidal eflecto Mateial FRP Inne wall conductive painting Resin with 45% Cu Fig.7 Inne view of the SC Range of displacement (Tx, Rx, each) Range of vaiable line length of VPS 3 mm 3 mm (75 mm 4) [db] -1 [db] -1 - S 11 S - S 11 S -1 1 z[mm] Tx : EUT 1.47 GHz Rx : sleeve antenna -1 1 z[mm] Tx : EUT 1.47 GHz (VSWR=3) Rx : sleeve antenna (a) Matched monopole (b) Mismatched monopole (VSWR=3) Fig.8 Compaisons of S 11 and S in the SC [db] PRT( = db) PRT( = -1. db) Diect measuement z[mm] fequency:1.47ghz Tx. and Rx. antennas:sleeve antennas Fig.9 Output Reflection coefficients of SC

8 1 H G F E D C B A Table Reflection compensated TRP (P(EUT):Received powe by displacement method) TRP [dbm] z[mm] Displacement method Displacement method + PRT Fig.1 TRPs by the displacement method and the eflection compensation method Z [mm] P(EUT) [dbm] K' [db] Γ [db] TRP [dbm] A B C D E F G H = -1.5dB.4 Round obin tests of TRP To veify the validity of TRP measuement by the SC, we conducted ound-obin tests in which the same EUTs ae measued by the own facilities of the diffeent oganizations. The paticipated oganizations and the facilities used ae listed in Table 3. The positione of TELEC (Telecom Engineeing Cente) has a metallic disc eflecto ove which an EUT is mounted with a space, and a pobe scans ove patial sphee (=(zenith) to typically 1 degees) [6]. The evebeation chambe of NICT (National Institute of Infomation and Communications Technology) is a athe lage chambe (5.5m x 4.5m x3m) with fou sties [7]. As EUTs, basic antennas, a dipole and a sleeve, and fou kinds of adio tansmittes wee used. To the basic antennas, powe of 1dBm was supplied by a signal geneato. The fou tansmittes, EUT(1) to EUT(4), have diffeent antennas o diffeent mounting positions as shown in Table 4. The EUTs in which monopole antennas ae mounted at the edge o the cone of the metallic body geneate coss polaization. Fig. 11 (a) and (b) show the oveall measued esults at 1.47 GHz and.41 GHz, espectively, whee P is tansmit powe deliveed to Tx antenna and PR is TRP. In 3D-scanning measuements achieved in anechoic chambe, thee is some diffeence (about 1dB) between Anitsu and TELEC at 1.47GHz, but at.4ghz the diffeence is small. This means accuate measuements ae not easy even in anechoic chambe. The data obtained by the evebeation chambe exhibits slightly highe TRP than othes. We can see that the SC gives TRP close to P fo all cases. Fo EUT(4) with mismatched antenna, all method excepting the SC exhibit educed TRP, while the SC gives high TRP because it compensates mismatch due to the displacement as descibed peviously.

9 ][dbppr/ipoleeevesl1.t(1)eu47ght()eut(3)eueut(4)db]p[pr/ipoleveslee.4(1)eut1gh()eut(3)eut(4)euttable 3 Test sites Facility Method of measuement Oganization Remaks Anechoic chambe 3D-Scanning (3m method) Anitsu Anechoic chambe 3D-Scanning Patial spheical scan TELEC (1m spheical positione) with a gound plane Revebeation 5.5m 4.5m 3m Random field method NICT chambe Fou sties Spheoidal couple Deteministic method Anitsu Majo axis = 1.m Eccenticity =.4 Table 4 Tested EUT Antenna Position EUT(1) Matched monopole Cente EUT() Matched monopole Edge EUT(3) Matched monopole Cone EUT(4) Mismatched monopole (VSWR=3) z3d-scan(anitsu) - z3d-scan(anitsu) D-Scan(TELEC) Rev.Chambe(NICT) SC(Anitsu) DCenteD D-Scan(TELEC) Rev.Chambe(NICT) SC(Anitsu) (a) 1.47 GHz (b).41 GHz Fig.11 Oveall esults of the ound obin tests

