03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 1 Open Spectrum New Technology / FCC SPTF Impact and Policy Implications Robert J. Berger Glocom Visiting Research Fellow [email protected] 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 2 Introduction  The FCC formed a Spectrum Policy Task Force (SPTF) in June 2002 to identify and evaluate changes in spectrum policy that will increase the public benefits derived from the use of radio spectrum.  This was the first time that there was a comprehensive and systematic review of FCC spectrum policy. 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 3 FCC Interest in Spectrum Policy Reform 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 4 Interest in new spectrum access models  FCC has not kept up with the pace of communication tech and accelerating demand for spectrum  Current “Command + Control” policy and procedures micromanage spectrum  Thus it is outmoded and obsolete  Need “out-of-the-box” ways to allocate and maximize spectrum access  Find ways to maximize public benefits delivered through spectrum based services and devices 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 5 Drivers for Spectrum Policy Reform  Explosive Demand for Spectrum-Based Services and Devices  Technological Advances: Enabling Changes in Spectrum Policy  Increased Access: Mitigating Scarcity of Spectrum Resources 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 6 Explosive Demand for Spectrum- Based Services and Devices  Consistently underestimated demand  1994 projected 54M mobile phone users for 2000  Actual number of users in 2000 was 110M  Unlicensed band (2.4Ghz) spurred explosion of new devices and services  $2.9B IN 2002  New tech allows for devices paid for & controlled by millions of end users  Old policies based on small number of licensees (broadcasters) 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 7 Tech Advances: Enabling Changes in Spectrum Policy  Digital Signal Processing allows for radical new modulation techniques  Wideband Spread Spectrum  Ultra-Wideband pulse  Very low power per hertz  Cognitive / Software Defined Radios  Dynamically and Intelligently utilize and share spectrum  Moore’s Law makes it practical and affordable (and inevitable) 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 8 Increased Access: Mitigating Scarcity of Spectrum Resources  Measurements of actual spectrum utilization in Metro areas showed:  Nearly 100% of spectrum allocated, but only 30% actually used  Looking for new ways to better utilize spectrum  Underlay Spectrum Commons (UWB, Spread Spectrum)  Cognitive Radios dynamically sensing and releasing spectrum  Secondary Markets 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 9 New Technologies Radically new way to utilize & expand the capacity of spectrum 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 10 Some Spectrum Basics image obtained from http://www.lbl.gov/MicroWorlds/ALSTool/EMSpec/EMSpec2.html 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 11 Time vs. Frequency  Time Domain  Wavelength or Pulse Width duration  Viewed with Oscilloscope  Frequency Domain  Cycles / Second  Viewed with Spectrum Analyzer (Spectrograph) Time Frequency 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 12 New tech facilitates sharing beyond 802.11Wireless LAN  Underlay legacy spectrum users  Wideband Spread Spectrum  Ultra-wideband nano-pulses  Pico-watts / Hertz  Intelligently utilize unused local spectrum  Cognitive / Software Defined Radios 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 13 Wideband Spread Spectrum  Trades off Spectrum for power  Wider spectrum produces more sharing and bandwidth  Called Process Gain  Can underlay legacy narrowband users Graphics from Spread spectrum communications by Jay Fitzsummons, Troy Morris and Tony Parezanovic http://murray.newcastle.edu.au/users/staff/eemf/ELEC351/ SProjects/Morris/project.htm 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 14 Direct Sequence Spread Spectrum  Pseudo-noise (PN- code) mixed with Data to produce the coded signal to modulate a carrier  Looks like noise source centered around the carrier with a bandwidth of the Pseudo-noise  Receiver knows PN- code to demodulate signal Graphics from Spread spectrum communications by Jay Fitzsummons, Troy Morris and Tony Parezanovic 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 15 Frequency Hopping Spread Spectrum  Data is conventionally modulated on carrier  SS Bandwidth is carved up to many narrow channels  PN-code selects which channel is utilized as the carrier  Hopping rate is in order of milliseconds / hop thus minimizing interference with legacy narrowband users Graphics from Spread spectrum communications by Jay Fitzsummons, Troy Morris and Tony Parezanovic 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 16 Ultra-wideband nano-pulses  Extremely short pulses instead of carrier waves  10 - 1000 of picoseconds (trillionths of a second) wide in time  1 - 10 Gigahertz wide in frequency  Picowatts of power per hertz (in the noise floor)  Radios can create output signal directly with digital techniques  High precision timing, but