Government Information Quarterly 29 (2012) 532–542

Contents lists available at SciVerse ScienceDirect

Government Information Quarterly

j ourna l homepage: www.e lsev ie r .com/ locate /gov inf

The wireless abyss: Deconstructing the U.S. National Broadband Map

Tony H. GrubesicGeographic Information Systems and Spatial Analysis Laboratory, College of Information Science and Technology, Drexel University, Philadelphia, PA 19104, USA

E-mail address: grubesic@drexel.edu.

0740-624X/$ – see front matter © 2012 Elsevier Inc. Aldoi:10.1016/j.giq.2012.05.006

a b s t r a c t

a r t i c l e i n f o

Available online 31 July 2012

Keywords:WirelessBroadbandUncertaintyNational Broadband MapPolicyGIS

The U.S. National Broadband Map (NBM) is arguably the most complex articulation and synthesis of telecommu-nications data ever generated by the federal government. Drawing upon information collected by fifty U.S. states,five territories and the District of Columbia, broadband provision is tabulated at the Census block level and madeavailable to the general public in a variety of formats (e.g.,maps, tabular databases, and geographic coverages). Onemajor policy challenge associated with deepening our understanding of wireless broadband provision in theUnited States is developing a methodological process for accurately rearticulating NBM wireless data collected atthe block level tomoremeaningful economic units (e.g., Census block groups or tracts).Without this ability, policyanalysis and an objective evaluation of the goals set forth in the National Broadband Plan are compromised. Thepurpose of this paper is to outline such a methodology, while simultaneously highlighting several consistencychecks for ensuring completeness and data aggregation integrity.

© 2012 Elsevier Inc. All rights reserved.

1. Introduction

The National Broadband Map (NBM) is a collective effort betweenthe National Telecommunication and Information Administration(NTIA), the Federal Communications Commission (FCC), fifty states,five territories, and the District of Columbia to provide a detailedsnapshot of broadband provision in the United States. According tothe FCC, broadband in the United States constitutes download (i.e.,to the customer) speeds of at least 4 megabytes per second (Mbps)and upload (i.e., from the customer) speeds of at least 1 Mbps (FCC,2010a,b). To confuse matters, the NTIA defines broadband as768 Kbps download and 200 Kbps upload speeds. Both definitionshave evolved significantly over the past decade, when as recently as2004 the FCC defined broadband as download speeds of at least200 Kbps (FCC, 2004). Clearly, as delivery technologies continue toevolve and improve, so too will the definition of broadband.

The NBM is a small, but significant facet of a much larger NationalBroadband Plan (“Plan”) (FCC, 2010b). The Plan outlines a strategicagenda for both developing and enhancing broadband infrastructurebecause of its perceived importance to a variety of critical sectors inthe U.S. economy (e.g., health care, education, energy) as well asgovernment performance, public safety and civic engagement (FCC,2010b). Understandably, one of the major challenges set forth in thePlan was to determine broadband provision levels throughout theUnited States. As noted by Grubesic (2012), issues of informationasymmetry have plagued broadband-related development efforts inthe U.S. for many years. One major reason that asymmetries exist isthe lack of quality data regarding broadband provision, pricing and

l rights reserved.

quality of service (QOS) at the local level (Greenstein, 2007). Forexample, prior to the release of the NBM, the only viable broadbandprovision data available to analysts was the FCC Form 477 database,which had been aggregated to the ZIP code or Census tract level.While these data were suitable for a rough snapshot of advanced tele-communications provision, their lack of spatial resolution was asignificant hindrance to understanding disparities in broadband andrelated competitive effects. In addition, because the Form 477 datadid not include pricing information, robust evaluations of theeconomic impact of broadband and associated telecommunicationspolicies were difficult.

Sadly, while pricing information is still unavailable in the NBM, thespatial resolution of provision data is greatly improved. Detailed provi-sion data are now available at the Census block level (e.g., providers,upload/download bandwidth, etc.), which is the smallest geographicunit that the Census Bureau publishes decennial survey informationon. As noted by Grubesic (2012), however, there are several majorproblems with the NBM data. First, provider participation in the NBMvaried significantly between states, ranging from 27% (Virginia) to100% (Indiana, Illinois, and six others). Second, issues of data uncertain-ty for digital subscriber line (xDSL) service were not rectified, likelyleading to a significant overestimation of broadband xDSL provisioncoverage in the U.S. Third, the NBM currently identifies thousands ofzero population blocks (i.e., no residents or businesses) as havingbroadband. In part, these errors can be attributed to providers claimingan ability to provide service to these eligible locations within a 10 daywindow (NTIA, 2011), but as Grubesic (2012) notes, there is an impor-tant difference between regions that could have broadband and regionsthat have broadband. Finally, the sheer size of the databases associatedwith the NBM creates problems. For example, the wireline providerdatabase contains approximately 12.5 million records. As a result, any

Table 1Wireless broadband overview.

Technology Air interfacea Data rate (Mbps)

Downlink Uplink

WiMAX OFDM, OFDMA 75 75UMTS-TDD TDMA 16 163GPP LTE OFDM, OFDMA 100 50CDMA2000/EVDO FDMA 3.1 1.83GPP2 ultra mobile broadband OFDMA 275 75MBWA OFDMA 1 1

OFDMA = Orthogonal frequency-division multiple access.TDMA = Time division multiple access.FDMA = Frequency division multiple access.Source: Kong, D.T., Liang, P-Y. and Chang, Y. (2009). Wireless Broadband Networks.Wiley: Hoboken, NJ.

a OFDM = Orthogonal frequency-division multiplexing.

1 Both Indiana and Ohio have a good mix of urban, suburban and rural settings, pro-viding a fairly representative snapshot of wireless coverage and technology for the U.S.

533T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

effort to aggregate these data to alternative units for analysis is compu-tationally burdensome, time consuming and prone to error.

