Research Article Measurements and Analysis of Secondary User Device Effects on Digital Television Receivers

Transcription

1 Hidawi Publishig Corporatio EURASIP Joural o Advaces i Sigal Processig Volume 2009, Article ID , 13 pages doi: /2009/ Research Article Measuremets ad Aalysis of Secodary User Device Effects o Digital Televisio Receivers Timothy R. Newma, 1 Daiel DePardo, 2 Alexader M. Wygliski, 3 Joseph B. Evas, 2 Rakesh Rajbashi, 2 Victor R. Petty, 2 Diesh Datla, 1 Frederick Weidlig, 2 Paul J. Kolodzy, 4 Michael J. Marcus, 5 Gary J. Mide, 2 ad James A. Roberts 2 1 Electrical ad Computer Egieerig, Virgiia Tech, Blacksburg, VA 24061, USA 2 ITTC, Uiversity of Kasas, Lawrece, KS 66044, USA 3 Worcester Polytechic Istitute, Worcester, MA 01609, USA 4 Kolodzy Cosultig, P.O. Box 1443, Cetreville, VA 20120, USA 5 Marcus Spectrum Solutios LLC, 8026 Cypress Grove Lae, Cabi Joh, MD 20818, USA Correspodece should be addressed to Joseph B. Evas, evas@ittc.ku.edu Received 7 July 2009; Accepted 12 August 2009 Recommeded by K. Subbalakshmi This article presets results from a study of the potetial effects of secodary users operatig i uoccupied televisio spectrum. Televisio spectrum is kow withi the wireless commuicatios commuity as beig uderutilized, makig it a prime cadidate for dyamic spectrum access. The proposed useof this ope spectrum has prompted questios cocerig the quatity of available chael space that could be used without egative impact o cosumers who view digital televisio broadcasts ad the viability of secodary use of ope chaels immediately adjacet to a digital televisio broadcast chael. I this work, we ivestigate secodary device operatio i the chaels directly adjacet to a desired televisio chael, ad the effects upo a selectio of cosumer digital televisio (DTV) receivers. Our observatios strogly suggest that secodary users could operate White Space Devices (WSDs) i uoccupied chael badwidth directly adjacet to a desired digital televisio (DTV) chael, with o observable adverse impact upo the receptio of the desired chael cotet. Copyright 2009 Timothy R. Newma et al. This is a ope access article distributed uder the Creative Commos Attributio Licese, which permits urestricted use, distributio, ad reproductio i ay medium, provided the origial work is properly cited. 1. Itroductio The growig demad for wireless services ad applicatios shows o sig of slowig dow. However, the curret commad-ad-cotrol regulatory structure for licesig spectrum has bee uable to cope with the dramatic growth of the wireless idustry. This has give rise to a artificial scarcity with regard to spectrum, resultig i prices for spectrum liceses becomig so cost prohibitive that may small to medium size busiesses are preveted from eterig the wireless market. Numerous studies have begu to examie how licesed spectrum is actually used, with the goal of ot oly rethikig the spectrum licesig regime, but also opeig certai uderutilized prime spectrum to ulicesed ad licesed secodary usage. It has bee show that several spectral bads, icludig the televisio spectrum, are uderutilized [1]. There have bee regulatory ad legislative efforts to allow ew wireless devices access to televisio (TV) bad white space o a per market basis. This approach, referred to as dyamic spectrum access (DSA), allows ulicesed devices to trasmit i parts of the spectrum uoccupied by the licesed sigals. I Jue 28, 2006, the Seate Commerce Committee adopted The Advaced Telecommuicatios ad Opportuity Reform Act of 2006 (S. 2686), which built upo the May 2004 Federal Commuicatios Commissio (FCC) Notice of Proposed Rulemakig (NPRM) [2] allowig ulicesed devices to utilize uused spectrum i the TV bad. This legislatio required the FCC to cotiue with rule makig procedures goverig the opeig of TV chaels

2 2 EURASIP Joural o Advaces i Sigal Processig Sidebad splatter DTV broadcast 6MHz 1 Chael +1 Figure 1: Adjacet Chael Iterferece. Table 1: Adjacet chael measuremets. Chael Receiver 1 Receiver 2 Receiver 3 Offset (set top) (LCD DTV) (set top) 1 34 dbm 30 dbm 29 dbm dbm 18 dbm 26 dbm 2 15 dbm 16 dbm 22 dbm 6 N/A 15 dbm N/A 8 N/A 14 dbm N/A Table 2: Co-chael desired/udesired ratio measuremets. ATSC Receiver 1 Receiver 2 Receiver 3 (set top) (LCD) (set top) Threshold ( 102 dbm) ( 96 dbm) ( 97 dbm) 68 dbm ( 81 dbm) ( 79 dbm) ( 80 dbm) 61 dbm ( 73 dbm) ( 72 dbm) ( 72 dbm) 2 51 (54 MHz 698 MHz) for use by wireless broadbad services ad other DSA-eabled devices. The FCC proposal also icludes the reallocatio of TV chaels (698 MHz to 806 MHz) to public safety