Jump to content, skipping navigation

A-GPS Over-the-Air Test Method - Business and Technology Implications White Paper

With the rise in Location Based Services (LBS) applications and the need to meet E911 requirements, the number of mobile cellular devices supporting Assisted GPS (A-GPS) is steadily growing. As one of the enabling LBS technologies, A-GPS offers customers higher position accuracy, quicker location fixes, and improved coverage of service in difficult locations, such as urban and in-building environments.

This white paper presents an overview of the business and technology drivers for OTA A-GPS testing, which describes a method to satisfy the radiated testing requirements of all involved parties. While the white paper is intended for Department Managers, Lab and R&D Managers, and Engineers already working with A-GPS or OTA, the introductory content in the following sections is also beneficial for those unfamiliar with these concepts.

    * Required Field

    Cancel

    White Paper A-GPS Over-The-Air Test Method: Business and Technology Implications Ron Borsato, Spirent Communications Dr. Michael D. Foegelle, ETS-Lindgren www .spirent.com 2www .spirent.com With the rise in Location Based Services (LBS) applications and the need to meet E911 requirements, the number of mobile cellular devices supporting Assisted nullS (AnullnullS) is steadilnull gronullngnull As one of the enabling LBS technologies, AnullnullS offers customers higher position accuracnull, quicnuller location nulles, and improved coverage of service in difnullult locations, such as urban and innulluilding environmentsnull As a result, mobile operators and device manufactures are loonullng for testing choices that quantifnull and benchmarnull realnullorld device performancenull nullntil recentlnull, all industrnullnullenulled AnullnullS test methodologies focused on testing the performance of a device over a cabled null connection, bnullassing the nullS antenna and associated circuitrnull, as shonull in nullgure 1null nullis approach does not give the complete picture of realnullorld device performance and its impact on the endnullser enullerience of LBS applicationsnull nullo achieve this, nullS performance testing needs to include all relevant componentsnull An nullernullnullenullAir (nullnullA) test methodolognull, shonull in nullgure null is the best solution to address this neednull nullis nullite paper presents an overvienull of the business and technolognull drivers for nullnullA AnullnullS testing, nullich describes a method to satisfnull the radiated testing requirements of all involved partiesnull While the nullite paper is intended for nullpartment nullanagers, Lab and nullnullnullanagers, and Engineers alreadnull nullrnullng nullth AnullnullS or nullnullA, the introductornull content in the follonullng sections is also benenullial for those unfamiliar nullth these conceptsnull Introduction Figure 1. Conducted RF Signal Figure 2. Over-The-Air RF Signal 3 www .spirent.com Conventional Standalone GPS With conventional standalone nullS, the nullS receiver in the monullile device is solelnull responsinullle nullor receiving satellite signals and computing its location. This method renullires the device to tracnull at least nullour satellites to compute its location. The same method is used nullnull nearlnull all nullersonal nullvigation Devices nullnullsnull nullnullone nullo has used these devices has enullerienced the long time delanull associated nullith getting a position null nullen the device is nullst ponullered on. nullDs also have limitations in onullaining a position null in challenging environments, such as indoors, in nullrnullan cannullonsnull and under dense nulloliage. Figure 3. GPS User in Urban nullnvironnullnt Conventional Standalone GPS vs. Assisted GPS Assisted GPS nullsing nullireless connectivitnull, monullile devices can also support assisted nullS null-nullSnull nullnullS improves the location determination pernullormance nullnull onullaining nullssistancenulldata nullom the netnullornull over the nullireless communication channel, enanulllingnull null Faster initial acnullisition onull satellites. null null enullective increase in nullS sensitivitnull, nullich can result in position nulles in more challenging environments. null Some position calculations to null onulloaded to a remote server, nulleeing the devicenull processor to service more critical nullunctions. These advantages are the reason nearlnull all monullile devices nullith nullS chipsets support nullnullS. nullobile devices can manulle use of several different approaches to determine their current locationnull Some of these null nullell nullentinullation, nullplinnull nullme nullfference of Arrival (nullnullnullA), Advanced nullornullrd Linnull nullrilateration (Anullnull, and Enhanced nullserved nullme nullfference (Enullnull) nullrelnull on the cellular netnullrnull Another popular method, used bnull devices that support Wireless LAnull(WLAnull, is mapping nullonull WLAnullaccess points then using this information to appronullmate a devicenull current locationnull nullnullever, since the manullritnull of these approaches have limitations, including accuracnull and availabilitnull, the most common method emplonulled in mobile devices todanull is the nullobal nullositioning Snulltem (nullS)null nullS is a nullobal nullvigation Satellite Snulltem that has been fullnull operational since 199null nullvices nullth embedded nullS receiving capabilitnull are able to accuratelnull compute their current position almost annullhere on earth nullere satellite signals can be receivednull nulle reliabilitnull, accuracnull, and availabilitnull of this technolognull have driven nulldespread adoptionnull nullobile devices nullth nullS have tnull options nullen determining their current locationnull 1) nullonventional standalone nullS capabilities and null Assisted nullS (AnullnullS) 4www .spirent.com Recentlnull, industrnull organinullations, including CTnull, have recogninulled the need to create standardinulled test procedures nullor nullnullS nullTnulltesting to onullnullectivelnull specinullnull and validate acceptanullle pernullormance. nullsunullgroup onull the CTnull organinullation has completed nullornull to include nullor the nullst time, a section on nullnullS nullTnulltesting. This is incorporated in the version nullnullrelease onull the nullnull nullest nulllan for nullobile Station nullernullnullenullAir nullerformance nullereanullter renullerred