Before the
FEDERAL COMMUNICATIONS COMMISSION
Washington, DC 20554
In the Matter of
Inquiry Regarding Software Defined Radios ) ET Docket No. 00-47
COMMENTS OF THE NATIONAL TELECOMMUNICATIONS
AND INFORMATION ADMINISTRATION
Gregory L. Rohde Kathy Smith
Assistant Secretary for Chief Counsel
Communications and Information
William Hatch
Associate Administrator
Office of Spectrum Management
Edward Drocella
Telecommunications Specialist
Office of Spectrum Management
National Telecommunications and
Information Administration
U.S. Department of Commerce
Room 4713
1401 Constitution Avenue, N.W.
Washington, DC 20230
(202) 482-1816
June 16, 2000
Table of Contents
Section
I. INTRODUCTION
II. WHAT FEATURES IN A RADIO ARE APT TO BE CONTROLLED BY
SOFTWARE? FOR EXAMPLE, COULD THE OPERATING FREQUENCY,
OUTPUT POWER, AND MODULATION FORMAT BE SOFTWARE CONTROLLED?
III. WHAT ARE THE SPECIFIC LIMITATIONS OF CURRENT SOFTWARE DEFINED RADIO TECHNOLOGY? WHAT ARE THE COST IMPLICATIONS?
IV. WHAT CAPABILITIES COULD SOFTWARE DEFINED RADIOS HAVE THAT
ARE NOT FOUND IN CURRENT RADIO TECHNOLOGY? .
V. WHAT WORK IS BEING DONE ON SOFTWARE DEFINED RADIOS INTERNATIONALLY, AND ARE THERE ANY STEPS THE COMMISSION SHOULD TAKE TO ENCOURAGE THIS WORK?
VI. TO WHAT EXTENT CAN SOFTWARE DEFINED RADIOS IMPROVE INTEROPERABILITY BETWEEN DIFFERENT PUBLIC SAFETY AGENCIES?
VII. TO WHAT EXTENT CAN SOFTWARE DEFINED RADIOS IMPROVE THE EFFICIENCY OF SPECTRUM USAGE?
VIII. WHAT PARTICULAR FUNCTIONS RELATED TO SPECTRUM USAGE
COULD SOFTWARE DEFINED RADIO PERFORM? COULD IT LOCATE
FREE SPECTRUM, DYNAMICALLY ALLOCATE BANDWIDTH, AND
ENABLE BETTER SHARING OF THE SPECTRUM?
IX. SHOULD WE APPROVE THE RADIO HARDWARE, THE SOFTWARE OR THE COMBINATION OF THEM?
X. ARE THE CURRENTLY REQUIRED MEASUREMENTS IN PART 2 OF THE
RULES APPROPRIATE FOR SOFTWARE DEFINED RADIOS?
XI. SHOULD WE REGULATE WHO CHANGES THE SOFTWARE AND THE
MANNER IN WHICH IT IS DONE? IF SO SHOULD THE COMMISSION
MAINTAIN RECORDS OF EACH MODIFICATION?
XII. WHAT ARE THE VARIOUS MEANS THAT MAY BE USED TO
DOWNLOAD NEW SOFTWARE?
XIII. SHOULD WE REQUIRE ANTI-TAMPERING OR OTHER SECURITY
FEATURES? HOW WOULD SUCH SECURITY FEATURES WORK?
XIV. COULD EQUIPMENT BE DESIGNED TO PREVENT IT FROM
TRANSMITTING IN CERTAIN DESIGNATED FREQUENCY BANDS,
SUCH AS THOSE ALLOCATED EXCLUSIVELY FOR GOVERNMENT
USE, AS A SAFEGUARD AGAINST INTERFERENCE?
XV. IS THERE A NEED FOR SUCH AN APPROVAL SYSTEM, AND IS IT
FEASIBLE AND PRACTICAL?