commuicatios as well as for auctio. The NPRM specifies that ay devices certified to use TV spectrum white spaces should employ either agile or cogitive radio techology i a dyamic spectrum access cofiguratio, such that these devices would ot iterfere with primary rights holders, for example, televisio broadcasters. Further legislative directio was provided i the Balaced Budget Act of 1997 (H.R.2015), Title III - Commuicatios ad Spectrum Allocatio Provisios [15]. I November 3, 2008, FCC commissioers voted uaimously to officially approve the use of these bads by white space devices (WSDs). I a DSA approach, the secodary users must ot cause ay harmful iterferece to the primary users as well as the other secodary users sharig the same portio of the spectrum. The defiitio of harmful is outside the scope of this article, but is typically associated with the affected applicatios ad their hardware characteristics. Sice primary users hold exclusive rights to the spectrum, it is ot their resposibility to mitigate ay additioal iterferece caused by secodary device operatio. These devices will eed to periodically sese spectrum i order to detect primary ad secodary user trasmissios, ad have the capability of adaptig to the varyig spectrum coditios for mutual iterferece avoidace [3]. Groups represetig traditioal icumbet users, such as televisio broadcasters, feel that opeig up the white space to DSAeabled secodary users will cause ueeded iterferece ad disrupt commuicatios for televisio broadcasters ad public safety commuicatio systems. Recetly the Natioal Associatio of Broadcasters ad Associatio for Maximum Service Televisio wet so far as to sue the FCC over its decisio to authorize the operatio of WSDs i the uused televisio chaels. The availability of the uderutilized televisio spectrum is ot disputed. However, there are two remaiig techical issues that the regulatory ad busiess commuities must address. The regulatory commuity must determie the techical rules that devices must use whe accessig this spectrum i order to prevet harmful iterferece to the primary devices (i.e. DTV receivers). Additioally, the device maufacturig commuity must determie if cost effective devices ca be created that meet the techical rules for operatio. The techical rules that are of particular iterest to the FCC ivolve device i-bad ad out-of-bad emissios (OOBEs). The device maufacturers must comply with these rules through the selectio of appropriate modulatio, amplifier, ad filter characteristics i both the trasmitter ad receiver chais of WSDs. The expected RF eviromet also has a direct impact o the receiver characteristics ad thus must be well uderstood. I this effort, we preset a feasibility study of devices performig DSA i uderutilized televisio bads whe televisio sigals are preset. The study is divided ito two parts. I the first part, the impact of trasmissios o the video quality of digital televisio sigals is determied for several scearios. This will provide the basis i determiig the emissio levels that ca be tolerated by DTV receivers. I the secod part, we focus o adjacet chael iterferece, ad measure the affects of a sigle adjacet chael secodary trasmissio o a selectio of DTV receivers. 2. Dyamic Spectrum Access Backgroud Substatial research efforts have bee aimed at the utilizatio of vacat portios of FCC allocated televisio spectrum usig DSA techiques. IEEE is a wireless stadard that focuses o reuse of the vacat TV spectrum without causig ay harmful iterferece to the primary users [4]. Some of the importat issues that have bee addressed iclude the

3 EURASIP Joural o Advaces i Sigal Processig 3 3dB coupler Ohm broadbad matchig etwork PC Spectrum aalyzer ATSC receiver uder test MPEG-2 aalysis software OFDM modulator QPSK/16 QAM/64 QAM 6MHzBW 2 K/8 K carriers Video moitor MPEG-2 stream IEEE-1394 (FireWire/i.Lik) Figure 2: KU White Space Device Emulator ad Testbed (KUWSDET). feature detectio of TV sigals [5], collaborative sesig for improved detectio capabilities [6], detectio of the presece of receivers i the viciity of a secodary device [7], ad effective methods for uoccupied TV spectrum access [8]. Our previous work explored the feasibility of secodary user commuicatio from the side of the secodary users [9]. Although it has bee demostrated that DSA methods are effective i avoidig harmful iterferece to televisio receivers, there is still debate regardig whether devices ca operate i uderutilized spectrum without simultaeously causig iterferece. There are may who claim that WSDs will cause harmful