to as the CTnull nullTnullTest nullannull The general test methodolognull denulled in this specinullation is enulllained in the nullollonulling sections onull this nullite paper. The certinullation organinullation nullor nullonullal Snullstem nullor Monullile Communications nullSMnulland nulliversal Monullile Telecommunications Snullstem nullMTSnulldevices sold into nullrth nullerica, the nullS Tnulle Certinullation RevienullBoard nullTCRBnull are linullelnull adopters onull the nenullversion onull the CTnull nullTnullTest nullan. null is also linullelnull that other industrnull nulldies nullill adopt similar methods in the nulluture. nulldustrnull organinullations are not the onlnull ones interested in mandating nullnullS nullTnullrenullirements. Mannull netnullornull operators also nulllieve this testing is vernull important and some alreadnull have nullnullS nullTnulltest programs in place. nullnullthat version nullnullonull the CTnull nullTnullTest nullan is nullalinulled, mannull onull these netnullornull operators are enullected to adopt the methodolognull in this specinullation to help ensure the pernullormance onull the nullnullS-capanullle devices thenull onuller their customers. About CTIA - The Wireless Association nullnullnullnulle Wireless Association is an industrnull consortium representing the nullreless communications industrnull in the nullnited Statesnull nullounded in 19null, this organinullation represents netnullrnull operators, device manufacturers, nullreless datanulletnullrnullng companies, and other contributors to the nullreless sectornull null addition to lobbnullng the nullnullnullnullongress and nullnullnullon behalf of the nullreless industrnull and operating one of the industrnulls largest trade shonull, nullnull maintains a nullreless device certinullation program intended to ensure a high standard of qualitnull performance for consumersnull Over-The-Air Testing The need nullor nullTnullpernullormance testing onull cellular and WiFi nullireless devices has long nullen a nullenull renullirement in the overall testing process. nuller the nullears, standard nullTnullpernullormance test plans have nullen created nullnull organinullations such as CTnull - The Wireless nullsociation null , nullnull, and Wi-Fi nulllliance null . nullTnull testing is pernullormed in a controlled radiated environment, called an anechoic chamnullr, using specialinulled enullipment to provide a nullonull signal to the device under test. nullnullenull aspect onull this testing is that all signals are transmitted and received nullirelesslnull, as thenull are in the real nullorld. This ensures that all interaction nullactors nulltnulleen the radio and the rest onull the nullireless platnullorm, including radiation pattern and platnullorm internullerence, are tanullen into account nullen determining overall nullireless pernullormance. nulltil recentlnull, all nullnullS testing to industrnull standards nullas pernullormed using a canullled RF connection. null a consenullence, devices that pass all tests in the enullisting connullormance standards manull pernullorm poorlnull in the real nullorld. This results in an innullerior end-user enullerience onull LBS applications and has led some industrnull leaders to create their onull nullnullS nullTnullsolutions. 5 www .spirent.com nullis section gives an overvienull of the AnullnullS nullnullA test method specinulld in the version nullnullrelease of the nullnull nullnullA nullest nulllannull AnullnullS nullernullnullenullAir testing requires specialinulled equipment benullnd that required for conducted testing over an null cablenull nulle test method described in this section applies to nullnull, nullSnull and nullode nullvision nullultiple Access (nullnull) devicesnull Required Equipment and Setup The goal onull nullTnulltesting is to onullain a nullnapshotnullonull the pernullormance onull the device-under-test nullnullnullin all directions around the device. For enullample, consider a renullirement to compare the amount onull light emitted nullom a light nullulnullaround the room in all directions. null is necessarnull to loonull at the light nullulnullnullom all directions to measure and compare the results. The Dnull is connullured nullor tnullical use cases. For a monullile device, this includes use onull a phantom head and hand to simulate the enullects onull a device held against the human head. For hand-held applications, such as personal navigation using nullnullS, a phantom hand is used to hold the device in the same nullanull a user tnullicallnull nullould. Thus, the RF shadonulls and near nullld enullects caused nullnull the pronullimitnull to these phantoms can null tanullen into account nullen determining the device pernullormance. The radiated energnull nullom or to the Dnull is measured nullnull placing a Measurement nulltenna nullnull a nulled distance anullanull nullom the device. Because the Dnull can null randomlnull oriented nullith respect to the Mnull, a dual polarinulled measurement antenna is used to measure tnullo orthogonal polarinullationsnullrecording the total radiated energnull irrespective onull the relative orientation. null all linullelihood, the device nullill null operating in a highlnull scattered environment nullen operating near the limit onull its sensitivitnull. null this case, the device does not nullavor annull particular polarinullation. The test methodolognull nullor nullnullS nullTnull testing utilinulles an Mnullnullith linear polarinullation, as opposed to circular polarinullation to remain compatinullle nullith the enullisting CTnull nullTnullTest nullan. To cover all points on the surnullace onull a sphere surrounding the device, it is necessarnull to null anullle to move the Mnullrelative to the Dnull in tnullo orthogonal anulles. nullagine loonulling at a glonull onull the Earth and nullanting to ensure that nullou have onullserved evernull part onull its surnullace enullallnull. nullou nullould have to move north and south, as nullell as east and nullest, to cover the entire glonull. This A-GPS Over-The-Air Test Method movement onull the Mnullrelative to the Dnull renullires some nullorm onull spherical positioning snullstem. There are tnullo common nullanulls to achieve this. The nullst is to mount tnullo orthogonal positioners, one on top onull the other, to rotate the Dnull in tnullo anulles. null this comnullined anullis scenario, the Mnullremains nulled, nullile the Dnull rotates in tnullo anulles. The second involves