XVI. THERE ARE INTERFERENCE ISSUES RELATED TO SOFTWARE
DEFINED RADIO RECEIVERS THAT SHOULD BE ADDRESSED
BY THE COMMISSION
XVII. THE COMMISSION SHOULD ADOPT A DEFINITION FOR SOFTWARE
DEFINED RADIO
XVIII. CONCLUSION
COMMENTS OF THE NATIONAL TELECOMMUNICATIONS
AND INFORMATION ADMINISTRATION
The National Telecommunications and Information Administration (NTIA), an Executive Branch agency within the Department of Commerce, is the President's principal adviser on domestic and international telecommunications policy, including policies relating to the Nation's economic and technological advancement in telecommunications. Accordingly, NTIA makes recommendations regarding telecommunications policies and presents Executive Branch views on telecommunications matters to the Congress, the Federal Communications Commission, and the public. NTIA, through the Office of Spectrum Management, is also responsible for managing the Federal Government's use of the radio frequency spectrum. NTIA respectfully submits the following Comments in response to the Commission's Notice of Inquiry in the above-captioned proceeding.(1)
Wireless technology advances promise to make radios based on Software Defined Radio (SDR) technologies more prevalent within the Federal and non-Federal communication infrastructure. In the commercial sector, there is great interest in new business opportunities afforded by SDR. For example, SDR reconfigurability could provide access to a multitude of communications services worldwide and allow service customization according to user preferences. From a regulatory perspective, this flexibility challenges the current telecommunications rules, which were designed to manage radio devices having a static set of regulatory attributes. The Federal Communications Commission (Commission) has initiated the SDR Notice of Inquiry (SDR NOI) to obtain views regarding the potential impact of SDR technology on the use and regulation of the non-Federal portion of the radio spectrum.
The SDR NOI contains a series of questions to help elicit comments. The questions are grouped into the following categories: state of software defined radio technology, interoperability, improving spectrum efficiency and spectrum sharing, and equipment approval process.
In parallel with the tremendous growth and demand for wireless communications, the underlying radio techniques have benefitted from major leaps in enabling technology. Two significant trends are the advent of microprocessor-controlled radios and the related rapid advances in integrated circuit technologies. These developments have led directly to the evolution of software programmable capability, digital signal processing, and multiband, multimode radios. The resultant enabling feature of these radio technologies is the ability to rapidly perform many control functions within the software domain, without altering the hardware.
The Federal Government agencies are placing an increasing emphasis on SDR technology because of its potential for lower cost and reconfigurability for multimode/interoperable communications. The SDR is viewed as a critical technology to satisfy requirements for interoperability among the various services and with foreign allies, as well as a means to achieve information superiority, operational flexibility, and cost benefits. An example of this developing radio concept is the Joint Tactical Radio System (JTRS).(2) JTRS represents the new generation of SDR that is reprogrammable to address multiband, multimode, and network capabilities to provide voice, data, and video communications. JTRS architecture development and validation is scheduled for completion in late 2000. This procurement of JTRS is intended to meet communications requirements by capitalizing on commercial technologies and processes to the maximum extent possible. These radios will be based on a common open software architecture developed by industry. The Navy's Digital Modular Radio (DMR) is another example of the next generation of SDR technology that will provide interoperability with new tactical systems (e.g., Have Quick and the Single Channel Ground Air Radio System (SINCGARS)).(3) The DMR is fully programmable and includes embedded software programmable cryptography. DMR development contracts were let in 1998 and a low rate production contract was let in January 2000. The Federal Aviation Administration's NEXCOM program is examining the utilization of SDR technology to replace analog radios. Federal law enforcement agencies have also considered SDR as a possible solution to satisfy interoperability communication and multiband radio requirements.
Although the SDR NOI is intended to elicit comments from the "public", the Federal Government and the commercial sector have common interests in SDR technology and the resolution of the questions posed by the Commission. Specifically, common interests include positive spectrum control of SDR emissions to ensure electromagnetically-compatible operations in authorized frequency bands, interoperability between various radio waveforms/standards, and the potential use of adaptive SDR capabilities to improve spectrum access for Government and non-Government users.
NTIA applauds the Commission for its efforts in this inquiry to obtain information on the current state of SDR technology. NTIA believes that this inquiry is necessary to determine whether changes to the current rules are required to facilitate the deployment of SDR technology. NTIA offers the following comments to specific questions raised in the SDR NOI that NTIA believes could have a direct impact upon the operations of these types of radios.