Although the problems associated with wireline NBM data arefairly well understood, much less attention has been paid to the wire-less NBM data. This is a notable gap because there is growing senti-ment that wireless options may have a disruptive effect on theoverall broadband market, making wireline options (e.g., fiber tothe home) less attractive (Middleton & Given, 2011). This suggests,that now more than ever, developing an understanding of wherewireless broadband options are available is critical to evaluatingdisparities in broadband and evaluating the relative success or failureof the National Broadband Plan over time. Unfortunately, in their cur-rent form, the NBMwireless provision data are both unwieldy and theantithesis of user-friendly. With over 50 million individual recordsfor wireless provision, efforts to manipulate, analyze and visualizethese data at a national scale are both time consuming and computa-tionally intensive. Thus, the purpose of this paper is to provide amethodological framework for rearticulating the raw, wireless broad-band provision data from the NBM to more meaningful economicunits for policy evaluation, spatial econometric analysis andgeographic visualization. Specifically, provision data are aggregatedfrom Census blocks to block groups using a multistep process thatleverages the data manipulation abilities of a geographic informationsystem (GIS). A variety of data consistency checks for ensuringcompleteness and data aggregation integrity are also detailed.

2. Wireless broadband in the United States

Wireless broadband comes in many forms, connecting a home orbusiness to the internet without wires, typically via a radio link be-tween a customer's location and a facility operated by a service pro-vider (Sawada, Cossette, Wellar, & Kurt, 2006). A simple typology todifferentiate between types of wireless broadband is fixed andmobile.There are also subtle differences between platforms that use licensedand unlicensed spectrum (Sirbu, Lehr, & Gillett, 2006). Where the lat-ter is concerned, unlicensed spectrum is shared among internetservice providers, while licensed spectrum is dedicated to a singleprovider. For wireless platforms, fixed technologies allow subscribersto access the internet from a fixed point (while stationary), and usu-ally require a direct line-of-sight between the wireless transmitterand receiver. Fixed wireless technologies include WiFi and WiMAX(Abichar, Peng, & Chang, 2006; Vaughan‐Nichols, 2004) and haveproven to be popular in rural and remote areas where wireline andmobile technologies are not as widespread (Zhang & Wolff, 2004).Mobilewireless connections provide broadband in specific geographiclocations to mobile objects (cars, trucks, boats, pedestrians, etc.)using spectrum that is dedicated to an internet service provider.Mobile wireless technologies include 3GPP Long Term Evolution(LTE) and CDMA2000 (EVDO) among others (Agashe, Rezaiifar, &Bender, 2004; Dahlman, Parkvall, Skold, & Beming, 2008). Finally, itis important to note that satellite broadband technologies fromproviders such as Hughes, Wild Blue and Spacenet are also usedthroughout the United States, although the number of householdssubscribing to satellite services remains very small (~1 million duringthe first quarter of 2010) (NSR, 2010). For a brief overview of wirelessbroadband platforms and their associated speeds, see Table 1.

To put the U.S. wireless market in perspective, consider the recentstatistics published within the National Broadband Plan (FCC, 2010b).Wireless broadband use is growing exponentially, with Ciscoprojecting that wireless networks in North America will carry approx-imately 740 petabytes per month by 2014, a 40-fold increase from2009 (~17 petabytes). In part, this massive increase is attributableto the growing use of smart phones, but it also fueled by the use ofLTE-enabled laptop computers and tablet devices. The FCC (2010a,b,77) also notes that machine-based wireless communications will in-crease dramatically within the next few years, as sensor networks

and “smart devices take advantage of the ubiquitous connectivityafforded by high-speed, low-latency, wireless packet data networks.”

The notion of broadband ubiquity is interesting. While there is nodoubt that the number of smart devices leveraging wireless broadbandnetworks is on the rise, the ubiquity of wireless broadband is less certain(Middleton & Bryne, 2011; Sawada et al., 2006). Fig. 1 provides someperspective on the spatial dimensions of new technology rollouts byprivate telecommunications providers. Specifically, it highlights localesthroughout Indiana and Ohio that have access to the new wireless 4GWiMAX network built by Clear Communications.1 This system isdesigned to provide average download speeds of 3 to 6 Mbps, withbursts up to 10Mbps. Although this does not reflect “true” 4G speeds asdefined by the ITU (100 Mbps) (ITU, 2008), it is representative ofaverage 4G speeds for mobile devices in the U.S. These speeds can sup-port a variety of internet-based activities ranging from streaming audioand video to online gaming (GAO, 2010). Dark green portions of Fig. 1denote areas with excellent coverage and high bandwidth capacities,light green areas have only partial coverage and lower available band-width. Areas with no green shading represent coverage gaps in theClear Communications network. Given this information, it is evidentthat portions of the Cincinnati and Columbus, Ohio metropolitan areashave excellent coverage and high bandwidth capacities (dark green),but the Dayton, Ohio metropolitan area (pop. 847,502) is without anycoverage (Fig. 1) from Clear Communications. Further, the secondlargest metropolitan area in the Midwest, Indianapolis, Indiana (pop.1.83 million), is only partially covered and likely under-capacitated interms of available bandwidth (light green). While early, this geographicperspective on the rollout of the much touted 4th generation networkhighlights underserved regions that might benefit from policy interven-tions if these gaps in provision persist. Granted, this is a geographic snap-shot of a single provider in amixed urban/suburban/rural region, but thenotion of 4G ubiquity remains relatively far-fetched, particularly formore rural areas. Further, there are concerns that the architecture andcapabilities of wired and wireless access networks will never converge,primarily because of the limitations associated with wireless spectrumavailability and its associated capacity (Lehr & Chapin, 2010). That said,the FCC recently moved to open television spectrum to wireless broad-band in the hope of relieving the strain on existing spectrum allocations(Benton Foundation, 2010). Regardless of one's stance on these issues,the ability to identify heterogeneities in provision between urban, subur-ban, rural and remote communities is critical to obtaining a better under-standing of wireless broadband provision and access in the United Statesand supporting the National Broadband Plan.

Stronger Coverage

Weaker Coverage

Fig. 1. Clear Communications 4G WiMAX coverage: portions of Indiana and Ohio.Source: http://www.clear.com/coverage.

534 T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

3. Wireless provision data and the national broadband map

The National Broadband Map details broadband provision infor-mation for wireline and wireless services (including satellite), aswell as documenting “anchor institutions” that can serve as gatewaysfor broadband access (e.g., schools, colleges and libraries).2 Similar tothe wireline data collection and reporting process, wireless broad-band provision data are linked to Census blocks in the nationalmap. Table 1 illustrates a sample of the wireless provision data atthe block level for the state of Ohio and provides the working defini-tion of the fields in the database (USDOC/NTIA, 2010). Perhaps themost interesting column in Table 1 is labeled “pct_blk_in_shape”.The supporting documentation available from the NBM websitedoes not contain any reference to this field, nor does it provide adefinition of it. Upon closer inspection, however, this column beginsto make sense when one deconstructs the methodological processused to generate wireless provision data for Census blocks in theNBM (Table 2).