placing the Dnull on a turntanullle and using a separate positioner to move the Mnullup and donull around it. null either case, nullom the vienulloint onull the Dnull, the Mnullmoves northnullouth nullheta nullqnullanullisnulland eastnullest nullhi nullfnullanullisnull around it, resulting in nullull spherical coverage. To avoid unnullanted internullerence nullom outside signal sources, and prevent internullerence nullith other communication snullstems, the Dnull and Mnullmust null shielded nullom the outside nullorld. This is done nullnull placing them inside an RF shielded room. nullnullever, nullile the shield renullcts enullernal energnull anullanull nullom the Dnull, it also renullcts energnull radiated nullom the Dnull nullacnull tonullards the Mnulland vice-versa. This can result in the energnull nulling measured more than once. This duplication occurs nullcause the energnull can null measured directlnull nullom the Dnull, as nullell as anullter it renullcts onull the nullalls onull the room. To prevent this nullom happening, the room must null lined nullith RF anullsornulling material to reduce unnullanted renullctions. The result is a nullullnull anechoic chamnullr nullere all onull the nullalls, the nullor, and the ceiling are lined nullith RF anullsornullr. nulltside the chamnullr, the measurement antenna must null connected to test instrumentation to measure the ponuller radiated nullom the Dnull, or to transmit signals at a nullonull level to the Dnull to determine its receiver sensitivitnull. The path loss associated nullith canullling, measurement antenna gain, and range path loss must null applied to correct the test enullipment reading to correspond to nullat is occurring at the Dnull. To determine radiated ponuller nullom the Dnull, a signal analnuller or ponuller meter is tnullicallnull used. To determine the receiver sensitivitnull onull the Dnull, a nulltnullornull Emulator nullEnull or satellite simulator in the case onull nullnullS testing, provides the nullonull donulllinnull signal. 6www .spirent.com Depending on nullat test instrument must null connected to the Mnull, it is onullten not practical to maintain the communication linnull to the Dnull through the Mnull. Thus, a separate communication antenna is tnullicallnull used to provide a dedicated communication path nulltnulleen the null and Dnull. This can provide a lonullloss uplinnull path nullen the Mnullis used nullor donulllinnull-onlnull tests. null can also provide nulli-directional communication signaling nullen the Mnullis connected to a signal analnuller nullor ponuller measurement. Because most communication test enullipment is designed nullor conducted testing, additional signal conditioning components are usuallnull renullired to adapt the nuller-The-nullr signals to the availanullle dnullnamic range onull the instrumentation. null RF snullitch matrinull is used to provide all onull the necessarnull routing nulltnulleen the component parts onull the snullstem. Finallnull, a null running test automation sonulltnullare is used to control the positioning snullstem and capture the desired measurements nullom all orientations around the Dnull. Figure null illustrates a tnullical test snullstem in nullich the Dnull is rotated in tnullo anulles and nullich is capanullle onull pernullorming nullTnulltesting nullor a numnullr onull technologies. Test Procedure and Interpretation of Results Because the nullS radio is receive-onlnull, the main interest is in evaluating receiver sensitivitnull nullom various directions around the device. The resulting Enullective nullotropic Sensitivitnull nullnullnullpattern then determines the average radiated receiver sensitivitnull across the entire sphere around the device, renullerred to as Total nullotropic Sensitivitnull nullnullnull or across a portion onull the sphere. null addition to determining the nullaseline radiated sensitivitnull onull the nullS receiver, the enullect onull cellular communication on the nullS receiver is evaluated to ensure that the nullS receiver pernullormance is not degraded due to internullerence nullom the monullile phone transmitter. Traditionallnull, a Tnull measurement nullhe measurement onull an Enull patternnullis determined nullnull pernullorming a sensitivitnull search at each point around the device. The signal level transmitted to the device is lonullered until a target error rate is reported nullnull the device. That denulles the limit onull the devicenull receiver sensitivitnull nullor that direction. The result is a contoured radiation pattern nullere the peanulls represent nulls in the antenna pattern nullere more ponuller nullas renullired to get the signal through, and the vallenulls correspond to the peanulls in the antenna pattern nullere the device is the most sensitive. A-GPS Over-the-Air Test Method nullontnullnull Figure null Tnullical OTA nullnullinullent Setunullnulliagranullused nullth nullrnullssion onullnullS-nullndgrennull DUT Multi-Axis Positioner Dual Polarized Measurement Antenna Fully Anechoic Chamber RF Absorber Material Communication Antenna Measurement Signal Path Communication Signal Path Base Station Emulator Dynamic Range Signal Conditioning RF Switch Matrix Broadband Signal Analyzer Positioning Controller Satellite Simulator PC Running Test Automation Software Fiber Optic Control Lines Network Emulator 7 www .spirent.com Determination onull the Tnull nullor an nullnullS device is complicated nullnull the time involved in determining a nulloodnullvs. nulladnullresult. nullsingle nullnullS null can tanulle over null seconds, and repeated nulles are renullired as the ponuller is lonullered. null is also renullired to onullain a level onull statistical connullence that the appropriate sensitivitnull level has nullen determined. To do this nullom all directions around a device could easilnull renullire danulls onull testing. null an alternative, a method has nullen developed to tanulle measurements nullom the nullnullS device itselnull to help determine the radiation pattern onull the device. The resultant pattern is then normalinulled to a single Enull sensitivitnull search to determine an estimate nullor the entire Enull pattern. The test procedure consists onull nulle stepsnull null nulltenna pattern null Linearinullation null Radiated sensitivitnull null Tnull, nullnull, and nullnull calculation null nulltermediate channel degradation null addition to understanding the test method nullor nullnullS nullTnull, it is important to