II. WHAT FEATURES IN A RADIO ARE APT TO BE CONTROLLED BY SOFTWARE? FOR EXAMPLE, COULD THE OPERATING FREQUENCY, OUTPUT POWER, AND MODULATION FORMAT BE SOFTWARE CONTROLLED?(4)
Nearly every radio frequency (RF) attribute of an SDR subject to regulation is potentially controllable via software. JTRS is expected to be equally controllable by software. A particularly important aspect of SDRs is the possibility of re-programmability or the ability to control certain radio parameters using software. Re-programmability can be limited to a fixed set of choices, thus providing a limited number of configurations, or the re-programmability can allow variations over a range of values for each radio function providing an unlimited number of possible configurations. NTIA has performed an initial review of the equipment characteristics required for the system review process to determine if changes would be required to accommodate SDRs.(5)
NTIA's initial review of transmitter information identified that the following transmitter parameters could be reprogrammed in SDRs:(6)
Tuning Range
RF Channeling Capability
Emission Bandwidth
Maximum Bit Rate
Maximum Modulation Frequency
Use of Pre Emphasis
Deviation Ratio
Power
Pulse Characteristics
Update Rate of Digital-to-Analog Converter
NTIA's initial review of receiver information identified that the following receiver parameters could be reprogrammed in SDRs:(7)
Tuning Range
RF Channeling Capability
RF Selectivity
Intermediate Frequency (IF) Selectivity
Maximum Bit Rate
Maximum Post Detection Frequency
Minimum Post Detection Frequency
IF Frequency
Analog-to-Digital Sample Rate
Anti-Aliasing Filter Response
Part of the answer to the Commission's question is whether the re-programmability of the SDR is limited to a fixed set of choices for the reprogrammable parameters (resulting in a finite set of standard configurations) or allows selection over a wide range of values for the reprogrammable parameters. Based on an initial review, NTIA offers the above listed transmitter and receiver parameters as a set of the possible parameters of an SDR that can be controlled by software. NTIA also recommends that the method of re-programmability for each reprogrammable parameter be described.
III. WHAT ARE THE SPECIFIC LIMITATIONS OF CURRENT SOFTWARE DEFINED RADIO TECHNOLOGY? WHAT ARE THE COST IMPLICATIONS?(8)
The use of SDR technology within a radio may be limited by size, weight, power, performance, and the current state of the SDR technology. Early SDR implementations have shown that the multifunction flexibility afforded by a software implementation usually comes at the price of added size, weight, and power, relative to a single function radio implemented in hardware. When an SDR can replace several hardware-based single-function radios, these disadvantages may be acceptable.
The computationally most demanding waveform that an SDR must handle will influence its complexity and cost. This is a limitation to the extent that it is currently possible to develop lower-cost hardware-based radios for specific waveform processing functions. Also, it is possible to define complex waveforms that cannot be processed in real time using commercial-off-the-shelf digital signaling processors.
Digital signal processing (DSP) techniques are key to the implementation of SDR. However, all RF communications eventually requires an analog interface with the antenna for transmission and reception. This requires all SDRs to contain some analog components that clearly are not software defined but may be software controlled. Since analog components cannot be reconfigured to the same extent as software, SDRs will be limited in their ability to reconfigure their RF front ends and transmitter output stages.
IV. WHAT CAPABILITIES COULD SOFTWARE DEFINED RADIOS HAVE THAT ARE NOT FOUND IN CURRENT RADIO TECHNOLOGY?(9)
SDR technology can allow one radio to interface with multiple telecommunications services. The interfaces could be radio initiated (i.e., adaptive) or user initiated. The availability of multiple services would expand competition for services. SDRs would likely be integrated into other products such as laptop computers, personnel digital assistants, and automobiles to provide reconfigurable information links. SDR flexibility could also provide services better tailored for each user's needs or interests. Additionally, capabilities could be software configurable to adapt to internationally different assignment of frequencies or waveform standards such as cellular telephone standards (e.g., Time Division Multiple Access and Code Division Multiple Access in the United States or Global Systems for Mobile in Europe).
For example, Motorola has used their WITS-6004 to demonstrate the feasibility of interoperability between dissimilar communications services. JTRS will interoperate with each other and legacy radios in the same manner to enhance each U.S. military component and allied forces components to communicate with each other on the tactical battlefield.