low complexity Graphics from Scientific American: Wireless Data Blaster by David G. Leeper 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 17 Ultra-wideband Modulations  Many ways to modulate pulse streams  No Multipath fading  Main issue is precision synchronization  Applications  Communications  Sub-centimeter positioning  Thru-wall/ground radar Graphics from Scientific American: Wireless Data Blaster by David G. Leeper 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 18 Ultra-wideband Status  Feb 2002: US FCC allowed limited use  Less than Part 15 levels below 3.1Ghz  Some restrictions on applications  Several Chip Vendors  Some samples  XtremeSpectrum  Most announcements for mid to late 2003 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 19 Cognitive / Software Defined Radios  Cognitive radio “understands” local conditions and user requirements  Will aggregate bands of spectrum that may be allocated but not being used locally  Software Defined Radio (SDR)  Radio signal modulated/demodulated in software  Can create arbitrary signals  Could be Spread Spectrum, UWB or traditional 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 20 Status of SDRs  Some commercial implementations  Very Limited Applications like multi-band / multi- standard cell phones  Vanu Inc.  SDR Software Developers Kit  Gnu-Radio  Open Source SDR  Military most advanced  DARPA NeXt Generation Communications 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 21 Mesh Networks  End points can relay through other user nodes  Low Power  Route around obstacles  Cooperation Gain  Total Capacity of Mesh increases with added users / relay nodes  Matches low power / high process gain tech like Spread Spectrum & UWB 0 12.5 25.0 37.5 50.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0 3.3 6.5 9.8 13.0 Mesh Network Capacity vs Station Density Total Capacity Number of Stations Per Station Capacity 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 22 Status of Mesh Networks  Limited commercial deployment  Nokia Rooftop  First generation product very limited throughput, proprietary and expensive.  For residential / infrastructure use only  http://www.wbs.nokia.com/  Mesh Networks Inc.  Initial product proprietary  Promising an 802.11 based product that supports infrastructure & end user relaying  http://www.meshnetworks.com/  Long history of Military development 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 23 The SPTF Report The Groundbreaking FCC Spectrum Policy Task Force Report 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 24 Rather Incredible Document from a Federal Bureaucracy  9 months from start to finish  Introduces and promotes several concepts that would have been unthinkable only a year ago  Not perfect, several contradictory positions  Will mark a milestone for new regulatory thinking 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 25 Spectrum Rights Models  Command and Control  Current style of regulatory policy  Exclusive Use  Spectrum as Private Property  Spectrum Commons  Technology used to share and manage spectrum 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 26 Command and Control Model  Traditional management of spectrum for the last 80 years  Government agency micromanages all spectrum allocation  FCC  National Telecommunications and Information Administration (NTIA)  Spectrum allocated to a specific entity for a specific use and specific technology  Little or no flexibility how licensees can utilize spectrum 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 27 Exclusive Use Model  Allocates spectrum as property  Spectrum holder can do whatever they want with it  Within the power and interference technical requirements  Can lease/resell all or portions to create secondary markets  Economists believe this will evolve spectrum to its “Highest Value” 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 28 Spectrum Commons Model  Unlimited unlicensed users share spectrum via technological mechanisms  802.11 WLANs proved the value  Already US$2B Industry and growing rapidly  Still rules and limitations on how Spectrum is used  Power per hertz, freq range, geographical, etc.  Marketplace of devices, services and technology 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 29 Creating a Spectrum Commons 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 30 Sharing Spectrum thru Technology, Not Politics  Spectrum is infinitely divisible  Tech determines physical access to usable spectrum  Tech possibilities are just beginning  Most legacy systems are based on early 20th century technology  Cell phones and 802.11 use primitive sharing  Multi-dimensional real-time sharing by space, frequency, time, coding, mesh has no comparison to today’s limited capacity Data Link (MAC) Network Transport Session Presentation Application Physical Economics Politics 802.11 IP TCP/UDP 2 3 4 5 6 7 1 8 9 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 31 FCC SPTF Commons Suggestions  A Low Power Underlay  Spread Spectrum, UWB, and other tech to utilize new capacity in the noise floor of legacy systems  Interference Temperature defines noise floor  Dynamic Reuse of idle spectrum  Sense and utilize local spectrum being unused by primary, release as soon as primary uses it.  