understand the signinullance onull each measurement and honullit is used to nullantinullnull the nullnullS pernullormance onull devices. This allonulls device manunullacturers to create nulltter-pernullorming devices and helps netnullornull operators ensure that devices launched on their netnullornull pernullorm nullell. Antenna Pattern The nullst part onull the nullnullS nullTnullTest nullan calls nullor measurement onull the nullS antenna pattern. null antenna pattern can null represented visuallnull to identinullnull the nullireless devicenull anullilitnull to enullectivelnull receive signals nullom dinullerent directions. nullagine the antenna at the center onull the shape in Figure null the areas nullith large peanulls signinullnull the directions nullom nullich the antenna receives signals most enullectivelnull. Figure null Tnullical GPS Antenna Pattern nulltenna pattern measurement is important in nullantinullnulling the true pernullormance onull nullS antennas in monullile devices. null devices nullcome smaller, more ponullernullul, and priced lonuller, the trade-onulls nulltnulleen sinulle, cost, and pernullormance nullcome more dinullcult. This is also true nullor the nullS antennas nonullemnulldded in nearlnull all high-end monullile devices, and an increasing numnullr onull mid- and lonullend devices. For these devices to deliver a good user enullerience nullor location-nullased applications, the nullS antenna pattern should null compromised as little as possinullle. The antenna pattern onull a device can null impacted nullnull a numnullr onull nullactors including, nullut not limited tonull null nullS antenna design null Device nullorm nullactor null Location onull the nullS antenna in the device null nullesence onull a human head or hand near the device 8www .spirent.com Figure nullillustrates the impact onull a human head on a nullS antenna pattern. nullte the large vallenull at the location onull the head. Figure null nullnullct onullnullnulln nullad on Antenna Pattern Figure nullstresses the importance onull nullS antenna location. null this case, the antenna is at the nullttom onull the device nullotice the peanulls nullacing donull ardsnullnullen the device is held upright. Since the device clearlnull nullails to enullectivelnull receive nullS signals nullom directlnull overhead, it is linullelnull to null a relativelnull poor pernullormer. Figure null Poorlnull Pernullrnullng Antenna Pattern While itnull not uncommon to pernullorm passive tests to evaluate an antenna radiation pattern nullnull nulleeding it nullith an RF canullle, this is avoided nullor nullTnulltesting nullor several reasons. The RF canullle itselnull can drasticallnull change the radiation pattern onull a device, especiallnull nullor electricallnull small devices linulle a tnullical monullile phone. null addition, the results nullom interactions nulltnulleen the radio, antenna, and device platnullorm are not the same as the pernullormance onull the individual snullstem components. For nullnullS nullTnulltesting, the antenna pattern is estanulllished nullnull radiating a nullonull nullS signal ponuller level and onullaining nullull spherical coverage around the device. Bnull nulleeping tracnull onull the nullS ponuller levels that the Dnull measures, it is possinullle to plot honullnullell the device receives nullS signals at dinullerent angles onull arrival. For nullnullS, the metric used to characterinulle the antenna pattern is the carrier-to-noise ratio nullnull null nullonull the nullS signal. For the CTnull-denulled tests, null discrete positions are renullired nullor nullull spherical coverage. Measurements are made in tnullo anulles nullve angles in the theta nullqnullanullis and null in the phi nullfnull anullisnull nullditionallnull, tnullo orthogonal antenna polarinullations null.g. parallel to the theta nullqnulland phi nullfnulldirections onull motionnullmust null measured to determine the total ponuller received at each point, nullor a total onull nullnullmeasurements. See Figure nullnullor an illustration. Figure null nullal Anulls Rotation nullth Tnull Antenna Polarinullation Eight nullS satellites are simulated during the antenna pattern measurement. The Cnull null ratio is measured nullnull the device under test nullor each individual satellite, and the average Cnull null is used as the metric nullor each discrete antenna pattern measurement. A-GPS Over-the-Air Test Method nullontnullnull 9 www .spirent.com Linearization The antenna pattern produced in this manner relies on the Dnull to pernullorm measurements on the received nullS signals. nullnullever, the Dnull is not a measurement device nullith a traceanullle calinullation. null order to provide that traceanullilitnull, the pattern measured nullnull the Dnull needs to null corrected to eliminate annull non-linearities introduced nullnull the Dnull. Bnull mapping the average or median Cnull null report nullom the Dnull nullacnull to a range onull signal levels generated nullnull the calinullated signal source null.g. nullS satellite simulatornull a set onull corrections nullor the pattern data can null onullained, essentiallnull transnullerring the calinullation traceanullilitnull onull the signal generator to the Dnull. This linearinullation process results in much more accurate antenna pattern data once this correction is applied. The enullact linearinullation procedure can null carried out in multiple nullanulls and is not covered in this nullite paper. Section nullnull.nulland nullpendinull E.nullin version nullnullonull the CTnull Test nullan nullor Monullile Station nuller-The-nullr nullernullormance descrinulls this procedure in more detail. nullease renuller to the CTnull Certinullation Wenullsite at httpnullnullnull.ctia.orgnullertinullation nullor nullurther innullormation. Radiated Sensitivity nullother important test step is to measure the radiated sensitivitnull, or Enullective nullotropic Sensitivitnull nullnullnull onull the device. nullverage nullS signal levels in clear snullnull conditions are vernull lonull, tnullicallnull -nullnulldBm, nullich are much lonuller than cellular signal levels. null is important nullor a nullS-enanullled monullile device to null anullle to receive in a lonullsignal environment. nulldevicenull nullS sensitivitnull renullcts, to a great enullent, the anullilitnull onull its antenna to receive lonullponullered signals. The nullS pernullormance onull monullile devices is closelnull correlated nullith the user enullerience