V. WHAT WORK IS BEING DONE ON SOFTWARE DEFINED RADIOS INTERNATIONALLY, AND ARE THERE ANY STEPS THE COMMISSION SHOULD TAKE TO ENCOURAGE THIS WORK?(10)
SDR has tremendous potential in the near-term to accommodate multiple bands/standards in the United States and internationally to integrate third generation (3G) wireless applications. SDR technology has the potential to unite a world of diverse standards and technologies, and frequency bands. SDR could provide a near-term solution that builds on, rather than replaces, successful Second Generation (2G) systems and later generation radios and services. SDR is particularly well suited to provide a long-term solution for 3G wireless, and is included as an element of the International Telecommunication Union (ITU)/International Mobile Telecommunications-2000 (IMT-2000) work that is currently underway.(11)
SPEAKeasy an early SDR project, under U.S. Government leadership, was the catalyst for the U.S.-led Modular Multifunction Information Transfer System (MMITS) forum. MMITS' global participation included: Alcatel (France), Ericsson (Sweden), Keio University (Japan), Motorola (United Kingdom), Nokia (Finland), Orange Personal Communications (United Kingdom), Rhode and Schwarz (Germany), Samsung Electronics (South Korea), and Siemens (Germany) among others. The recent renaming of the MMITS to the SDR Forum (SDRF) signals a shift to a greater commercial emphasis in open architecture standards for SDR. The SDRF is based in the United States and seeks the global harmonization of SDR concepts on the basis of requirements, rather than a technology-driven philosophy. A major SDRF activity is the development of the SDR system architecture which has the following desired characteristics: flexibility, upgradeability, scalability, and extensibility.(12)
In addition to the efforts ongoing in the SDRF, the European Commission has sponsored precompetitive software radio programs in its R&D in Advanced Communications in Europe (RACE) and Advanced Communications Technology and Services (ACTS) programs.(13) The Flexible Integrated Radio System and Technology (FIRST) and Future Radio Wideband Multiple Access System (FRAMES) projects of the ACTS program have investigated next generation air interfaces using prototype software radios and have examined many aspects of software-reconfigurable air interface implementation.(14)
The ITU provides an international treaty-based regulatory framework consisting of allocations to radio services, technical conditions to avoid interference to services of other countries, technical/operating characteristics in very limited cases to ensure global interoperability (e.g., maritime mobile), and voluntary standards known as Recommendations. The ITU may be the most suitable forum to address the wide range of global regulatory issues that surround SDR technology.(15)
A Draft New Question has been introduced in ITU-Radiocommunication Sector (ITU-R) Study Group 8 that addresses SDRs.(16) This Draft New Question will allow administrations to study:
- the key technical characteristics associated with the design and application of SDR;
- the frequency band considerations that are important to the application of SDR;
- the special interference considerations that may be required in SDR applications;
- the operational implications of SDR to mobile radio systems;
- the appropriate definition for SDR; and
- the technical considerations that are necessary to insure conformance with ITU Recommendations and Radio Regulations.
The results of the studies called for in this Draft New Question are to be completed by the 2003 World Radiocommunication Conference.
NTIA recognizes that there is a great deal of interest in SDR technology globally and global interest in addressing the associated regulatory issues. The international forum for SDR (The SDR Forum) is based in the United States, signifying that the United States is a world leader in SDR technology development. Therefore, NTIA recommends that the U.S. Government actively participate with industry representatives within the SDRF to complete the studies called for in the ITU-R Draft New Question. Working within the ITU forum with other global regulators can set the stage for the development of a favorable global regulatory environment for SDR. This will help promote the growth of U.S.-based SDR technology development and will create an environment for those firms that supply and use SDR to have a global marketplace.
VI. TO WHAT EXTENT CAN SOFTWARE DEFINED RADIOS IMPROVE
INTEROPERABILITY BETWEEN DIFFERENT PUBLIC SAFETY AGENCIES?(17)
The Public Safety Wireless Advisory Committee (PSWAC) Final Report defines interoperability as the ability of two or more public safety communications systems to interact with one another and exchange information according to a prescribed manner in order to achieve predictable results.(18) Public safety includes individuals in Federal and non-Federal (state and local) public safety agencies, generally made up of law enforcement/police services, fire and rescue services, emergency medical services, and emergency management services. Three types of public safety interoperability missions each with different requirements have been identified: day-to-day interoperability, mutual aid interoperability, and task force interoperability.(19) At this time, police and Federal law enforcement vehicles often have multiple mobile radios, while multiple portable radios are somewhat common in the fire services. This allows interoperability across bands or system protocols, but causes vehicle space problems or burdens public safety officials with additional weight and cost penalties.