iterferece to primary users [10], while others argue that DSA ca be doe i a trasparet maer [11] ad ca be safely implemeted usig the latest radio techology commuicatios techiques [12]. Propoets of the DSA approach have ivestigated the use of ope televisio spectrum for several reasos. First, there is a substatial amout of uused spectrum available for DSA. Furthermore, the propagatio properties i these frequecy rages, such as low propagatio atteuatio, are beeficial for log rage, mobile, ad o-lie-of-sight (NLOS) commuicatios [8]. Fially, the fixed chael allocatios result i determiistic usage patters that are favorable for accurate spectrum sesig [13]. There are several challeges that must be addressed i order to eable white space device operatio i ope TV spectrum. The presece of strog TV sigals ear the secodary user ca lead to the geeratio of spurious sigals, itermodulatio products, ad saturatio effects [14]. These problems might occur at the trasmissio source, at the DTV receiver, or at the secodary-use receiver. I additio, the secodary device could potetially cause iterferece to primary users if the spectrum sesig mechaism of the device fails to idetify the presece of primary users, or if there is sigificat out-of-bad power leakage from secodary trasmissios ito the primary user chaels. I previous studies, we preseted related results cocerig o the impact of TV trasmissios o secodary users at varyig distaces from a trasmissio tower, ad we also provided results that idicated typical curret geeratio cosumer DTV receivers were capable of error-free receptio of adjacet chael DTV broadcasts. These results, while ot comprehesive, provide icetive to further explore the iterferece relatioships betwee secodary users ad DTV receivers. Ithisstudy,weprovideabriefreviewofoursecodary user iterferece feasibility aalysis ad expad upo the results by focusig o adjacet chael DTV iterferece. A udesired iterferig sigal, trasmitted i a chael adjacet to a desired sigal chael, ca itroduce thirdorder itermodulatio (IM 3 ) products geerated by trasmitter oliearities ito the desired chael badwidth, resultig i iterferece to the desired chael cotet. I the case of a DTV broadcast, the oise-like characteristics of a 8VSB-modulated digital televisio sigal create IM 3 products with the same oise-like properties. These products are ofte referred to as sidebad splatter or shoulders ad result i what is essetially co-chael iterferece to other sigals occupyig directly adjacet chaels, as illustrated i Figure 1. The threshold for harmful iterferece (iterferece levels that impact the operatio of the DTV receivers to such a extet that the sigal cotet is severely degraded) varies cosiderably, ad is depedet upo the techical characteristics of the TV receiver ad the secodary user device. Therefore, stadardized procedures for testig the iterferece-limitig capabilities of devices should be devised for various spectrum eviromets ad scearios. Ideally, these techiques ad procedures would be based

4 4 EURASIP Joural o Advaces i Sigal Processig DTV sigal Co-chael iterferece White Space Device upo iput from both the academic commuity ad primary rights holders [16]. A ivestigatio studyig the operatio of public safety trasmissios across televisio spectrum, whe both digital ad aalog televisio sigals were preset, has bee coducted [17]. Although several isights were obtaied regardig the iteractio betwee licesed ad ulicesed trasmissios, the ivestigatio did ot quatify the impact o the video quality of the televisio sigal or was there a aalysis of trasmitter oliearities Adjacet Chael iterferece Desesitizatio iterferece I termodulatio iterferece 2d adjacet chael iterferece + x Figure 3: Types of potetial WSD iterferece to DTV sigals. Chael error display artifacts Figure 4: Displayed effects of DTV chael errors. 3. Viability of DTV Coexistig with Cogitive Radios 3.1. KU White Space Device Emulator ad Testbed. The KU White Space Device Emulator ad Testbed (KUWSDET), show i Figure 2, was iitially cofigured to simulate white space devices operatig i the 54 MHz to 806 MHz frequecy rage usig OFDM modulatio. WSD emulatio was accomplished usig a desktop computer equipped with a PCI form factor modulator, which was capable of QPSK, 16 QAM, ad 64 QAM, usig 2000 or 8000 carriers, ad various code rates ad guard itervals, with a 6 MHz chael badwidth. The RF output level ca be software cotrolled over a 31.5 db rage. Additioal RF amplificatio ad step atteuatio were iserted ito the WSD emulator trasmit chaiasrequiredisupportofspecifictestparameters.the