Unlike the process associated with estimating provision forwireline carriers (Grubesic, 2012), wireless provision at the blocklevel is estimated through a series of interpolation techniques in aGIS environment.3 While there are some minor variations in themethodologies applied and modeling details given by each of theagencies charged with the provision of tabulation in each state, thebulk of them used wireless signal propagation modeling to generatea geographic coverage associated with each tower and their anten-nae. For example, the end product for the state of Ohio “depicts agraphical illustration of the theoretical propagation characteristicsof a selected frequency range based on defined variables (receiversensitivity of the home/mobile device, foliage factor, and digitalelevation terrain input)” (Connect Ohio, 2011). Put more simply, Con-nect Ohio (the grantee charged with developing both wireline and

2 For more details on broadband use in anchor institutions such as public libraries,see McClure, Jaeger, and Bertot (2007) and Jaeger, Bertot, McClure, and Langa (2006).

3 Readers are referred to the NBM methodologies page (http://www2.ntia.doc.gov/files/broadband-data/All_Grantees_Methodologies_December-2010.zip) for further details.

wireless broadband provision data for Ohio) provided the NTIA witha map of wireless coverage, for all of the participating providers, forthe state. Again, the approach used for generating these coveragesvaried somewhat for each state.

There are some rather awkward questions that arise regarding theNBM data once it is realized that these raw geographic coverages ofwireless broadbandprovision are used to estimate broadbandprovisionat the block level. For example, consider Fig. 2, which displays theprovision footprints associated with each of wireless broadband pro-viders in the state of Ohio. Please note that the information representedin Fig. 2 corresponds to the ArcGIS compatible shapefile available fordownload from the NBM site (USDOC/NTIA, 2010). Aside from thediversity in geographic footprints associated with each provider, thereare a number of important observations associated with the underlyingdata thatmust be noted. First, returning to the discussion of Table 1 for amoment, it becomes clear that these wireless coverage footprints areused to generate the values in the “pct_blk_in_shape” column. Forexample, Census block 391336007032011 is served by MikulskiCommunications LLC. However, only 6.98% of this particular block iswithin the Mikulski service footprint. Can this block actually be consid-ered served byMikulski? If not, howmuch coverage is necessary to con-sider a block served? Fifty percent? Seventy five percent or greater?Second, issues ofwireless network ubiquity are also at issue. In this par-ticular instance, in addition to Mikulski, block 391336007032011 is100% covered by three other providers (Cellco [doing business asVerizon], Sprint, and AT&T). That said, there are many others instanceswhere blocks have partial coverage only. For example, consider block390019901001000, also located in Ohio, which is 40% covered by oneScioto Wireless coverage zone, 20% covered by another and 1.9% fromVerizon. Again, does this instance qualify as coverage? Fig. 3 providesa spatial perspective on this issue, highlighting the geographic manifes-tations of coverage for this block. There are notable gaps, but it is impor-tant to remember that these wireless coverages are estimates based oncellular propagation models. Actual coverage may vary. Thus, from apolicy perspective, can one consider this block as having been providedwireless broadband? Regardless of how one might answer these typesof questions, it is clear that a continuum of wireless provision exists

Table2

Nationa

lBroad

band

Map

WirelessBroa

dban

dProv

isionData,

Decem

ber20

10.

FRN

PROVNAME

DBA

NAME

HOCO

NUM

HOCO

NAME

cens

usbloc

k_fips

pct_blk_

in_sha

peTR

ANST

ECH

MAXADDOW

NMAXADUP

TYPICD

OW

NTY

PICU

Pob

jectid

1711

6831

MikulskiC

ommun

ications

LLMikulskiN

et90

0470

MikulskiC

ommun

ications

LLC

3913

3600

7032

011

0.06

9870

43

00

3-1-39

-6-22

1711

6831

MikulskiC

ommun

ications

LLMikulskiN

et90

0470

MikulskiC

ommun

ications

LLC

3913

3601

0001

020

0.78

8670

43

00

3-1-39

-6-22

1711

6831

MikulskiC

ommun

ications

LLMikulskiN

et90

0470

MikulskiC

ommun

ications

LLC

3913

3600

8003

014

0.67

6370

43

00

3-1-39

-6-22

1979

2696

Mechc

omDot

Net

Mechc

omDot

Net

9005

41Mechc

omDot

Net

3913

9003

0022

004

0.02

1570

55

55

3-1-39

-6-23

1979

2696

Mechc

omDot

Net

Mechc

omDot

Net

9005

41Mechc

omDot

Net

3913

9001

0009

024

0.06

2470

55

55

3-1-39

-6-23

1979

2696

Mechc

omDot

Net

Mechc

omDot

Net

9005

41Mechc

omDot

Net

3913

9002

1014

005

0.64

7370

55

55

3-1-39

-6-23

FRN—FC

CRe

gistration

Num

ber.

PROVNAME—

Prov

ider

Nam

e.DBA

NAME—

Doing

Busine

ssAsNam

e.TR

ANST

ECH—Te

chno

logy

Code

(see

below

forva

lidva

lues).

MAXADDOW

N—Max

imum

Adv

ertisedDow

nloa

dSp

eed(see

below

forva

lidva

lues)from

reco

rdleve

l.MAXADUP—

Max

imum

Adv

ertisedUploa

dSp

eed(see

below

forva

lidva

lues)from

reco

rdleve

l.TY

PICD

OW

N—Ty

picalD

ownloa

dSp

eed(see

below

forva

lidva

lues).

TYPICU

P—Ty

picalU

ploa

dSp

eed(see

below

forva

lidva

lues).

DOW

nloa

dSPE

ED—Max

imum

Adv

ertisedDow

nloa

dSp

eedifprov

ided

from

Ove

rview

table.

UPL

OADSP

EED—Max

imum

Adv

ertisedUploa

dSp

eedifprov

ided

from

Ove

rview

table.

Source:U.S.D

eptof

Commerce,N

ationa

lTelecom

mun

icationan

dInform

ationAdm

inistration,

StateBroa

dban

dInitiative

(Decem

ber30

,201

0).

535T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

throughout the United States, manifesting at a highly local level. Thisissue will be discussed more thoroughly later in the paper.