The current frequency allocations for public safety land mobile channels are scattered over five disparate segments of the frequency spectrum between 25 MHz and 1 GHz. There are public safety frequency allocations in 30-50 MHz (VHF Low Band), 162-174 MHz (VHF High Band), 406-420 & 450-512 MHz (UHF), 764-776 & 794-806 MHz (700 MHz), and (806-940 MHz (800 MHz).(20) As a result, radios in one band cannot currently interoperate with radios in another band. Multiple agencies converging on a single incident with communications systems that do not share the same frequency, face a much greater communications challenge than those who share common frequencies or even a common frequency band. SDRs capable of operating in multiple frequency bands are a possible solution to one aspect of the public safety interoperability problem.
In addition to fragmented frequency allocations for public safety land mobile communications, different air-interface standards(21) or protocols are typically employed in each band, making interoperable communications impossible. Functional integration is used in the SDRs to reduce the number of radio types into a single general-purpose programmable waveform processor. It is therefore possible for multiple law enforcement air-interface standards to be implemented in a single radio, despite different physical layers (modulation, forward error correction), link layers (link acquisition protocols, link maintenance, frame/slot processing), network layers (network protocols, media access protocols, network time maintenance), upper layers (source coding), time bases, and bandwidths.(22) SDRs capable of supporting different air- interface standards are a possible solution to the public safety interoperability problem.
The PSWAC Final Report considered software programmable radios as a possible solution for the interoperabiltiy problems encountered with multiple frequency bands and air- interfaces. At the time the PSWAC Final Report was written, there were many technical challenges identified to producing a practical software programmable radio. Antenna systems must operate across a wide frequency range; a single multiband antenna system is preferable to many antennas for different frequency bands. These antenna systems could be augmented with "smart" antenna technology. Other enabling technologies included multiband power amplifiers, tunable preselectors, interference cancellers, low noise synthesizers, wideband low noise amplifiers, wideband linear mixers, high throughput digital signal processors, and smaller chip packaging.(23)
In addition to the technical challenges identified by the PSWAC, software radios were found to be much more expensive than hardware based radios, with the market being confined to the military, big business, and government applications. However, it was anticipated that over time, the cost of the software radio enabling technologies would decrease thereby reducing the cost of the SDRs.(24)
NTIA participated in the PSWAC, and based on the available information supported the conclusions of the PSWAC Final Report regarding software radios. As stated in the PSWAC Final Report there are technical and economical factors that must be taken into consideration. NTIA recognizes that since the PSWAC Final Report was written, there have been advances in the SDR enabling technologies such as analog-to-digital converters, field programmable gate arrays, digital signal processors, and linear power amplifiers. These advances in the enabling technologies will make it possible for SDRs to play an increasing role in improving interoperability between public safety agencies. However, the cost difference between SDRs and radios employing hardware solutions is still a major concern. Federal, state, and local public safety agencies have limited monetary resources, which at this point in time is still a major obstacle in using SDRs to resolve interoperability between public safety agencies.
VII. TO WHAT EXTENT COULD SOFTWARE DEFINED RADIOS IMPROVE THE
EFFICIENCY OF SPECTRUM USAGE?(25)
In general spectrum efficiency includes elements and parameters within transmitters and receivers. The processors built into SDRs can be used to implement more efficient channel access schemes. Adaptive access schemes that sense RF activity should be able to provide more efficient spectrum utilization than the present fixed assignment process. Rather than being assigned a single frequency, adaptive systems could be given numerous frequencies from which to select from on a real-time basis. Conceivably, this technique could permit demand access to a common pool of frequencies and allow conventional netted systems to approach the efficiency of trunked systems. However, the extent of the actual improvement in spectrum efficiency has not been quantified.