WSD output ad the feed from a roof-mouted cosumer grade directioal TV atea were fed ito a 3 db coupler, ad the combied output was switched betwee a spectrum aalyzer ad the DTV receiver uder test. I the case of DTV receivers equipped with IEEE-1394 (FireWire/i.Lik) output, the KUWSDET has the additioal capability of MPEG-2 trasport stream statistics aalysis, which provides more precise DTV chael performace testig. Tests to date have focused o the effects of WSD trasmissios o cosumer grade DTV receivers Types of Iterferece. Whe wireless trasmissios operate i close proximity to each other i the frequecy domai, there exists the potetial for these sigals to iteract. This iteractio ca egatively impact the ability of a receiver to perfectly recover the desired sigal. By characterizig the spectral characteristics of the sigals located withi a frequecy rage of iterest, it is possible to classify the type of iterferece expected at the receiver. Five types of iterferece that could exist betwee a primary DTV sigal ad a secodary trasmissio i a dyamic spectrum access etwork are show i Figure 3. The differeces betwee each of these types of iterferece are based o the relative spacig betwee the two trasmissios, ad their relative trasmissio power levels. For istace, whe the DTV sigal spectrum is located at chael, ad the secodary trasmissio is also located at the same chael, this is referred to co-chael iterferece. I this sceario, the desired DTV chael would be severely corrupted by secodary device operatio due to its iability to resolve the two sigals. Aother type of iterferece ca occur if the secodary sigal is located i a adjacet

5 EURASIP Joural o Advaces i Sigal Processig 5 KCTV KTWU KSMO KTKA Chael/frequecy Effective radiated power Atea height above average terrai KTWU-DT CH 23/ MHz 960 kw 281 meters KCTV-DT CH 24/ MHz 1000 kw 319 meters KSMO-DT CH 47/ MHz 1000 kw 356 meters KTKA-DT CH 48/ MHz 55 kw 451 meters Figure 5: Adjacet chael DTV broadcast receptio i Lawrece, KS. KTWU CH 23 ( MHz) 40 dbm/5.38 MHz KCTV CH 24 ( MHz) 54 dbm/5.38 MHz Figure 6: KTWU ad KCTV Adjacet chael broadcasts. chael, such as chael + 1. I this case, the DTV sigal may experiece adjacet chael iterferece from the secodary sigal sice the trasmitted spectrum of the latter may ot be totally cofied to its allocated bad. Note that as the amplitude level of the secodary trasmissio is icreased, so does the amout of out-of-bad eergy that could iterfere with the DTV sigal. If the secodary sigal is located further away from the DTV sigal, such as the secod adjacet chael, the impact of adjacet chael iterferece is substatially reduced, relative to secodary sigals operatig closer to the DTV sigal, give the same power levels. However, if the power level of the secodary sigal is icreased, it is possible that out-of-bad sigal eergy may iterfere with the DTV sigal. I fact, whe the secodary sigal is substatially stroger tha the DTV sigal ad is located withi the geographical viciity of a desired frequecy, desesitizatio iterferece ca potetially occur. I this sceario, the secodary sigal overloads the receiver, ihibitig its ability to fully recover the desired DTV sigal. Receiver itermodulatio iterferece occurs whe two or more sigals are preset withi the same frequecy rage, ad are mixed i a receiver RF amplifier or mixer stage durig oliear operatio, producig a sigal that iterferes with a desired sigal. Cosequetly, these receivergeerated sigals could prevet the display of the cotet of a desired DTV chael. The visible effects of DTV receiver iterferece ca rage from mild error artifacts to complete loss of chael cotet. Figure 4 is a example of moderate display errors Prelimiary DTV Adjacet Broadcast Chael Observatios. As previously discussed, televisio broadcasters have expressed the cocer that secodary use of ope spectrum immediately adjacet to a operatioal digital televisio chael could create iterferece for viewers of the DTV chael cotet. It is iterestig to ote that despite the best efforts of the FCC to allocate adjacet chael frequecies to trasmitters with sufficiet geographic separatio, it is ot particularly uusual for viewers to be located i areas where televisio markets overlap, allowig adjacet broadcast DTV chael situatios to occur. The Iformatio ad Telecommuicatios Techology Ceter (ITTC) at the Uiversity of Kasas is located approximately halfway betwee Kasas City ad Topeka, with access