Geographical / Guard Bands  Max power determined by local conditions + rules set by FCC and/or primary licensee 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 32 Interference Temperature  A paradigm for assessing the interference in an environment  A quantitative measurement that allows for technology based access control to spectrum  Measures the RF power available at the receiving antenna per unit bandwidth. 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 33 Interference Temperature  Metric to establish maximum permissible levels of interference  Characterizes the worst case environment in which a receiver would be expected to operate.  Different threshold levels could be set for each band, geographic region or service, 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 34 Creating an Underlay Commons Distance from licensed transmitting antenna Power at Receiver 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 35 Agile Radio Enables Dynamic Sharing  Also known as Cognitive or Software Defined Radio (SDR)  Dynamically discovers + utilizes local unused spectrum in real-time  Takes advantage of “White Space” in spectrum allocations  Releases slices of spectrum if primary licensee starts to use it also  Scales power based on application, local condition and rules set by FCC and/or Primary Licensee  Combine with Spread Spectrum and UWB 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 36 Example Agile Use of Spectrum Channel 6 Channel 7 Channel 5  Legacy uses such as TV do not allow adjacent channels in the same geographical location due to primitive receivers  Agile Radio could use low power channel 5 inside of channel 6 and 7 coverage areas  Could use any channel at higher power outside of their and adjacent channels in areas where they are not allocated 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 37 SPTF Recommendations  Designate additional bands for unlicensed use  Pursue secondary markets for use of licensed spectrum  Government granted easements to licensed spectrum to enable a commons for low power non-interfering users  Promote spectrum flexibility in rural areas  Promote experimental spectrum allocations 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 38 Some Public Responses 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 39 Big Bang Auction of Spectrum  Championed by FCC Economists Evan Kwerel and John Williams  Existing spectrum licensees incentivized to put “their” spectrum up for auction  Not required, but if they don’t, their use of the spectrum continues to be bound by old rules  Get to keep all the proceeds from their sale  Government (including military) puts all its spectrum in the auction  Government can “buy back” spectrum for government (military, public safety) or public “Spectrum Parks”  Purchasers can aggregate spectrum  Corporations or Organizations can buy spectrum for “unlicensed” uses  WiFi Alliance could buy spectrum for 802.11 for instance 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 40 Faulhaber & Farber Proposal  Place all spectrum into the market, using Big Bang  Ownership model: Fee simple with non-interference easement  I own the spectrum and have absolute use priority; others can use it but only if they don’t interfere with this absolute use priority  UWB, agile radio, mesh networks OK; “virtual commons”  Monitoring and enforcement = transactions costs  Gov’t (at all levels) and private groups can own spectrum and make it available for commons use: “spectrum parks” 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 41 Concerns with Faulhaber & Farber Proposal  Better than a pure Private Property Model  Still assumes spectrum as private property should be the dominant model  Does not show that the benefits outweigh its costs  Treats the Commons as a hedge  Still constrains Commons opportunity  Promotes permanent grant of private property  Provides no revisability when technology or applications change  Not enough information to make such final and irrevocable decision 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 42 Technology based Commons  Advocated by many technologists and consumer advocates  David Reed (MIT), Yochai Benkler (NYU School of Law), Lawrence Lessig (Stanford) Dewayne Hendricks (Dandin Group) among others  New Tech utilizes spectrum more efficiently  Works best with large swaths of spectrum  Underlays & Agile radios can allow for transition from legacy  Industry Standards and technology manage most sharing issues 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 43 National Association of Broadcasters (NAB)  Mostly concerned with protecting their existing “rights”  Maintain that they are defenders of public interest with “free” Television  Against auctions because they undervalue the public interest benefits  Claim they are already “setting new standards in spectral efficiency”  Against commons for fear of interference with old radios and TVs  Consider the Cellular industry their biggest threat in terms of spectrum 