Standardizing the algorithms permitted in certain frequency bands can maximize the effectiveness of adaptive spectrum access. Any required standards need to be reflected in the Commission's rules. Organizations such as the SDRF may be helpful in coordinating industry efforts to develop standards for adaptive access. The rules would need to be changed to allow assignments to frequency bands rather than specific channels and the equipment approval process would need to verify that the radio implements the access protocol properly.
Standardized algorithms for adaptive spectrum access may also allow more intensive sharing of the spectrum. For frequency bands where such sharing could be implemented, exclusive Government or non-Government frequency allocations would not be required.
From the receiver spectrum efficiency perspective, NTIA has been a long time advocate of employing receiver standards to improve overall spectrum efficiency. NTIA has also recommended that the Commission promote the development of adequate industry receiver standards to minimize the potential for interference from adjacent band transmitters.(26) We are aware that high order, steep roll-off filters constructed of electronic components (analog filters) are expensive and difficult to realize in practice. Also, as the steepness of the roll-off is increased, the phase response tends to become more nonlinear. This can create distortion of the desired receive signal since different frequencies within a signal can be delayed in time by different amounts.(27) However, in SDR receivers, many filtering functions that cannot be implemented in analog hardware can be implemented in software. An example is the design of finite impulse response (FIR) filters that simultaneously can achieve sharp roll-off and linear phase response.(28)
NTIA has adopted technical standards applicable to receivers operating in certain frequency bands, such as fixed and land mobile services in the 406.1- 420 MHz band(29) and the fixed service in the 1.71-15.35 GHz band.(30) If SDRs were permitted to operate in these bands, they would be required to comply with these standards which specify spurious response attenuation levels, adjacent channel selectivity, intermodulation rejection requirements, and conducted spurious emission levels. The NTIA will request that Federal agencies provide technical characteristics data and the description of the SDR receivers to determine whether SDR receivers are capable of meeting applicable standards.
NTIA continues to believe that receiver standards are an extremely important element in improving spectrum efficiency. The ability of SDR receivers to implement filtering with greater adjacent band selectivity using software that cannot be implemented using analog hardware has the potential to increase spectrum efficiency and reduce the potential for interference from adjacent band transmitters. NTIA will continue to work with the Federal agencies to address the technical issues related to SDR receivers. NTIA recommends that the Commission as part of this rulemaking process examine the feasibility of promoting the development of industry standards for SDR receivers to improve the overall efficient use of the spectrum.
VIII. WHAT PARTICULAR FUNCTIONS RELATED TO SPECTRUM USAGE COULD A SOFTWARE DEFINED RADIO PERFORM? COULD IT LOCATE FREE SPECTRUM, DYNAMICALLY ALLOCATE BANDWIDTH, AND ENABLE BETTER SHARING OF THE SPECTRUM?(31)
NTIA recommends that the Commission permit the first generation SDRs to seek dynamic assignments in the bands allocated for their service. NTIA is also exploring the use of dynamic channel assignment by Federal Government SDRs. In general we believe that dynamic bandwidth allocation could allow better sharing in time, space, and frequency domains. We believe that the Commission should entertain policy and rulemaking approaches to demonstrate and validate these concepts. The partnership between the NTIA, the Commission, and the private sector should be formed for this purpose.
IX. SHOULD WE APPROVE THE RADIO HARDWARE, THE SOFTWARE OR THE COMBINATION OF THEM?(32)
The answer to this question might vary according to the radio application. NTIA believes that it will be necessary to require software and hardware tests. SDR technology has not matured to the point where it is possible to predict software radio RF parameters from examining only the software or only the hardware.
The basic building blocks of the SDR transmitter are shown in Figure 1. The digital software functions are grouped as a single entity, block A in Figure 1.