6 6 EURASIP Joural o Advaces i Sigal Processig PC1 OFDM modulator 6MHzBW Udesired sigal A1 R1 A2 R Ohm broadbad matchig etwork Spectrum aalyzer MPEG-2 aalysis software 3dB combier PC2 8VSB modulator 6MHzBW Video moitor ATSC receiver uder test Desired sigal MPEG-2 trasport stream IEEE-1394 (FireWire/i.Lik) Figure 7: KU White Space Device Emulator ad Testbed (KUWSDET) provides a cotrolled eviromet for observig the effects of oliear operatio. to off-the-air DTV cotet from both media markets, as show i Figure 5. Two separate istaces of adjacet DTV chael pairs ca be received at this locatio, ad tests have demostrated that a selectio of cosumer DTV receivers are capable of receivig ad properly displayig the cotet of all four chaels. The KTWU trasmitter i Topeka, KS, USA which broadcasts o chael 23 ( MHz), ad the KCTV facility i Kasas City, MO, USA broadcastig o chael 24 ( MHz) are the first pair of adjacet DTV chaels observed, illustrated i Figure 6. It should be oted that while FCC TV chael allocatio badwidth is 6 MHz, the 8-VSB DTV sigal-3 db badwidth is 5.38 MHz. Cotet from both chaels was displayed durig a live demostratio to FCC Commissioer Michael Copps durig a visit to ITTC usig a uexceptioal cosumer LCD DTV. Our observatios suggest that ope adjacet chael badwidth could also be used for other applicatios. I the followig sectios, we actively explore this possibility by geeratig simulated WSD sigals o chaels adjacet to DTV trasmissios ad observig the effects. 4. Measuremets of TV Bad White Space Devices ad Digital Receivers Iitial experimets revealed that a relatively high WSD chael power level is required before the output egatively impacts a DTV test receiver. Some prelimiary adjacet chael measuremets are illustrated i Table 1. Show are the WSD trasmitter levels (i dbm/6 MHz BW) required to cause visible impairmets to the DTV receptio whe the received DTV sigal is 61 dbm. Prelimiary co-chael measuremets are show i Table 2. The measuremets listed are the co-chael desired/udesired () ratios required to cause visible impairmets to the DTV receptio. Note that there is ot yet a stadard for for o-tv (ATSC/NTSC) sigals. The prelimiary experimetal results from a limited umber of test receivers idicate that the proposed U- D operatio i the televisio bad ca be accomplished without sigificat impact upo DTV receivers i the viciity. Experimets are ogoig ad will be thoroughly reported i future publicatio submissios. 5.SecodaryUserItermodulatio Product Aalysis 5.1. Itermodulatio Measuremet System. As DTV systems have replaced aalog TV equipmet, televisio broadcast professioals have foud the eed to develop ew system performace measuremet methods ad termiology. Ehacemets to the traditioal third-order itermodulatio descriptio for aalog sigals have bee detailed i the article Iterferece Mitigatio for Improved DTV Receptio [18], where the authors ecourage the additio

7 EURASIP Joural o Advaces i Sigal Processig 7 Filtered udesired 14 dbm Filtered udesired 16 dbm = 37 db Desired 53 db Udesired IM3(DTV-T) 2 = 39 db Desired 53 db 1 (a) (b) Filtered udesired 14 dbm Filtered udesired 16 dbm Udesired IM3(DTV-T) 62 dbm Udesired IM3(DTV-T) 60 dbm 1 Udesired IM3(DTV-T) 6 MHz chael power measuremet BW 6 MHz chael power measuremet BW (c) +1 (d) Figure 8: (a) = 53 dbm; udesired sigal at 1; miimum IM3(DTV-T) cofiguratio (b) = 53 dbm; udesired sigal at + 1; miimum IM3(DTV-T) cofiguratio (c) IM3(DTV-T) level preset i desired chael. udesired sigal at -1; Miimum IM3(DTV-T) Cofiguratio. (d) IM3(DTV-T) Level Preset i Desired Chael. Udesired Sigal at + 1; Miimum IM3(DTV-T) cofiguratio. of ew techical terms to describe DTV trasmitter ad receiver characteristics. The proposed DTV third-order itermodulatio product IM3(DTV), third-order DTV Itercept Poit IP3(DTV), DTV Trasmitter-geerated itermodulatio product IM3(DTV-T), ad DTV Receiver-geerated itermodulatio product IM3(DTV-R) parameters offer refiemets to the geeric IM3 ad IP3 terms ad ecourage more specific aalysis of the origis of DTV itermodulatio iterferece. The sigle iterferer adjacet chael measuremets cotaied i this article are preseted usig the techiques ad termiology described i the above-metioed article. IM3(DTV) values i Tables 3 ad 4 are direct measuremets; IP3(DTV) values were calculated usig IP3(DTV) = U + U IM3(DTV). 2 (1) I order to have a cotrolled eviromet for testig ad aalysis we ehaced the capabilities of the KUWSDET with ATSC 8VSB sigal geeratio capability, which was used to create test DTV sigals at desired chael locatios, supplemetig the use of off-the-air broadcast sigals. ATSC 8VSB emulatio was accomplished usig a modestly priced PCI form factor digital modulator ad RF frequecy sythesizer similar to the hardware used to geerate OFDM sigals. The 8VSB ad OFDM systems are housed i separate desktop computer cases; modulated test sigals are geerated usig trasport stream cotet stored o hard drives i each computer. Off-air digital broadcast sigals are also available for experimets via a directioal rooftop-mouted atea. The KUWSDET is capable of accommodatig a variety of experimetal scearios through frequecy agility, ad flexible cofiguratio of RF amplifier gai blocks,