There are several additional aspects of the wireless broadband provi-sion data from the NBM that are worth highlighting. As noted previously,there are 50,013,289 records forwireless broadbandprovision in theNBMdatabase. Needless to say, this is a massive amount of information and isextremely difficult to manipulate, visualize and analyze without a signifi-cant commitment of computational resources. The sheer size of this data-base also makes aggregation of wireless broadband provision data tomore meaningful administrative units such as block groups, transporta-tion analysis zones, or municipal boundaries difficult. Granted, analystscould break the national database into more manageable file sizes, andto their credit, the NTIA has already done this for the state level. However,to conduct any national-level analysis, this means analysts mustre-aggregate state-level data to compile a national database. Moreover,this process still does not address the question of how to efficiently aggre-gate 50 million block data records to alternative geographical units. Thereason this is so important is becausemuch of the truly rich demographic,socio-economic and business data are only available from the Census Bu-reau and other agencies at larger levels of aggregation. Further, becausethere are so many different types of administrative boundaries (e.g., con-gressional districts, school districts, state legislative districts, urbanizedareas, etc.) that are not neatly tied to the Census administrative hierarchy,more simplistic aggregation approaches that are available in statisticalanalysis packages like SAS and SPSS may not be viable. So, if one is inter-esting in modeling the economic impacts of wireless broadband(Thompson & Garbacz, 2011), particularly at a local level, both provisiondata and information on the demographic/socio-economic/business com-position of an administrative unit are required. Further, an efficient andreliable methodology for rearticulating the raw wireless provision datamust be developed. In the next section, a process is outlined to makethis more manageable for analysts interested in policy and spatial econo-metric analysis of wireless broadband in the United States.

4. Data and methodology

As highlighted in the previous section, two databases wereacquired from the National Broadband Map for analysis. The first isthe complete U.S. broadband availability database, by comma sepa-rated values (CSV) (USDOC/NTIA, 2010). The second is the completeU.S. broadband availability database, by shapefile (SHP), a populargeospatial vector data format compatible with a wide range of GISsoftware (ibid). Both databases correspond to the December 2010NBM records on broadband provision.

There are some fundamental differences between these two filesthat are worth detailing. For example, the raw version of the CSV fileon broadband provision from the NBM contains 50,013,289 individualrecords at the block level. Considering that there were only 8,262,363Census blocks in the United States during 2000, which is the vintageof the geographic base files from the Census used for tabulating theNBM (Grubesic, 2012), this equates to an average of six provisionrecords for each block. Again, the structure of this database ishighlighted in Table 1. This inflation of raw record counts occurs be-cause the NBM contains a record for each instance of provision foreach block. Thus, while some blocks in the United States do not appearin the NBM database, others appear multiple times. The SHP file forbroadband provision contained 29,088 unique wireless coverages inthe United States, but these coverages were composed of thousands ofindividual and/or multipart polygons. Multipart polygons refer to cov-erages that consist of non-contiguous (i.e., spatially separate) compo-nents that are stored as a single feature. For example, T-Mobile'swireless coverage in Ohio consists of two unique coverages, one witha maximum download speed of 4 and a maximum upload of 2 andthe other with a maximum download of 6 and a maximum upload of4. Other providers in the state have up to four different coverages. Toput this in perspective, NBM records in the state of Montana included

0 40 80 120 16020Kilometers

Wireless Broadband Coverage

Cincinnati

Cleveland

Fig. 2. Wireless broadband coverages for the state of Ohio, December 2010. Source: USDOC/NTIA (2010). U.S. Dept of Commerce, National Telecommunications and InformationAdministration, State Broadband Initiative (SHP format December 30, 2010).

536 T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

16,258 unique coverages for Montana Internet Corporation alone.While there was certainly some variation in speed between these cov-erages, a closer examination of the data for Montana suggests that themountainous terrain throughout the state helped generate thousandsof non-contiguous patches of wireless coverage.

Given this data ecosystem within the NBM, analysts are basicallyforced tomake a choice between: (a)managing 50 million block recordsin a CSVfile and determining away to aggregate these data to alternative

0 0.2 0.4 0.60.1

Fig. 3. Marginal wireless broadband coverage for a census block in rural Ohio. Source: USDOAdministration, State Broadband Initiativ

administrative units such as block groups, congressional districts oranother geographic unit without double counting provision instances,or, (b) developing a strategy for handling 29,088 geographic coveragesfor accomplishing the same task. Considering that the CSV file tooknearly 40 min to open in SPSS (while caching locally) on a sixteen core,Intel Xeon system with 8 gigabytes of memory, the latter option seemssensible, but is certainly not the only option. One major advantage todealing with geographic coverages in the NBM is that they served as

0.8Kilometers Verizon 1

Scioto Wireless 1

Scioto Wireless 2

Block 390019901001000

C/NTIA (2010). U.S. Dept of Commerce, National Telecommunications and Informatione (SHP format December 30, 2010).

WirelessNBM Data: GeographicCoverages

by State

Disentangle Multipart Provider

Polygons

Start

Generate unified wireless

broadband provision coverage

layer

Repair Polygon Geometry

ArcGIS Explode Command

Automated inspection of each feature in a wireless

coverage for geometry problems. Upon discovery of

a geometry problem, a relevant fix will be applied

ManualRecord

Correction

Reconstruct Wireless

CoveragesArcGIS Dissolve Command

Are ProvidersConsistent?

No

Compile Unique Wireless

Coverages

Yes

ArcGIS Dissolve Command

Aggregate to Census Unit?

Evaluate GeographicIntersection

CensusBlock

Groups,Tracts, etc.

End

No

ArcGIS Assign Data bySpatial Location Command (e.g. complete containment,

intersection, etc.)

ArcGIS Merge Command

Fig. 4. Aggregation strategy for NBM wireless broadband data.

537T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

the foundation for determining broadband provision to Census blocks intheCSVfile. Thus, choice (b)will providemore control over the rawNBMdata for the rearticulation process to alternative geographic units, al-though there are some computational complexities that make this op-tion challenging as well. These will be discussed in the next section.

4.1. Deconstructing NBM wireless coverages

Fig. 4 details the multistep process used for rearticulating the NBMwireless data aggregation to alternative economic units for policy andspatial econometric analysis. The purpose of this subsection is to provide

538 T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

some contextual details on the process, highlighting several potentialpitfalls for analysts when attempting to compile a comprehensive wire-less database using NBM data.