onull location-nullased applica- tions. When using devices indoors, or in areas nullere the snullnull is onullstructed nullnull trees or other onullstacles, the alreadnull-lonullnullS signal levels are nullurther attenuated. null a result, devices nullith good nullS sensitivitnull nullornull in mannull situations nullere others nullith poorer sensitivitnull do not. Some devices on the marnullet todanull can use nullS signal levels nulllonull-nullnulldBm. Radiated sensitivitnull is measured nullnull lonullering the nullS signals until the Dnull is unanullle to meet the specinulld pernullormance renullirements. The test is pernullormed at the device orientation and Mnullpolarinullation that resulted in the highest Cnull null measurement in the upper hemisphere. The satellite scenario and pernullormance metrics used nullor the test are in accordance nullith the industrnull standards nullor the respective nullireless standard in use nullnullnullTS null.nullnullnullor nullTS, nullnull TS null.nullnullnullnullor nullM, or Tnull -nullnullnullor CDMnullnullnullith the enullception that the actual sensitivitnull level is nullound, as opposed to determining passnullail at a particular signal level. nullce the Enull has nullen determined at this one point, the linearinulled pattern is normalinulled to the corresponding device orientation and polarinullation and then sunullracted nullom the Enull nulln dBnullto produce an Enull pattern. Thus, the remaining Enull points are estimated nullom the one measured Enull value and the measured pattern shape, rather than measuring each Enull point individuallnull. Figure null Average GPS Signal -13nulldnull at Surnullace onullnullarth 10www .spirent.com TIS, UnullS, PIGS Calculation nullce the complete Enull pattern is determined, the Tnull, nullnull, and nullnull nullall onull nullich are isotropic sensitivitnull measurements nullcan null calculated. null discussed previouslnull, Tnull is a metric that represents the average sensitivitnull onull a device in a radiated environment. null represents the lonullest signal level that the device nullould null anullle to operate nullith inull it nullas radiated nullith enullal ponuller level nullom all directions. Tnull is convenient nullcause it is a single metric that represents the overall radiated sensitivitnull pernullormance onull the device, manulling it easnull to nullnchmarnull devices against each other. For Tnull, the entire spherical antenna pattern is used nullee Figure nullnull nullpper nullmisphere nullotropic Sensitivitnull nullnullSnullis similar in concept to Tnull nullut it represents the average radiated sensitivitnull pernullormance onull a device anullve the devicenull horinullon nullee Figure nullnull nullnull is calculated nullnull integratingnullveraging the Enull pattern over the upper hemisphere nullom theta nullqnullnullnullto null degrees. Similarlnull, nullartial nullotropic nullS Sensitivitnull nullnullSnullis calculated using antenna pattern data nullom the upper hemisphere as nullell as null degrees nulllonullthe horinullon nullee Figure nullnull A-GPS Over-the-Air Test Method nullontnullnull Figure 13. Good UnullS Pernullrnullnce Figure 1null Tnull Figure 1null Poor UnullS Pernullrnullnce Figure 11. UnullS Figure 1null nulldoors nullth Renullctions nullonullFloor Figure 12. PnullS Figure null and null illustrate the importance onull nullnull. The devices in nullth images shonullthe same antenna pattern, nullut the one in Figure null is inverted. Despite identical Tnull values, the device in Figure null nullields nulltter nullnull, nullith nulltter pernullormance in an environment nullith partial clearance nullere onlnull the overhead snullnull is unonullstructed. nullnull is an important metric nullcause devices onullten receive signals renullcted onull the ground, nullor enullample, nullile standing indoors nenull to a nullindonull, as shonull in Figure null. nullother advantage onull using nullnull is the nullact that the device nullill not null held in a completelnull vertical orientation nullith respect to the ground, so it can null considered to account nullor some range onull variation around the vertical orientation. 11 www .spirent.com Intermediate Cnullannel nullenulladation nullCnull null addition to measuring the Enull pattern to determine Tnull and other related metrics, an nulltermediate Channel Degradation nullCDnulltest is pernullormed nullor each nulland supported nullnull the monullile device. nullnullS pernullormance manull null anullected nullnull the devicenull active cellular connection due to the cellular sunullsnullstems internullering nullith the nullS receiver. null a result, nullS pernullormance can degrade due to selnull-nullamming nullen dinullerent cellular channels are used. These enullects can onlnull null measured enullectivelnull using nullTnulltesting, since the internullering cellular signal does not reach the nullS receiver in a conducted test. nullD is an important measurement, nullcause user enullerience can null severelnull impacted nullen nullS pernullormance degrades due to the use onull cellular nullenullencies that manull null specinull to a given netnullornull Even inull a device is targeted nullor one netnullornull operator marnullet and its associated nullenullencies, users manull roam to other netnullornulls nullile traveling. Figure null illustrates this potential pronulllem. I’m not in the states! Figure 1null Pernullrnullnce Problenullnullile Roanullng The nullD procedure tests the nullnullS pernullormance across a varietnull onull nullireless operating channels nullereanullter renullerred to as intermediate channelsnull To test this, a Cnull null measurement is pernullormed at the nullid channelnullnullenullencnull in a particular nullireless operating nulland at the same pattern peanull used nullor the nullS sensitivitnull measurement. The same Cnull null measurements are repeated at various intermediate channels nullor that particular operating nulland. The nullal nullD measurement is denulled as the dinullerence nulltnulleen the Cnull null measurement at the mid channel and the lonullest Cnull null at annull intermediate channel nullncluding the mid-channelnull Therenullore, the nullS intermediate channel degradation is alnullanulls nullero or greater. The nullD measurement is pernullormed nullor each operating nulland that the device supports. 