8 8 EURASIP Joural o Advaces i Sigal Processig Filtered udesired 32 dbm Filtered udesired dbm = 21 db = db Udesired IM3 (DTV-T) Desired 53 db Desired 53 db Udesired IM3 (DTV-T) (a) (b) Filtered udesired dbm Filtered udesired 8dBm Udesired IM3 (DTV-T) 65 dbm Udesired IM3 (DTV-T) 66 dbm MHz chael power measuremet BW 6 MHz chael power measuremet BW (c) (d) Figure 9: (a) = 53 dbm; udesired sigal at 1; sigificat IM 3(DTV-T) cofiguratio (b) = 53 dbm; udesired sigal at +1; sigificat cofiguratio (c) IM 3(DTV-T) level preset i desired chael. udesired sigal at 1; sigificat IM 3(DTV-T) cofiguratio. (d) IM 3(DTV-T) level preset i desired chael. Udesired sigal at + 1; sigificat IM 3(DTV-T) cofiguratio. filters, step atteuators, switches, ad power combiers. The cofiguratio used for the IM 3(DTV-T) measuremets preseted here is show i Figure 7. Noliear characteristics of the RF power amplifier, labeled A2 i Figure 7, were used to create IM 3(DTV-T) iterferece. The udesired sigal IM 3(DTV-T) iterferece level preset at the receiver uder test iput was determied by the iput level to A2. The degree to which A2 was drive ito oliear operatio was cotrolled usig the step atteuator labeled R1, while the value of the step atteuator labeled R2 determied the fial udesired sigal RF chael power level. Sigle iterferer DTV receiver tests were coducted i the upper ad lower adjacet chaels to a desired DTV sigal at two itermodulatio iterferece levels. R1 was used to adjust the Udesired Sigal for maximum RF chael power with the miimum IM 3(DTV-T) iterferece preset i the adjacet desired chael, ad maximum RF chael power with very sigificat levels of IM 3(DTV-T) Observatios. Receivers were first tested usig a desired chael power level of 53 dbm measured i a 6 MHz badwidth. Chael 31 ( MHz) was selected as the desired frequecy, ad the ATSC sigal was created usig the KUWSDET 8VSB modulator ad a loopig HDTV MPEG-2 trasport stream. The adjacet chael udesired sigal was geerated usig the testbed OFDM-modulator ad a loopig trasport stream. The filtered udesired sigal level was cofigured to geerate the miimum level of IM 3(DTV-T) iterferece while maitaiig operatio withi the regio of the 1dB compressio poit (P1dB) of amplifier A2. The adjacet chael udesired sigal levels required to degrade the 53 dbm desired chael to TOV are idicated i Table 3(a). Example data from the measuremet techique referece [18] is also icluded for compariso purposes. Spectrum plots represetative of a selectio of Table 1 measuremets are cotaied i Figures 8(a) 8(d). The DTV receivers were the tested usig a filtered udesired sigal cofigured to

9 EURASIP Joural o Advaces i Sigal Processig Filtered udesired 1 dbm = 27 db 9 Filtered udesired 4 dbm = 24 db Desired 28 db Desired 28 db Udesired IM3(DTV-T) 2 Udesired IM3(DTV-T) 1 (a) +2 (b) Filtered udesired 1 dbm 1 +1 Filtered udesired 4 dbm Udesired IM3(DTV-T) 46 dbm Udesired IM3(DTV-T) 53 dbm 6 MHz chael power measuremet BW 6 MHz chael power measuremet BW (c) +1 (d) Figure 10: (a) = 28 dbm; udesired sigal at 1; miimum IM3(DTV-T) cofiguratio (b) = 28 dbm; udesired sigal at + 1; miimum IM3(DTV-T) cofiguratio (c) IM3(DTV-T) level preset i desired chael. Udesired sigal at 1; Miimum IM3(DTV-T) Cofiguratio. (d) IM3(DTV-T) Level Preset i Desired Chael. Udesired Sigal at + 1; Miimum IM3(DTV-T) cofiguratio. geerate cosiderably higher levels of IM3(DTV-T) iterferece i the adjacet desired chael badwidth tha i previous test with the same 53 dbm/6 MHz desired sigal used i the previous test. Adjacet chael udesired sigal levels required to degrade the 53 dbm desired chael to TOV are idicated i Table 3(b). Spectrum plots represetative of a selectio of Table 3(b) measuremets are cotaied i Figures 9(a) 9(d). The desired chael power level was icreased to 28 dbm measured i a 6 MHz badwidth for the ext receiver test. Chael 31 ( MHz) was agai selected as the desired frequecy. The ATSC sigal was created usig the KUWSDET 8VSB modulator ad a loopig HDTV MPEG-2 trasport stream as i the precedig tests. The filtered