The first step of this process is tomerge all of the separate SHP files(one for each state) published by the NBM to create a single geo-graphic coverage for the United States. By creating a single, masterdatabase of wireless broadband provision for the U.S., one will notneed to recreate subsequent steps for each state individually.4 Theend product is a GIS shapefile that consists of all fifty states (andthe District of Columbia) and their associated wireless coverages.

The second step is to explode each of the individual wireless cover-ages associated with each provider in the master database. Thisserves two purposes. First, because the raw NBM data consisted ofhundreds (and sometimes thousands) of individual coverages forthe same provider in certain states (e.g., Montana), not accountingfor these duplicitous, multipart polygons could create aggregationproblems (i.e., double counting) later in this process. Second, thiswill allow analysts to create a single, nationwide coverage for eachprovider in the NBM database (e.g., Verizon), rather than multiplecoverages from single providers in all 50 states. In a nutshell, the ex-plosion process will help reduce data redundancy and improve thegeographic consistency of wireless provision information.

The purpose of the third step is to improve the geometric charac-teristics of each provider's wireless coverage area. In this context, it isimportant to remember that the shapefile (.shp) is an open format towhich many software packages write. As noted by ESRI (2012), errorsare common because of bugs in the software or programs not follow-ing the documented specification of the shapefile format. Because theNBM consists of data collected from a diverse range of agencies across50 states, bugs and errors in the generated shapefiles are virtuallyguaranteed. Therefore, one way to ensure the objective evaluationof wireless broadband provision when using NBM geographic cover-ages is to mend geometric errors in the database. This is an automatedprocess that both identifies and repairs geometric inconsistencies.Specifically, ArcGIS version 10, a commercial geographic informationsystem, was used to identify and repair errors in the wireless cover-ages.5 In total, 181 errors in wireless coverage geometry were foundfor the 50 states and the District of Columbia and all of them wererepaired successfully.6

Once the geometry was repaired for each of the wireless cover-ages, individual coverages were re-assimilated into a single coveragefor each provider in the United States. Specifically, using the dissolvecommand in ArcGIS allowed for each of the individual polygons asso-ciated with a provider to be compiled into a single polygon based onthe operating carrier's name. Although using provider names is not anideal option for conducting this type of geoprocessing, there is a com-pelling reason for this approach. Surprisingly, the wireless NBM data-base does not contain column that assigns a unique identificationnumber to individual providers. Worse, although the Federal Regis-tration Number (FRN) should serve as a unique identification foreach provider, it does not. For example, in the state of Ohio, thereare seventeen providers that have a FRN code of 9999.7 Holding com-pany number (HOCONUM) is not a viable option either because it isnot included as a field in the wireless coverage shapefile. As a result,

4 If the study area of interest consists of a single state, this step is unnecessary.5 For more details on the identification and repair process for geometric errors in

ArcGIS, see http://tinyurl.com/czggd4n.6 The majority of these errors were polygon “self intersections,”where areas of over-

lap within a polygon are dissolved. Other fixes included the removal of null or emptygeometries and incorrect ring ordering in the polygon (where its boundaries are mixedclockwise and counterclockwise orientations). For more details and a general overviewof geometric errors in spatial data coverages, see Ubeda and Egenhofer (1997).

7 Wilkshire Communications, Wabash Communications, Slane Telecom, SkymaxBroadband, SAA bright.net, Redbird Internet Services, RAA Services, Mango Bay Com-munications, LightSpeed Technologies, Jenco Wireless, g wireless, Coyote WirelessBroadband, Champaign Telephone, Blu SkyWireless, Access Ohio Valley, Computers4U,and Avolve.

the only columns which are remotely close to being unique in the NBMdatabase are PROVNAME or DBANAME. For this paper, PROVNAMEwas utilized to ensure that providers doing business with differentnames in a region were not double counted (e.g., Clearwire Corporationas “Clearwire” and/or “Clear”). Alas, using the PROVNAME column wasdifficult due to numerous spelling errors and typos. For example, therewere many instances where “T-Mobile USA, Inc” was entered as“T-MobileUSA Inc” (missing a comma). Similar errors for other providerswere also found. Individually, these errors are not hugely problematic,but collectively, they all required correction before the final uniquewire-less coverages were compiled for the U.S.

Once the manual process of fixing errors in the PROVNAME col-umn was complete, the ArcGIS dissolve command is used for thefinal time to compile the unique wireless coverages for each providerin the United States. The end product is a shapefile that consists of819 unique wireless coverages for the U.S. that is geographically con-sistent, geometrically correct and do not contain any duplicitous en-tries.8 The final wireless coverage map for the United Statesincludes all terrestrial fixed wireless (licensed and unlicensed) aswell as terrestrial mobile wireless (Fig. 5). Furthermore, not only isit an exact cartographic match to the map displayed by the NationalBroadband Map site for 2010 data, but it consists of a radically simpli-fied underlying database with a single coverage for each unique pro-vider. From this point, the aggregation process is relativelystraightforward. For example, to count the number of unique wirelessbroadband providers for each block group in the United States, a com-mercial GIS system must evaluate individual wireless coverage inter-sections (n=819) with each of the block groups (n=207,507),yielding nearly 170 million combinations. Fortunately, ArcGIS is rela-tively efficient at this, although it does take time and resources. Theentire process required 23 h to complete on a 24 core IBM serverwith 16 gigabytes of RAM. It is important to note that computationaleffort increases/decreases linearly with the number of administrativeunits.

Given these computational challenges, one must ask if the tradeoffassociated with geoprocessing 170 million record combinations hasany advantages over processing 50 million in standard tabular for-mat. Recall that the primary motivation for the spatial deconstructionof the wireless provision data is because of the added flexibility andcontrol it offers for data aggregation to alternative units. Again, oncethe process is complete, one can aggregate to block groups, tracts,ZIP codes, counties, school districts, state legislative districts or anyother unit of interest. Further, because this process maintainsprovider coverages in their native format, it allows for geometric in-consistencies to be identified and enables analysts to explore alterna-tive coverage operators. Specifically, there are a number of options forevaluating the presence of wireless coverage, including operators forintersection, contain, within, and closest functions (Table 3). For thepurposes of this paper, an intersection rule was applied. That is, if awireless coverage shape intersected a block group, it was considered“covered” by the wireless provider. This is a relatively liberal defini-tion of coverage that ensures the inclusion of any provider with apresence in a census block group. Rightly or wrongly, this isessentially how the NBM captures wireless provision (recall the ex-ample illustrated in Fig. 3). A more strict interpretation of wirelessprovision would be to consider a block group covered only when aprovider's geographic service area completely contains the blockgroup in question. This would effectively eliminate the partial cover-age questions associated with providers illustrated in Fig. 3 and

8 For the purposes of this paper, satellite providers were excluded from the final da-tabase. There are two reasons for this. First, many of the geographic coverages associ-ated with these providers were complete state boundaries. Second, because somestates included these boundaries in their report to the NBM and others did not, the sat-ellite broadband provision data in the NBM are highly inconsistent. These data uncer-tainties combined with the relatively small market for satellite broadband make itsexclusion relatively low impact when evaluating provision nationally.