10 3. TRS Measuements 3.1 Theoy and measuement method To chaacteize the sensitivity of a eceive, conventional conducted measuements have been used. The conducted sensitivity is defined by the input powe to the eceive Ps that achieves the pedetemined theshold of BER (bit eo ate) as shown in Fig. 1. On the othe hand, in OTA measuement, sensitivity is expessed in tems of TRS (Total adiated sensitivity). Denoting the powe of incoming adio waves fom ( diection with -polaization that achieve the theshold BER by EIS (), EIS is witten by PS EIS, (1) G, whee G () expesses opeating gain of the Rx antenna. Similaly fo -polaization, EIS is given by PS EIS, (11) G, TRS is defined as 4 TRS (1) 1 1 sindd EIS, EIS, The denominato is ewitten as 1 G, G, sindd (13) P S Using the elations of the opeating gain and the diective gain, the equation (13) becomes L G, 4 L m m GD D d d P, sin (14) S PS whee is antenna efficiency, L m is mismatch loss, and G D () and G D ()ae diectional gains. Then TRS is obtained as PS TRS (15) Lm The SC is vey suitable fo measuing with a simple facility and in a shot time. If an EUT is well matched, o L m can be measued by any means, we can ealize a simple and apid TRS measuement. The configuation of TRS measuement using the SC is shown in Fig. 13. The measuement pocedues ae following: PN signals fom the teste ae tansmitted by the efeence antenna into the SC at downlink fequency f 1, the eceived powe level and demodulated BER of the EUT ae looped back to the efeence antenna at the uplink fequency f. Then applying displacement method, we find optimum position of the EUT and the efeence antenna. At the optimum position, vaying the tansmit powe, we find the theshold powe, P th that achieves BER with the defined theshold. The elation between P th and P S that achieves the theshold BER is witten as P th C PS 1 (16) whee C is an unknown coefficient which should be detemined by calibation. Afte that, we calibate the system using a configuation s shown in Fig. 14. The eceived powe by the spectum analyze P R is witten as PR PSG 1 C (17) Finally TRS is obtained as

11 P TRS P P 1 P R 1 S th SG (18) EIS (,) PS EIS, G, PS EIS, G, Ps Receive BER (theshold) Ps Receive BER (theshold) (a) Conducted sensitivity (b) TRS Fig.1 Measuements fo eceive sensitivities Test antenna EUT ( :ad. efficiency) Mod. Signal (f 1 ) Teste Anitsu MT88 P th Rec. Powe Ps (f BER ) Fig.13 TRS measuements by spheoidal couple Test antenna Refeence antenna ( : ad. Efficiency) CW (f 1 ) P R SG P SG SA Fig.14 Calibation of the SC system

12 3. Measued esults fo actual UMTS teminals Fig. 15 shows a set-up fo TRS measuement. The paametes of the SC ae the same as those used fo TRP measuements. As EUT, two UMTS teminals wee used. The uplink and downlink fequencies ae 1.9GHz and.1ghz, espectively. As the teste, Anitsu MT88 (Radio Communication Analyze) was used. Fig. 16 shows the measued eceived powe of EUT s looped-back to the teste. The coupling facto is obtained as the diffeence between the tansmit powe fom the teste and the eceived powe at the optimum position. The coupling factos (CF), the diffeence between the tansmit powe and the eceived powe by EUT, fo the sample 1 and the sample wee 4.5 db and 3.7dB, espectively. In this measuement, we did not conduct the calibation fo simplicity, because both of the SC loss and the input eflection of the test antenna ae consideed to be sufficiently small, namely C in Eq.(16) is unity, and is zeo. The coupling facto is intepeted as the antenna efficiency. Table 5 shows the esults of conducted sensitivities measued in diffeent envionments. In the measuement in an usual office oom, moe tansmit powe and the eceive input powe Ps ae equied to achieve the theshold of BER than those in the SC. This is due to a eason why the SC povides a bette shielded envionment and suppess extenal noise. Table 6 shows the measued esults of TRS. Since the ecommended TRS of 3GPP is -14 dbm, it can be said that the two tested teminals have sufficient sensitivity chaacteistics. Spheoidal Couple (a = 1.m) Wideband Antenna W-CDMA Teminal Received powe [dbm] Tansmit powe : 33.3 dbm 3.7dB Coupling Facto: CF 4.5dB Sample-1 Sample- Fig.15 Set-up fo TRS measuement Table 5 Conducted measuements of sensitivity z [mm] Fig.16 Displacement chaacteistics of EUT eceived powe EUT Envionment Tx powe [dbm] P s [dbm] BER [%] Sample-1 Sample- Office oom SC [ * ] Office oom SC [ * ] [ * ] The spheoidal couple is used just as a shield oom. Table 6 OTA measuement EUT BER [%] P s [dbm] CF [db] TRS [dbm] Sample Sample