udesired OFDM sigal was cofigured to geerate miimal levels of IM3(DTV-T) iterferece i the adjacet desired chael badwidth, while maitaiig operatio withi the regio of the 1dB compressio poit (P1dB) of amplifier A2. Adjacet chael udesired sigal levels required to degrade the 28 dbm desired chael to TOV are idicated i Table 4(a). Example data from the measuremet techique referece [18] is agai icluded for compariso purposes, ad spectrum plots represetative of a selectio of Table 3 measuremets are cotaied i Figures 10(a) 10(d). The same 28 dbm desired sigal used i the precedig test was also used for the fial receiver test. For this test, the udesired sigal was agai cofigured to produce sigificat levels of IM3(DTV-T) iterferece. Adjacet chaeludesired sigal levels required to degrade the 28 dbm

10 10 EURASIP Joural o Advaces i Sigal Processig Filtered udesired 8 dbm = db Desired 28 db Udesired IM3(DTV-T) 2 1 (a) Filtered udesired 6 dbm = 22 db Desired 28 db Udesired IM3(DTV-T) (b) Filtered udesired 6 dbm Filtered udesired 8 dbm Udesired IM3(DTV-T) 41 dbm 1 Udesired IM3(DTV-T) 40 dbm 6 MHz chael power measuremet BW 6 MHz chael power measuremet BW (c) +1 (d) Figure 11: (a) = 28 dbm; Udesired sigal at 1; sigificat IM3(DTV) cofiguratio (b) = 28 dbm; udesired sigal at + 1; sigificat IM3(DTV) cofiguratio (c) IM3(DTV) level preset i desired chael. Udesired sigal at 1; sigificat IM3(DTV) cofiguratio. (d) IM3(DTV) level preset i desired chael. Udesired sigal at + 1; sigificat IM3(DTV) cofiguratio. desired chael to TOV are idicated i Table 4(b). Spectrum plots represetative of a selectio of Table 4 measuremets are cotaied i Figures 11(a) 11(d). Isufficiet spectrum aalyzer frot ed atteuatio resultig i IM3(DTV-R) Additioal 10 db of frot ed atteuatio Figure 12: Example of IM3(DTV-R) usig a improperly cofigured spectrum aalyzer Receiver-Geerated Itermodulatio Products. Direct spectrum aalyzer measuremet of receiver-geerated itermodulatio products IM3(DTV-R) was ot covered i scope of this report, due to the ecessity of receiver disassembly ad measuremet port isertio. Future ivestigatios may iclude such a aalysis. However, through the use of a purposely miscofigured spectrum aalyzer, a effort has bee made to illustrate IM3(DTV-R), as show i Figure 12. The aalyzer display cotet was first captured with the iteral RF iput atteuator properly cofigured, the with the same sigal iput levels, the aalyzer atteuator settig

11 EURASIP Joural o Advaces i Sigal Processig 11 Table 3: (a) Desired chael power = 53 dbm/6 MHz; udesired sigal with miimum IM 3(DTV) ; (b) desired chael power = 53 dbm/6 MHz; udesired sigal with sigificat IM 3(DTV). (a) 1 filtered udesired + 1 filtered udesired Receiver 1 CH 30 IP3DTV + 1 CH 32 IP 3(DTV) 1 (1999 model year set-top) (32 LCD DTV) (Low cost set-top) ATSC A/74 Rhodes-sgrigoli example (Figure 4; Iterferece Mitigatio for Improved DTV Receptio ) Levels required to degrade desired chael to TOV IM 3(DTV) 57.8 IP 3(DTV) 0.5 (b) 1 filtered udesired + 1 filtered udesired Receiver 1 CH 30 IP 3(DT V) + 1 CH 32 IP 3(DT V) 1 (1999 model year set-top) (32 LCD DTV) (Low cost set-top) ATSC A/74 Levels required to degrade desired chael to TOV was maually reduced 10 db, resultig i the overload of oe or more aalyzer frot ed compoets. The goal of this simple demostratio is to show that ot oly ca trasmiter oliearities causes iterferece but a improperly cofigured receiver ca also result i sigificat iterferece. 6. Coclusio I this article, we have preseted a feasibility study of secodary trasmissios ito the uused TV spectrum ad have preseted the observed effects of a sigle iterferig secodary trasmissio o a selectio of digital televisio receivers. The results idicate that carefully implemeted secodary use of ope chael badwidth immediately adjacet to a licesed digital televisio trasmissio is a viable practice, give the expectatio that WSD desigs will place a emphasis o the protectio of televisio broadcasts. Device desig challeges specific to white space operatio iclude RF power amplifier liearity, sigal filterig implemetatio, developmet of suitable cogitive properties, ad refiemet of adaptive capabilities to provide the maximum amout of protectio to DTV broadcasts. Focusig research ad developmet o these challeges will help to esure the successful coexistece of primary ad secodary users i the televisio spectrum. Ackowledgmets The authors dedicate this work to the memory of their colleague ad metor, James A. Jim Roberts. His impressive