12www .spirent.com Anecnullic Cnullamnullr The anechoic chamnullr is a critical piece onull enullipment nullor nuller-The-nullr testing. null serves tnullo purposesnullit isolates the Dnull nullom outside signal sources that could internullere nullith radiated measurements nullile the special RF anullsornulling material inside the chamnullr prevents signal renullctions inside the chamnullr nullom corrupting measurements. null area around the Dnull, nullonull as the nulluiet nullonenull represents a nullalinulld test volume nullere signals nullom the measurement antenna produce a nullld nullith a nullonull level onull uninullormitnull. null long as the device is contained nullithin this test volume, dinullerences in the location onull the antenna on the device nullill have onlnull a small impact on the resulting Tnull measurement. The nullalitnull onull the nulliet nullone is anullected nullnull the range length and the overall chamnullr sinulle. nulleallnull the nulliet nullone nullould null located such that the measurement antenna is in the radiating nullar nullld onull the device and vice versa. The chamnullr must null large enough to cover the renullired test range, and, depending on the positioning snullstem arrangement used, manull need to null over tnullice the range length. null addition, larger chamnullrs achieve nulltter anechoic pernullormance nullnull increasing the angle onull incidence nullith the anullsornullr covered nullalls, therenullnull reducing the strength onull renullctions. Most anechoic chamnullrs used nullor nullTnulltesting are nullite large, enullending nullnullmeters in each dimension. Typical Test Solution Applicationnullspecinull test equipment is needed to accuratelnull perform this procedure per the nullnull nullnullA nullest nulllannull A tnullical test solution for AnullnullS nullnullA consists of the follonullng componentsnull null Anechoic chamber null Specialinulled chamber equipment such as device turnnull tablenullositioner, nullS antenna, cellular antenna, phantom headnulland, and null snulltch matrinull null nullellular netnullrnull emulator null nullS satellite simulator null Snullnull(nullnull) or nullE (nullnull) softnullre server for AnullnullS capabilitnull null Automation softnullre to control equipment, automate test procedure, and present results Cnullamnullr Equipment The chamnullr enullipment consists onull mannull components in addition to an anechoic chamnullr. nullmeasurement antenna is renullired to transmit the simulated nullS signals to the Dnull. For CTnull-denulled nullnullS nullTnulltesting, this must null a linearlnull- polarinulled antenna capanullle onull independentlnull transmitting tnullo orthogonal polarinullations and supporting the nullS Lnullnullenullencnull onull nullnull.null Mnull. The path loss nullom the satellite simulator through all snullitching, canullling, the measurement antenna, and the space inside the chamnullr to the center onull the nulliet nullone is calinullated to allonullthe nullS satellite levels to null renullerenced at the Dnull level rather than the satellite simulator. null addition to the measurement antenna, at least one communication antenna is needed to nullirelesslnull transmit and receive the nullM, nullidenulland Code Division Multiple nullccessnullWCDMnull, or CDMnullcommunication signals nulltnulleen the Dnull and a cellular netnullornull emulator. nullpositioning snullstem is renullired to move the measurement antenna relative to the device in order to pernullorm a spherical measurement around the device. Tnullo orthogonal anulles onull motion, corresponding to the theta nullqnull and phi nullfnullcoordinates onull the spherical coordinate snullstem, are renullired to move the antenna andnullr Dnull allonulling measurements to null made at discrete points around the device. Most snullstems start nullith a turntanullle that rotates the Dnull in one anullis. The second anullis can then null mounted on top onull the nullst to allonullrotation onull the Dnull in tnullo anulles nullithin the chamnullr. This renullires more support structure than nullst Figure 1null nullT and nullad Phantonullnullside an Anechoic Chanuller nullhoto used nullth nullrnullssion onullnullS-nullndgrennull 13 www .spirent.com a simple turntanullle nullhich can use an enullanded polnullstnullrene nulloam column nullor Dnull supportnullsince the second anullis must have an anullle and nullarings, etc., sunullcient to support the nulleight onull the Dnull. nullother alternative nullor spherical measurement is to rotate the measurement antenna up and donull around the device. nullnullever, this renullires a larger chamnullr to accommodate the same range length and test volume, since the measurement antenna must move to the same distance anullve and nulllonullthe Dnull. nullvariant on this nullould null to place multiple measurement antennas nullithin the chamnullr, eliminating the need nullor one or nullth anulles onull motion in nullavor onull electricallnull snullitching nulltnulleen measurement antennas. This adds complenullitnull and limitations to the snullstem including a nulled angular resolution, nullut can reduce test time due to the anullilitnull to rapidlnull snullitch nulltnulleen Mnull as opposed to phnullsicallnull moving the Dnull or Mnull. The device positioner is generallnull controlled nullnull test automation sonulltnullare, allonulling automated testing onull the nullnullS nullTnullprocedure. Tnullicallnull, a head andnullr hand phantom is used nullile testing monullile phones to simulate the impact onull having the phone nenull to a human ear or in a humannull hand. For nullnullS testing, the most linullelnull real-nullorld usage is nullith the device held in a personnull hand. nullving the phone hand-held nenull to a personnull head is more linullelnull nullen manulling an emergencnull call null.g. nullnullin the nullited States or nullnullin Europenull or possinulllnull nullen using turn-nullnull-turn direction services. The CTnull organinullation has vernull specinull renullirements nullor the characteristics onull these head and hand phantoms. null RF snullitch matrinull is also renullired outside the chamnullr in order to connect the test enullipment to the appropriate antennas inside the chamnullr. null addition to snullitching the correct RF signals, the snullitch matrinull enanullles automated snullitching nulltnulleen the tnullo polarinullations onull the measurement antenna. nulltnullornull Emulator nullnullS renullires a nullireless radio communication linnull in order to operate. nullcellular netnullornull emulator is a renullired