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 44 Cellular Telecommunications & Internet Association (CTIA)  Represents the Cellular / Mobile Phone Industry  Aggressively supports spectrum policy reform  Cellular industry wants more spectrum for more capacity  Against “giving non-viable incumbents flexibility to provide any service”  I.E. allowing TV stations to compete with Mobile Phone Companies  Against underlay in already allocated bands  Ok for new licenses where it can be explicitly stated 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 45 Motorola: Go slow, mostly licensed, some unlicensed ok  Offered two technical papers  Good analysis of mesh networks  Raises questions on ability of cognitive radios to release spectrum fast enough  Still seems more FUD to slow down new tech  Supports R&D in mesh and cognitive radios  But not immediate roll out or licensing  Supports more unlicensed spectrum  Some in 5Ghz, but most in 10Ghz and above 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 46 Satellite Broadcasting & Communications Association  Protect DBS receivers from Terrestrial Interference  Concerned with sharing spectrum with terrestrial Multi-channel Video Distribution and Data Service (MVDDS)  Against underlays and unlicensed spectrum  Current DBS receivers are easily interfered with  GPS is very sensitive to interference  Fear of near channel overlap on satellite radio 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 47 National Association of Amateur Radio (ARRL)  Amateur Operators use to be radio innovators  Since the IC and Digital revolutions Amateurs have not been as involved  Now are mostly legacy users  FCC should use the SPTF for planning  No Big Bang / privatization of spectrum  Unless Amateur’s get their own “Public Park”  Concerns of unlicensed spectrum  Proper policing of power and other tech constraints  Supports the need for regulation of receivers 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 48 Consumer Federation of America  Pro-consumer advocacy organization  Considers spectrum to be a “First Amendment (Freedom of Speech) asset of citizens  Privatization of spectrum would limit free speech  Selling of spectrum would accelerate consolidation of media  Supports Spectrum Commons 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 49 Microsoft: Unlicensed Spectrum will Unleash Broadband  Unlicensed wireless can break the broadband bottleneck  Allows the Internet to “route around” incumbents Telcos who are slow to build broadband  Allows end users to finance broadband builds  Supports Spectrum Commons and additional unlicensed spectrum  Believe there should be “rules of the road” to facilitate sharing of spectrum 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 50 Cisco: Unlicensed Spectrum for the Network Revolution  Similar to Microsoft’s points  Some additional points of Cisco:  More commons, less private spectrum ownership  Don’t get stuck on international spectrum harmonization 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 51 XtremeSpectrum: UWB; new understanding of Interference  Manufacturer of UWB technology  Current vague definition of interference be replaced by explicit definition  The new “Interference Temperature is a good start.  Specify minimal capabilities of receivers to reject interference  Translate that into max permissible emissions levels for underlay technology 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 52 Potential Action Items 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 53 Policy Recommendations  Avoid irrevocable and difficult to change policies  Privatizing large amounts of spectrum would be difficult to reverse  Technology is just beginning  Enable some form of “easements” on existing and new licenses to allow for underlays and agile radios.  Be explicit with legacy incumbent licenses  Much of the issues of the report were concerning incumbents, but were not openly discussed as such  Treat incumbents as a transition issue, not assume that they should be incumbents forever 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 54 Reality Checks  Spectrum Utilization Audits  See how spectrum is really being used in various regions  Analyze capacity of spectrum  Consider several dense usage scenarios using data from the spectrum audits and demand growth profiles  Calculate various spectrum utilizations with different technology assumptions  See how often there is really a “tragedy of the commons” 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 55 Consortium to develop Open Spectrum Technologies  Open Spectrum will be a boon for hardware device manufacturers  Japan could lead in Open Spectrum device technology  A consortium to develop core technologies could be highly leveraged  Radio Haven in a secondary market would be an excellent test environment 

03/31/2003 Copyright 2003 Robert J. Berger & GLOCOM 56 Status: Major Fork in the Road  FCC & Industry split  Economists & Incumbents like Property Models  Technologists & Internet types like Commons  Surprising support for commons though  Boxer/Allen Senate Bill  Pronouncements from Chairman Powell considering making unused TV bands unlicensed