12 12 EURASIP Joural o Advaces i Sigal Processig Table 4: (a) Desired chael power = 28 dbm/6 MHz; miimum udesired sigal IM 3(DTV-T) ; (b) desired chael power = 28 dbm/6 MHz; sigificat udesired sigal IM 3(DTV-T). (a) ATSC A/74 Receiver 1 (1999 model year set-top) 2 (32 LCD DTV) 3 (Low cost set-top) Rhodes-sgrigoli example (Figure 4; Iterferece Mitigatio for Improved DTV Receptio ) 1 filtered udesired 1 CH IP 3(DT V) Levels required to degrade desired chael to TOV + 1 filtered udesired + 1 CH IP3 DTV IP 3(DT V) (b) 1 filtered udesired + 1 filtered udesired Receiver 1 CH 30 IP 3(DT V) + 1 CH 32 IP3 DTV 1 (1999 model year set-top) 2 (32 LCD DTV) 3 (Low cost set-top) ATSC A/ Levels required to degrade desired chael to TOV career ecompassed a array of accomplishmets ragig from early work for the NASA Apollo program to Presidet ad Chief Operatig Officer of the Uiversity of Kasas Ceter for Research. Durig his teure at the Uiversity of Kasas, Jim s outstadig leadership ad research skills touched all of them at KU; his directio ad support served to elevate their idividual efforts, resultig i umerous collective achievemets. The work preseted here is but oe small example. Jim s legacy of talet, dedicatio, ad friedship, combied with the maer i which he lived his tragically shorteed life ad bravely faced the ed, stads as a shiig example to follow. This work was supported by NSF grats ANI , ANI , ad grats from the New America Foudatio. Refereces [1] M. A. McHery, NSF spectrum occupacy measuremets project summary, Shared Spectrum Compay, August [2] Federal Commuicatios Commissio, Ulicesed operatio i the TV broadcast bads, ET Docket No , May [3] The XG Visio: Request For Commets, versio 2.0, Tech. Rep., BBN Techologies. [4] C. Cordeiro, K. Challapali, D. Birru, ad N. Sai Shakar, IEEE : the first worldwide wireless stadard based o cogitive radios, i Proceedigs of the 1st IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 05), pp , Baltimore, Md, USA, November [5] A. E. Leu, K. Steadma, M. McHery, ad J. Bates, Ultra sesitive TV detector measuremets, i Proceedigs of the 1st

13 EURASIP Joural o Advaces i Sigal Processig 13 IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 05), pp , Baltimore, Md, USA, November [6] E. Visotsky, S. Kuffher, ad R. Peterso, O collaborative detectio of TV trasmissios i support of dyamic spectrum sharig, i Proceedigs of the 1st IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 05), pp , Baltimore, Md, USA, November [7] B. Wild ad K. Ramchadra, Detectig primary receivers for cogitive radio applicatios, i Proceedigs of the 1st IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 05), pp , Baltimore, Md, USA, November [8] M. McHery, The probe spectrum access method, i Proceedigs of the 1st IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 05), pp , Baltimore, Md, USA, November [9] V. R. Petty, R. Rajbashi, D. Datla, et al., Feasibility of dyamic spectrum access i uderutilized televisio bads, i Proceedigs of the 2d IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 07), pp , Dubli, Irelad, April [10] America Federatio of Musicias, Commets i respose to ET docket o , October [11] Wireless Iteret Service Provider s Associatio, Commets i respose to ET docket o , March [12] R. Rajbashi, A. M. Wygliski, ad G. J. Mide, Adaptivemode peak-to-average power ratio reductio algorithm for OFDM-based cogitive radio, i Proceedigs of the 64th IEEE Vehicular Techology Coferece (VTC 06), pp , Motreal, Caada, September [13] S. B. R. Cabric ad D. Mishra, Implemetatio issues i spectrum sesig for cogitive radios, i Proceedigs of the 38th Asilomar Coferece o Sigals, Systems ad Computers, vol. 1, pp , Pacific Grove, Calif, USA, November [14] C. W. Rhodes, Iterferece betwee televisio sigals due to itermodulatio i receiver frot-eds, IEEE Trasactios o Broadcastig, vol. 51, o. 1, pp , [15] H.R.2015-Balaced Budget Act of 1997, Title III-Commuicatios ad Spectrum Allocatios Provisios, thomas.loc.gov/cgi-bi/query/f?c105:1:./temp/ c105lvlw HQ:e21719:. [16] T. X. Brow, A aalysis of ulicesed device operatio i licesed broadcast service bads, i Proceedigs of the 1st IEEE Iteratioal Symposium o New Frotiers i Dyamic Spectrum Access Networks (DySPAN 05), pp , Baltimore, Md, USA, November [17] D. Predergast, M. Guillet, B. Caro, et al., The effects of public safety mobile systems operatios (i TV chaels 63/68) o DTV ad NTSC broadcastig, IEEE Trasactios o Broadcastig, vol. 51, o. 1, pp , [18]C.W.RhodesadG.J.Sgrigoli, Iterferecemitigatio for improved DTV receptio, IEEE Trasactios o Cosumer Electroics, vol. 51, o. 2, pp , 2005.