Fig. 5. Simplified wireless broadband coverage map derived from NBM data: December 2010.

539T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

provide an interesting, alternative perspective on wireless ubiquityfor the U.S.

5. Results and discussion

Fig. 6 illustrates the derived representation of wireless broadbandprovision in the United States at the block group level by standarddeviation. Specifically, this map is the final product of all the errorchecks and GIS manipulations outlined previously. The descriptivestatistics for wireless broadband provision by block group are:max=15, min=0, average=4.82 and a standard deviation of 1.73.The corresponding geographic patterns are not particularly surpris-ing; wireless broadband provision options are strongly tied to majormetropolitan areas and transportation corridors. However, there areseveral interesting anomalies in this wireless landscape. First, thereappears to be an agglomeration of providers along some state borders(e.g. Idaho and Washington; California and Nevada). In the case ofIdaho and Washington, there are a number of regional service hubsin this corridor, including the university cities of Pullman (Washing-ton State University) and Moscow (University of Idaho) as well as

Table 3Overview of selected spatial selection operators.

Operator Explanation

Within a distance of Creates buffers around a source feature and returns all the featurbuffer zones

Within The geometry of the target feature must fall inside the geometryOverlapping boundaries are permitted.

Completely within To be selected, all parts of the target features must fall inside thefeature. Touching of boundaries is not allowed.

Contain The geometry of the source feature must fall inside the geometryincluding its boundaries.

Completely contain All parts of the target feature must completely contain the geomeTouching of boundaries is not allowed

Centroid in A target feature is selected if the centroid of its geometry falls intfeature or on its boundaries

Intersect Returns any feature that either fully or partially overlaps the sour

the river ports of Clarkston, Washington and Lewiston, Idaho. All ofthese cities are connected by major state highways, and are servedby numerous wireless providers. In the case of California and Nevada,there are two facets of local geography at work. First, the presence ofLake Tahoe, numerous ski resorts, Interstate 80 and the cities of Reno,NV, Sparks, NV and Truckee, CA certainly enhance wireless provisionin the region. However, the size and geographic extent of blockgroups in this region also impact provider counts. As highlighted byGrubesic (2008), larger administrative units are prevalent in thewestern states and issues of broadband ubiquity and coverage are im-pacted by these regional geographic characteristics. Specifically, theenormous block group (~1162 km2) that basically mimics the bound-ary of Washoe County Nevada (oriented north/south) on the westernborder of Nevada has nine wireless providers and is between 1.5 and2.5 standard deviations above average for wireless broadband provi-sion in the United States. Much like the example illustrated in Fig. 3,this block group is collectingmarginal wireless coverage from a varietyof sources—some spilling in from California, some from Oregon andsome from nearby areas in Nevada. For verification purposes, oneonly needs to consult Fig. 5 to see that this region is a functional

Remarks

es intersecting the Select cities within 100 m of a river

of the source feature. The state of Minnesota can be selected even thoughit shares a boundary with the United States.

geometry of the source All counties within Ohio, but not spatially adjacentto its border would be selected

of the target feature, Inverse of the “within” operator

tries of the source feature. Inverse of the “completely within” operator

o the geometry of the source Works for points, lines and polygons

ce feature Can be used for point, line and polygon coverages

Fig. 6. Rearticulated wireless broadband provision by block group (Mean=4.82).

540 T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

dead-zone for wireless coverage, yet, the aggregation process hasattributed 9 providers to this block group. There are other instancesof these geographic aggregation errors for Death Valley, CA, the Na-tional Grasslands in Northern Colorado and the Grand Canyon (Arizo-na and Nevada border) among others. All of these large block groupshave collected marginal coverage from numerous providers in/around their borders.9

Clearly, this presents a significant problem for policy and spatialeconometric analysis of wireless broadband provision in the UnitedStates.While thewireless provider counts illustrated in Fig. 6 are accurate,it would be wrong to suggest that these marginally covered block groupsrepresent “core” wireless regions. If anything, the bulk of these areaswould be at the opposite end of the continuum—sparsely covered withpoor quality of service. Given this landscape and theproblemshighlightedwith the National Broadband Map data, where does this leave theNational Broadband Plan and telecommunications policy?

First, it is important to note that these aggregation errors are notfatal. Where spatial econometric analysis is concerned, the unifiedprovider coverages that are displayed in Fig. 5 can be used to tagregions exhibiting marginal geographic coverage. For example,cross-sectional econometric models that are using economic unitssuch as block groups can differentiate geographic space with spatialregime (i.e., dummy) variables. In this instance, if one is interestedin discriminating between large blocks with marginal coverage andthose with more complete coverage, one can generate a relevantbinary assignment variable (0, 1) as a control mechanism to differen-tiate coverage variation over geographic space (Anselin, 1992;Grubesic, 2003; Mack & Grubesic, 2009). Second, one could alter thecoverage operator. Instead of using intersect, it is possible that some

9 This may also be a function of residual geometric errors in the database. Accordingto the Census Bureau, Cartographic Boundary Files are vector files generalized from1:100,000-scale TIGER Line Files and are designed for use at scales from about1:500,000 to 1:5,000,000 (Census Bureau, 2010). At 1:100,000, the spatial accuracyof block groups would be +/−166.67 ft. Thus, non-matching geographic base files,unintended coverage spillover and marginal wireless coverage could all be contribut-ing to the derived results.

type of contain or within function may be more appropriate for por-tions of the West. Alternatively, one could simply remove theseoutliers/anomalies from the database prior to statistical analysis.This is a reasonable strategy for two reasons. Not only does it elimi-nate bias from statistical models, but it also reflects a hard truth—wireless providers are unlikely to make infrastructure investmentsin these regions.