component onull this solution. This instrument emulates all netnullornull components renullired to estanulllish monullile calls, enullchange necessarnull messages nullor nullnullS sessions, and retrieve nullS Cnull null measurements nullom the phone. The CTnull nullTnullTest nullan applies to all cellular devices, nullether thenull are nullTS, nullM, or CDMnull. nullll nullrth nullerican nullenullencnull nullands supported nullnull a device must null tested. null netnullornull emulator used to test such a device must null anullle to support WCDMnull, nullM, or CDMnullair internullaces at all supported nullrth nullerican nullenullencnull nullands, at a minimum. nullnullever, the devicenull nullnullS pernullormance manull also need to null evaluated in other nullands onull interest null.g. nullM nullnullMnull, nullM nullnull Mnull, and nullTS nullnull Mnullnull For enullample, to enanullle roaming, a tnullical nullTS device manull support three WCDMnulloperating nullands nullnull Mnull, nullnull Mnull, and nullnull Mnullnullplus nullour nullM nullenullencnull nullands nullnull Mnull, nullnullMnull, nullnull Mnull, and nullnull Mnullnull For this reason, it is desiranullle that the netnullornull emulator is anullle to support all onull the nullenullencnull nullands supported nullnull the device under test. nullditionallnull, a high level onull snullnchroninullation is renullired nulltnulleen the cellular and nullS emulators to meet the timing renullirements nullor nullTS solutions. Specinull timing renullirements include coarse time accuracnull delivernull nullnullnull ms uncertaintnull and calculation onull time-to-nullst null nullccuracnull nullnull ms uncertaintnull. The nullar is set even higher nullor CDMnull solutions, nullich renullire nearlnull pernullect snullnchroninullation. The minimum renullirement is nullnullns, nullut annull timing uncertaintnull nullill result in degraded device pernullormance. null is essential that timing snullnchroninullation accuracnull in nullnullS nullTnulltest solutions is high enough to prevent unnecessarnull device pernullormance degradation. 14www .spirent.com Typical Test Solution nullontnullnull GPS Satellite Simulator Emulation onull the nullS satellite constellation is an essential renullirement in this solution. The CTnull nullTnullTest nullan renullires up to eight nullS satellites to null nulloadcast simultaneouslnull. The simulator must accuratelnull control the ponuller level onull each satellite. The nullrinullontal Dilution onull nullecision nullDnullnullonull the satellite constellation is renullired to null null.nullnullor nullTS, null.nullnullor CDMnullnullS nullccuracnull, and null.nullnullor CDMnullnullS sensitivitnull, nullich is vernull precise. This manull implnull the need to automaticallnull renullind the nullS scenario at periodic intervals. The industrnull is also preparing nullor the adoption onull other nullSS constellations, such as nullnullnullS and nullalileo, as nullell as regional snullstems such as nullnullSS. With the enullectation that monullile devices nullill soon adopt these technologies, it nullill nullcome increasinglnull important nullor satellite simulators to also support these additional constellations. SnullC nullnullTSnulland Pnull nullnullAnull Softnullare Server The Serving Monullile Location Centre nullMLCnullis a nullTS netnullornull entitnull that manages several important tasnulls nullor nullnullS positioning. Firstlnull, the SMLC captures assistance data nullom a netnullornull onull nullS renullerence receivers and delivers this data to the monullile device during a positioning session. Secondlnull, the SMLC helps to calculate position accuracnull during MS-assisted positioning sessions. The SMLC sonulltnullare server must nullornull in connullnction nullith the satellite simulator and netnullornull emulator to provide the renullired assistance data correctlnull. The CTnull nullTnullTest nullan denulles vernull specinull assistance data parameters and it is necessarnull that the SMLC server complies nullith the plan. For those loonulling to test nullnullS nullTnullnullnullond the renullirements onull the CTnull nullTnullTest nullan, nullnullinullilitnull and programmanullilitnull onull the SMLC sonulltnullare server is essential. Fullnull characterinulling the sensitivitnull onull a device in the real-nullorld renullires dinullerent levels onull assistance data. Sensitivitnull, nullen tested nullith the manullimum level onull assistance data, is much greater than sensitivitnull tested nullith no assistance data. There nullill null a large spectrum onull pernullormance nullen tested nulltnulleen those tnullo enullremes. Test time can null miniminulled nullnull connulluring the SMLC in a nullanull that reduces the time it tanulles nullor devices to return position nulles. With a nullnullinullle SMLC, test enullecution time can null reduced nullnull over nullnullnullithout having a signinullant impact on the nullnullS nullTnulltest results. The nullosition Determination Entitnull nullDEnullis a CDMnull netnullornull entitnull that serves the same purpose as the SMLC in nullTS netnullornulls. The nullE sonulltnullare server must also nullornull in connullnction nullith the satellite simulator and netnullornull emulator to provide the renullired assistance data correctlnull. Automation Softnullare The automation sonulltnullare controls the entire test solution. This sonulltnullare provides a single user internullace nullor setting up test sessions, enullecuting tests, and analnulling results. null a minimum, the CTnull-denulled test method nullor nullnullS nullTnullshould null automated in this sonulltnullare. nullditionallnull, it manull allonull parameters to null modinulld nullor custominulled test scenarios. null nullenull nullnenull onull good automation sonulltnullare is that it reduces the complenullities onull the test procedures and instrumentation control, manulling user interaction nullith the solution intuitive and easnull-to-use. To save time and cost, the sonulltnullare should control all enullipment in the snullstem once a test session is started, reducing the need nullor user intervention and increasing the repeatanullilitnull onull tests. nullmanullor advantage is that custominulled test sessions can null saved and tested again at annull time. This allonulls an understanding as to honulldevice pernullormance changes as hardnullarenullonulltnullare modinullations are made, and to understand honullpernullormance varies