13 4. Multiple Reflection Suppession 4.1 Extaction of the pimay wave in the SC In the TRP measuement method based on the displacement method and the eflection compensation method descibed above, impedance mismatch of EUT own is compensated. In ode to measue TRP including mismatch effects, we popose anothe mode of opeation of the SC, multiple eflection suppession mode, which is quite diffeent with the fome measuement method caied out in multiple eflection envionment. We can undestand behaviou of adio waves in the SC by investigating time-domain phenomena. Fig. 17 shows a simulated impulse esponse of the input cuent of a Rx antenna. It is consideed that the pimay wave, i.e. the fist coming wave, maintains adiation chaacteistics of Tx antenna in fee space because it is not intefeed with highe ode eflected waves. Theefoe, if we can extact only the pimay wave by suppessing highe ode waves, we can measue TRP of EUT with inheent mismatch by using the SC. Fig. 18 shows a compaison of tansmission S1 of two diffeent dipoles, one is matched and the othe mismatched (VSWR=3). Values of S1 ae obtained by Fouie tansfoming the pimay wave cut fom the simulated time-domain data. As expected, fo thee SCs with diffeent eccenticity, the mismatched dipole has S1 of 1.38dB lowe than the matched dipole, coesponding to VSWR (VSWR=3 is equivalent to a eflection loss of 1.3dB). To suppess multiple eflections, we ae developing a SC system with the Rx antenna wapped with adio absobe as shown in Fig. 19. In the egion suounding the Rx antenna, since stong field is concentated, even adio absobe with small volume can efficiently absob multiple eflections. Fig. shows the effects of the Rx antenna. (a) and (b) show a simulated and a measued S1, espectively fo thee kinds of absobing mateial with diffeent absoption ates. In the measued data, small ipples disappea, this means the Rx antenna is effective fo suppession of multiple eflections. Received cuent [MA] Pimay Seconday ns 1.9 ns Thidly Time [ns] 15.7ns 5.3 ns 5.3 ns Fig.17 Impulse esponse of input cuent of Rx antenna S 1 [db] Matched dipole 1.38dB Mismatched dipole (VSWR 3) Eccenticity Fig.18 S1 deived fom the pimay wave

14 Fig.19 SC system with Rx antenna wapped by adio absobe Radio absobe A Radio absobe B Radio absobe C Radio absobe A Radio absobe B Radio absobe C S 1 [db] -3 S 1 [db] [GHz] [GHz] (a) Simulated (b) Measued Fig. Effect of the absobe wapped Rx antenna 5. Conclusions We poposed an estimation method of eflection coefficients of EUT by phase otation technique in TRP measuement using a spheoidal couple, that impoves accuacy and eliability of the displacement method. We conducted ound obin tests of TRP, and measued data wee compaed with those by othe conventional methods. Futhemoe, we descibed TRS measuement method, and showed measued TRS of actual UMTS teminals. Elimination of the conventional adio anechoic chambe and spheical positione cuts system size and cost. Futhemoe, thanks to vey high sensitivity, it makes possible to measue TRP of not only in-band signals but also low level spuious adiation. The SC system is consideed to be vey pomising fo futue OTA measuements. Acknowledgments This wok has been conducted unde the sponsoship of the R&D pogam fo expansion of adio esouces pomoted by the Ministy of Intenal Affais and Communications, Japan.

15 Refeences [1] IEC6498-1, Amendment Methods of measuement fo adio equipment used in the mobile sevices - Pat 1: Geneal definitions and standad conditions of measuement [] Cellula Telecommunications & Intenet Association, Test Plan fo Mobile Station Ove the Ai Pefomance; Method of Measuement fo Radiated RE Powe and Receive Pefomance, Revision.1, Ap. 5. [3] 3GPP TR V8.. (8-1)3 d Geneal Patneship Poject; Technical Specification Goup Radio Access Netwok; Measuements of adio pefomances fo UTMS teminals in speech mode (elease8) [4] Tasuku Teshiogi, et al, Total Radiated Powe Measuement fo Small Radio Teminals Using a Spheoidal Couple, AMTA8-3, Nov. 8, Boston [5] Tasuku Teshiogi, et al., Total Radiated Powe Measuement fo Antenna Integated Radios Using a Spheoidal Couple, Poc. ISAP9, PP , Oct. 9, Thailand [6] Tadahio Watanabe, et al., Total Radiated Powe Measuement above 1 GHz with Patially- Spheical Scanning of a Pobe, EMC 9/Kyoto, 1R3-3, pp , Kyoto, July 9 [7] T, Sugiyama, T. Shinozuka, K. Iwasaki, Estimation of adiated powe of adio tansmittes using a evebeation chambe, IEICETans.Commun., vol.e88-b, no.8, pp , Aug. 5