While the removal of these areas is couched in an analytical con-text, it does hint to the practical problems in 1) developing a ubiqui-tous wireless broadband network for the United States, and 2)obtaining good data for benchmarking network development andgenerating evidence-based policy using the data that are available.Both problems will be discussed in turn.

Where ubiquity is concerned, the architecture and capabilities ofwired and wireless access networks are slow to converge. As notedearlier, one major problem has been the limitations associated withwireless spectrum availability and its associated capacity (Lehr &Chapin, 2010). However, with additional TV spectrum being madeavailable for broadband (Benton Foundation, 2010), the landscape isabout to change. Broadband convergence aside, it is also importantto acknowledge that private firms have difficulty in adequately pro-viding “public goods” such as communications infrastructure in aspatially balanced manner (Grubesic, 2006). Simply put, privatefirms have no interest in making huge infrastructure investments inregions where populations are sparse and there is virtually nodemand for terrestrial or mobile wireless services. As a result, it ispossible that leaving portions of the United States unconnected is aviable national strategy given the tremendous expense associatedwith providing wireless (and wireline) broadband to rural andremote locations. This is undoubtedly a controversial view, but onethat needs to be considered given the economics associated withinfrastructure provision. It is possible that satellite broadband tech-nologies can fill in the gaps, although given the spotty data on satel-lite service provision, availability, its geographic reach and quality ofservice currently prevents analysts from making any definitive state-ments on this alternative.

This is not to say, however, that broadband options in rural areasare a vain hope. Broadband mesh networks are already providing

541T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

service for many of the more remote regions of the country, includingthose with challenging local terrain. Mesh systems are fully wirelessand employ multihop communications to forward traffic to/fromwired internet access points (Bruno, Conti, & Gegori, 2005;Shillington & Tong, 2011). These multihop systems effectively elimi-nate the need for “line-of-site” connections and allow regions withmountainous (or just hilly) topography to receive wireless broad-band. For example, in portions of southwestern Illinois, where theterrain is rolling and line-of-sight wireless systems are impractical,the Illinois Rural Electric Cooperative has employed a wireless broad-band mesh network that leverages internet connections from theutility's power substations to deliver last mile connectivity to coopmembers (IREC, 2012). The last mile connections are essentiallyWiFi, but the installed infrastructure from SkyPilot Networks helpsmitigate interference and subscribers are provided hardware thatsupports a 5 GHz network signal from up to 7 miles from a nodewith dedicated bandwidth (SPN, 2012). Pricing is reasonable, with abasic 512 k upload and 128 k download package for $27 per monthand a 2 MB download and 512 k upload for $60 per month. Of course,activation and installation fees apply (~$250).

Given the results of this paper and previous empirical work re-garding the National Broadband Map (Grubesic, 2012), it is alsoclear that the current iteration of the NBM wireless provision dataleave much to be desired. Aside from the sheer size of the NBM data-base and the inability of average desktop computers and software todeal with 50 million records efficiently, coverage geometries weredetermined to be problematic, there were thousands of duplicitouswireless records and hundreds of errors in the database (misspelledprovider names, lack of unique identification numbers). Not onlydoes this impact one's ability to conduct reliable analysis, it alsomeans that there is no convenient path to develop meaningful evalu-ations of the National Broadband Plan using NBM data, or broadbandtelecommunications policy in general. Again, while data resolution atthe census block is certainly a welcome change from ZIP codes, mostof the relevant demographic, socio-economic and firm-level data areonly available from larger economic units (e.g. block groups, tracts,etc.), forcing analysts to develop aggregation routines to make theNBM data truly accessible.

Finally, given the limitations of these data, can any meaningfulpolicy analysis be generated for wireless broadband provision in theUnited States? This is a difficult question to answer and the resultsof this paper are not meant to condemn nor condone inferencesthat may be drawn from the NBM data. Although the wirelesscoverages provided by the NBM are flawed and are extremely difficultto use (both in tabular and GIS format), they appear to give a relative-ly viable snapshot of where wireless broadband is available. The sub-tleties associated with measures like speed and quality of service aremore difficult to assess. For example, reconsider Fig. 3. If Verizon andScioto wireless are offering different speeds within their respectivecoverages, a subscriber using a mobile device (in a car) that crossesone of the coverage areas will not necessarily see an immediateand/or instantaneous drop-off in speed or service quality. Theinfluence of terrain, tower loads and differences in mobile devicequality can all impact service performance. Finally, the lack of pricingdata on wireless broadband from the NBM is a limitation that isshared with its companion, wireline broadband data. Pricing datamust be made available for policy evaluation. Without it, policy anal-ysis will continue to be hampered by information asymmetries(Grubesic, 2012).

6. Conclusion

The purpose of this paper was to outline a methodological frame-work for rearticulating the raw, wireless broadband provision datafrom the NBM to more meaningful administrative units for policyevaluation, spatial econometric analysis and geographic visualization.

Specifically, provision data were aggregated from Census blocks toblock groups using a multistep process that leverages the data manip-ulation abilities of a geographic information system. However, thetrue utility of the outlined approach is the flexibility to rearticulatethe wireless data to any administrative unit. A variety of data consis-tency checks for ensuring completeness and data aggregation integri-ty were also detailed. Results suggest that the National BroadbandMap is fraught with data inconsistencies, but if tracked carefully,can be managed and mitigated to develop a more accurate represen-tation of wireless broadband provision in the United States. It ishoped that the outlined strategy will make the NBM data more acces-sible to those interested in telecommunications policy evaluation,disparities in communications availability and the growing informa-tion economy. Further, although this paper is somewhat critical ofthe NBM, it is also important to note that these data, while imperfect,represent a major improvement when compared to previous efforts.

References

Abichar, Z., Peng, Y., & Chang, J. M. (2006). WiMax: The emergence of wireless broad-band. IT Professional Magazine, 8, 44–48.

Agashe, P., Rezaiifar, R., & Bender, P. (2004). CDMA2000® high rate broadcast packetdata air interface design. IEEE Communications Magazine, 42, 83–89.

Anselin, L. (1992). Spatial data analysis with GIS: An introduction to application in thesocial sciences. Technical Report 92‐10. Santa Barbara: National Center for GeographicInformation and Analysis, University of California.

Benton Foundation (2010). Opening TV spectrum to new wireless broadband services.http://benton.org/node/45917

Bruno, R., Conti, M., & Gegori, E. (2005). Mesh networks: Commodity multihop ad hocnetworks. IEEE Communications Magazine, 43(3), 123–131.