The digital processing functions are referred to as software objects. These software objects can be viewed as a computer program operating on some DSP platform. The output of the software object is a digital signal. Inherent to every transmitter, however, is the requirement that the digital signal be ultimately converted to a continuous-time signal with a continuous amplitude, hereafter referred to as an analog signal. The location for this conversion, along the signal processing chain, can occur at the baseband, IF, or RF level. If the conversion occurs at RF the mixer shown in Figure 1 is not required. There are hardware components required following the digital-to-analog converter (DAC); one is a smoothing filter (located at B in Figure 1), and the other is a power amplifier and optional filter following the power amplifier (located at C in Figure 1). The grouping of these hardware components are referred to as a hardware object.(33) In a SDR transmitter, the modulated signal to be transmitted is generated as a digital signal using DSP techniques and is referred to as a software object. The digital signal is then converted to an analog signal for transmission. A software radio will also consist of hardware objects (i.e., analog amplifiers and filters) which will have attributes set by the manufacturer that may be discerned by the software processor. This can be thought of as a software-readable object descriptor for the hardware object. The capability of a SDR can be defined in terms of configurations combining selected hardware and software objects.
Transmitter output characteristics of the amplifiers, filters, and oscillators, DAC, and harmonic emission levels need to be considered in the approval process as well as any effects of the DSP platform (i.e., finite word length effects). The concept of SDRs consists of configurations combining selected hardware and software objects operating on a DSP platform. If it is possible to show that particular hardware components when used with specific software objects operating on a given DSP platform, can be approved as meeting the applicable emission standards, then it may be possible to approve the software separately.
NTIA believes that separate hardware and software approval will only be possible if a consistent predictable connection between the software object (computer program) operating on a given DSP platform and the hardware objects can be established. If it is not possible to establish a consistent, predictable connection between the software objects operating on a given DSP platform, hardware objects, and the emission characteristics at the antenna input, then another method of approving the SDRs needs to be developed. In order to begin the process of determining if a consistent, predictable relationship between the software (computer code) operating on a given DSP platform and hardware objects can be adequately specified, technical characteristics for each hardware object (i.e., oscillators, amplifiers, filters) and a technical description of the software objects operating on a given DSP platform (i.e., computer code, digital-to-analog converter) for each configuration of software and hardware may be necessary.
X. ARE THE CURRENTLY REQUIRED MEASUREMENTS IN PART 2 OF THE RULES APPROPRIATE FOR SOFTWARE DEFINED RADIOS?(34)
There may be a need to test additional aspects of SDRs. SDRs may include adaptive access algorithms and control mechanisms and the test process needs to verify these features. Adaptive algorithms might be tested by connecting the SDR radio RF port to a simulator that would test a series of RF conditions to which the SDR would be required respond to in accordance with specifications in the Commission's rules.
Because SDRs may in principle be easily reconfigured, control measures must be implemented to insure that only authorized reconfiguration can be implemented. Tests should be included to verify the existence and effectiveness of the controls.
In Part 2 of the Commission's Rules, it is required that the following measurements be made on transmitters used on the licensed services: RF power, modulation characteristics, occupied bandwidth, spurious emissions at the antenna terminals, field strength of spurious emissions and frequency stability.(35) In addition to these technical characteristics, NTIA is also concerned that the digitization process in an SDR transmitter may contribute to signal distortions that are unique to the digitization process, and not present in traditional transmitters. For example, the digital-to-analog conversion can cause extraneous spectral emissions. Harmonics are accentuated by DAC update rates that are an integer multiple of the signal frequency. Some of these are due to nonlinearities in the DAC and some are due to the nature of digital signals converted to the analog domain resulting in spectral copies located around integer multiples of the DAC update rate.(36) While a concerted effort is usually made to remove these spectral components, there is the possibility of their existence. It is not known at this time what impacts these digitization effects will have on electromagnetic compatibility (EMC) with other systems in the environment. NTIA and the Commission should work with the manufacturers in the development of SDRs to ensure that these digitization effects will not impact the compatibility of SDRs. Similarly, NTIA through its spectrum certification process will work with the involved Federal agencies to address concerns about unintended effects of the digitization process because they can affect overall SDR performance including EMC. For example, during the procurement of fieldable JTRS radios, digitization effects will be evaluated by contractor and government tests. Spectrum certification data submissions for JTRS will reflect the effects of digitization and allow for NTIA review and comment.