across dinullerent devices. Criticallnull, automation sonulltnullare also stores and recalls results data nullom the tests that have nullen enullecuted, providing the anullilitnull to vienulland analnulle these results. For nullnullS nullTnulltests, it should null possinullle to carrnull out all the analnullses discussed in the nullTnullTest Method section nullithin the automation sonulltnullare itselnull. The antenna pattern graphs are particularlnull important nullith this test methodolognull. Finallnull, testing inevitanulllnull goes nullong at some point. Whether it is call set up, or understanding a particular protocol error, unenullected pronulllems can alnullanulls occur. nulltomation sonulltnullare should also allonulldenullugging onull unenullected pronulllems. null most cases, this is accomplished nullnull providing tools such as event, instrument communication and protocol logs. 15 www .spirent.com Conclusion The arrival onull nullnullS nullTnulltesting is a vernull signinullant event nullor the cellular industrnull. nulldustrnull nulldies clearlnull recogninulle the need to test nullnullS nullTnullpernullormance in the manner descrinulld anullve and are in the process onull manulling this a mandatornull test procedure. Companies that nullst understand honullto manulle and interpret these measurements have an advantage in selling Location Based Services or the platnullorms that deliver them. nulltimatelnull, nullnullS nullTnulltesting helps to assure the consumer onull a superior end-user enullerience onull LBS applications. Glossary of Terms Acronyms nullescription nullnull nulld nullneration nullartnership nullonullect nullLT nullvanced Fornullard Linnull Trilateration nullnullS nullsisted nullonullal nullositioning Snullstem Cnull null Carrier-to-nullise Ratio CDMnull Code Division Multiple nullccess CTnull Cellular Telecommunications and nullternet nullsociation Dnull Device-nullder-Test Enull Enullective nullotropic Sensitivitnull E-nullTD Enhanced nullnullserved Time Dinullerence FCC Federal Communications Commission nullSS nullonullal nullvigation Satellite Snullstem nullS nullonullal nullositioning Snullstem nullM nullonullal Snullstem nullor Monullile nullSMnullcommunications nullnull nullrinullontal Dilution onull nullecision nullD nulltermediate Channel Degradation LBS Location Based Services Mnull Measurement nulltenna null nulltnullornull Emulator nullTnull nuller-The-nullr nullE nullosition Determination Entitnull nullnull nullartial nullotropic nullS Sensitivitnull nullD nullersonal nullvigation Device nullTCRB nullS Tnulle Certinullation RevienullBoard nullnullSS nullasi-nullenith Satellite Snullstem SMLC Serving Monullile Location Centre Tnull Total nullotropic Sensitivitnull nullnull nullpper nullmisphere nullotropic Sensitivitnull nullTS nulliversal Monullile Telecommunications Snullstem nullTDnullnull nullp-Linnull Time Dinullerence onull nullrival WCDMnull Widenulland Code Division Multiple nullccess WLnull Wireless Lnull 16www .spirent.com Spirent Communications Performance Analysis, Wireless 541 Industrial Way West Eatontown, NJ 07724 USA Spirent Communications 1325 Borregas Avenue Sunnyvale, CA 94089 USA SALES AND INFORMATION sales@spirent.com www.spirent.com Americas T: +1 800.927.2660 Europe, Middle East, Africa T: +33 1 6137.2250 Asia Pacific T: +852 2511.3822 ETS-Lindgren Corporate Headquarters 1301 Arrow Point Drive Cedar Park, TX 78613 USA SALES AND INFORMATION info@ets-lindgren.com www.ets-lindgren.com Americas T: +1 512.531.6400 Europe T: +358.2.8383.300 Asia T: +65.6536.7078 nullnullnull Spirent Communications, nullc. nullll onull the compannull names andnullr nulland names andnullr product names renullerred to in this document, in particular the name nullpirentnulland its logo device, are either registered trademarnulls or trademarnulls pending registration in accordance nullith relevant national lanulls. nullll rights reserved. Specinullcations sunullnullect to change nullithout notice. Rev. D nullnullnull nullnald Borsato is a Solutions Architect at Spirent nullommunications in Eatontonull, null and is the chair of the nullnullnull nullS nullnullA Subgroup and the chair of the nullnullnull nullnull Anechoic nullhamber Subgroupnull nullor the past fourteen nullears, he has nullrnulled at manullr nullreless companies including nullerinullon Wireless, nullotorola and Lucent nullechnologies before nullining Spirent nullommunications in nullnullnull null is a recogninulled subnullct matter enullert in the development of radiated sensitivitnull testing for nullnull and nullS and has been a contributor and auditor of the nullnull nullest nulllan for nullobile Station nullernullnullenullAir nullerformancenullnull addition, he has participated in mannull other nullreless standardsnullertinullation nullrnullng groups including the nullnull nullnull and nullnull nulloups, the nullnullnullElectronullAcoustic Adnulloc nulloup, the AnullS nullnullnullcubator, the nullnull nullnull SubnullWornullng nulloup, the nullnull Audio SubnullWornullng nulloup, and the nullnull Bluetooth SubnullWornullng nulloupnull null can be contacted at nullnnullorsatonullpirentnullcomnull nullnullnullichael nullnullnulloegelle is the nullrector of nullechnolognull nullvelopment at Enullnullindgren in nulledar nullarnull nullnull null is the industrnull recogninulled subnullct matter enullert in radiated null testing nullth numerous publications in the areas of Electromagnetics, Enull, Wireless nullerformance nullesting, and nullondensed nullatter nullhnullicsnull null is conullhair of the nullnullnull nullonverged nullvices Adnulloc nulloup, has served as vicenullhair of its nullnull Anechoic nullhamber Subgroup and Winulli Alliance Winullinull nullobile nullonvergence nulloup, and is the editor and principal contributor for the Winullnullnullorum null nulldiated nullerformance nullests (nullnull for Subscriber and nullobile Stations nullest nulllannull null addition, he has been involved in numerous standards committees on Enull and nullreless, including the AnullSnullASnullnullnull null nullrnullng groups, the nullnull nullertinullation nullogram Wornullng nulloup on nullernullnullenull Air performance testing of nullreless devices, the nullEE nullnull11 nullasnull nulloup nullfor nullreless performance prediction of nullnull11 devices and mannull morenull null can be contacted at foegellenulltsnullindgrennullcomnull Ronald nullrsato Snullrent Conullunications null. nullichael null. Foegelle nullS-nullndgren