Census Bureau (2010). Cartographic boundary files. http://www.census.gov/geo/www/cob/scale.html

Connect Ohio (2011). Official April 2011 Update Submission to the NationalTelecommunications and Information Administration Under the State BroadbandData and Development Grant Program for the State of Ohio. Retrieved from.http://www2.ntia.doc.gov/files/broadband-data/All_Grantees_Methodologies_December-2010.zip

Dahlman, E., Parkvall, S., Skold, J., & Beming, P. (2008). 3G evolution: HSPA and LTE formobile broadband (2nd ed.). Academic Press.

Environmental Systems Research Institute [ESRI] (2012). Retrieved from http://resources.esri.com/help/9.3/arcgisengine/java/gp_toolref/data_management_toolbox/checking_and_repairing_geometries.htm

FCC (2004). Availability of advanced telecommunications capability in the United States.Washington, D.C.: Federal Communications Commission Retrieved from b http://www.fcc.gov/broadband/706.html>

FCC (2010a). Mobile broadband: The benefits of additional spectrum. Federal Communi-cations Commission Retrieved from b http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-302324A1.pdf>

FCC (2010b). Connecting America: The national broadband plan. Federal CommunicationsCommission Retrieved from b http://transition.fcc.gov/Daily_Releases/Daily/Business/2010/db0720/FCC-10-129A1.pdf>

GAO (2010). National broadband plan reflects the experiences of leading countries, but im-plementation will be challenging. Report: United States Government AccountabilityOffice: Report to Congressional Requesters http://www.gao.gov/products/GAO-10-825

Greenstein, S. (2007). Data constraints and the internet economy: Impressions and impre-cision. NSF/OECD meeting on factors shaping the future of the internet Retrievedfrom b http://www.oecd.org/dataoecd/5/54/38151520.pdf>

Grubesic, T. H. (2003). Inequities in the broadband revolution. The Annals of RegionalScience, 37, 263–289.

Grubesic, T. H. (2006). A spatial taxonomy of broadband regions in the United States.Information Economics and Policy, 18, 423–448.

Grubesic, T. H. (2008). Spatial data constraints: Implications for measuring broadband.Telecommunications Policy, 32, 490–502.

Grubesic, T. H. (2012). The U.S. national broadbandmap: Data limitations and implications.Telecommunications Policy, 36, 113–126.

Illinois Rural Electric Cooperative [IREC] (2012). http://www.e-co-op.com/International Telecommunications Union [ITU] (2008). Requirements related to technical

performance for IMT-Advance radio interface(s). http://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-M.2134-2008-PDF-E.pdf

Jaeger, P. T., Bertot, J. C., McClure, C. R., & Langa, L. A. (2006). The policy implications of inter-net connectivity in public libraries. Government Information Quarterly, 23(1), 123–141.

Kong, D. T., Liang, P. -Y., & Chang, Y. (2009).Wireless Broadband Networks.Wiley: Hoboken,NJ.

Lehr, W., & Chapin, J. M. (2010). On the convergence of wired and wireless accessnetwork architectures. Information Economics and Policy, 22, 33–41.

Mack, E. A., & Grubesic, T. H. (2009). Forecasting broadband provision. InformationEconomics and Policy, 21, 297–311.

McClure, C. R., Jaeger, P. T., & Bertot, J. C. (2007). The looming infrastructure plateau?Funding, connection speed, and the ability of public libraries to meet the demandfor free internet access. First Monday, 12(12–3).

542 T.H. Grubesic / Government Information Quarterly 29 (2012) 532–542

Middleton, C. A., & Bryne, A. (2011). An exploration of user-generated wirelessbroadband infrastructures in digital cities. Telematics and Informatics, 28,163–175.

Middleton, C. A., & Given, J. (2011). The next broadband challenge: Wireless. Journal ofInformation Policy, 1, 36–56.

NTIA (2011). National broadband map: Data comparison methodology. National Tele-communications and Information Administration. Retrieved from http://tinyurl.com/3sjv37u

Sawada, M., Cossette, D., Wellar, B., & Kurt, T. (2006). Analysis of the urban/rural broad-band divide in Canada: Using GIS in planning terrestrial wireless deployment.Government Information Quarterly, 23, 454–479.

Shillington, L., & Tong, D. (2011). Maximizing wireless mesh network coverage. InternationalRegional Science Review, 34, 419–437, http://dx.doi.org/10.1177/0160017610396011.

Sirbu, M., Lehr, W., & Gillett, S. (2006). Evolving wireless access technologies for mu-nicipal broadband. Government Information Quarterly, 23, 480–502.

Sky Pilot Networks [SPN] (2012). Rural utility delivers broadband services using scal-able wireless mesh. http://tinyurl.com/86cl7e2

Northern Sky Research [NSR] (2010). Satellite Broadband Holding Its Own. Now at 1 Mil-lion+ Subscribers URL: http://tinyurl.com/7h4kang

Thompson, H. G., & Garbacz, C. (2011). Economic impacts of mobile versus fixed broad-band. Telecommunications Policy, 35, 999–1009.

Ubeda, T., & Egenhofer, M. J. (1997). Topological error correcting in GIS. In M. Scholl, &A. Voisard (Eds.), Advances in spatial databases. Lecture Notes in Computer Science,http://dx.doi.org/10.1007/3-540-63238-7.

USDOC/NTIA (2010). US Dept of Commerce, National Telecommunications and Infor-mation Administration, State Broadband Initiative. SHP format December 30,2010.

Vaughan‐Nichols, S. J. (2004). Achieving wireless broadband with WiMax. Computer,37, 10–13.

Zhang, M., & Wolff, R. S. (2004). Crossing the digital divide: Cost-effective broadbandwireless access for rural and remote areas. IEEE Communications Magazine, 42,99–105.

Tony H. Grubesic is an associate professor in the College of Information Science andTechnology and Director of the Geographic Information System and Spatial AnalysisLaboratory (GISSA) at Drexel University. His research and teaching interests are ingeographic information science, regional development and public policy evaluation,critical infrastructure, geospatial intelligence and urban health disparities. Grubesicobtained a B.A. in Political Science from Willamette University, a B.S. in Geography fromthe University of Wisconsin-Whitewater, a M.A. in Geography from the University ofAkron, and a Ph.D. in Geographic Information Science from the Ohio State University.