XI. SHOULD WE REGULATE WHO CHANGES THE SOFTWARE AND THE
MANNER IN WHICH IT IS DONE? IF SO SHOULD THE COMMISSION
MAINTAIN RECORDS OF EACH MODIFICATION?(37)
As SDR technology migrates to a completely software radio (SR), there will be dynamic software-defined capabilities throughout the device including RF functions such as the frequency of operation and the transmit power. The SR will have the capability of being reconfigured by a software download possibly using an over-the-air transmission. It is the ability of SRs to be reconfigured that has control and regulatory implications. NTIA believes that it will be necessary to investigate what regulatory procedures are necessary for the entities that are permitted to issue or reissue radio software to ensure appropriate control of the software and to provide software traceability.
For SRs the avoidance of interference to other systems can be controlled to some extent by the Commission's type approval process for the hardware platform and the initial software package. However, subsequent software downloads and changes to the software could lead to incompatibilities causing unforseen problems. NTIA believes that a software configuration management process is one possible solution to regulate software changes for SRs. As part of the software configuration management process for an SR, a record of its current configuration and capabilities will be recorded. This information could then be made available on a timely fashion to any party with a need or a right to know. The software configuration management process would also record all subsequent modifications (versions) of the software. A configuration management process is under consideration to track approved hardware and software changes for the JTRS. NTIA recommends that the Commission in conjunction with industry representatives within the SDRF investigate the possibility of implementing a configuration management process to track and record changes to the software as part of the regulation of SRs.
XII. WHAT ARE THE VARIOUS MEANS THAT MAY BE USED TO DOWNLOAD NEW SOFTWARE?(38)
The implementation of a fully programmable radio system implies the definition of methods for software download. For base stations, the software download will normally be executed when a new software version is released, and therefore, will not be performed frequently. On the other hand, the software download for the mobile terminal is more critical because the task could be performed more frequently due to user mobility or to user needs. For example, the user needs to change the air interface for roaming access. Moreover, the software download in the mobile terminal must be as fast as possible, easy to perform, user transparent, secure, and error-free. In addition to the download methods proposed by the Commission, an alternate method of using a Smart Card for downloading software for SDR/SR should also be considered.
For Smart Card software loading, the customer could purchase the Smart Card containing the software from the operator. The software download is then performed when the Smart Card is inserted in the user terminal. Some problems could occur when the user needs to roam to other operations in the same or different country, using different standards. In this case, the user has to purchase several different Smart Cards and must be able to use the appropriate one. A possible solution to this problem could be the installation of equipment in airports, railway stations, and hotels, which would permit the Smart Card to be reprogrammed, upon arrival at the hosting country/region. The main disadvantage of the Smart Card is that the software download procedure is not effortless for the user.
NTIA believes that the possible advantages of the Smart Card include: 1) error-free software download ensuring greater software integrity; 2) fast download, comparable to that achievable with a computer; 3) no impact on the network (e.g., over-the-air software downloads require that an internationally standardized dedicated channel be established); and 4) provides a possible solution for the security and authentication issues that are associated with over-the-air software downloading. NTIA recommends that the Commission in coordination with the industry members of the SDRF consider using the Smart Card as a possible method of downloading software for SDR/SR.
XIII. SHOULD WE REQUIRE ANTI-TAMPERING OR OTHER SECURITY
FEATURES? HOW WOULD SUCH SECURITY FEATURES WORK?(39)
The ability of the SDR/SR to be upgraded, reprogrammed, or reconfigured over-the-air or over a network connection will permit improved capabilities to be incorporated easily. However, it is this ability to reconfigure the SDR/SR by using a software download that could make them extremely vulnerable to malicious attacks. Therefore, NTIA is exploring the feasibility of requiring SDR/SR radios to implement security features to ensure software integrity.
The process of reconfiguration of a SDR/SR either over-the-air or a network connection could include three actions for successful installation: 1) authentication; 2) acceptance; and 3) activation. To ensure the integrity of the SDR/SR each type approved component should require an attached or embedded certificate (code) that cannot be forged and is traceable back to the issuing authority. SDRs should be built to a common open architecture standard that uses authentication to ensure that radios could only run authorized waveform software, as well as other common software such as operating systems. These features should ensure positive spectrum control and prevent the use of unauthorized waveform software that could allow transmissions in authorized bands. Digital serial numbers encoded in the radios could identify groups of users such as commercial, public safety, government non-military, and military. The authentication process could use these serial numbers to determine if a software application is appropriate for a specific SDR device. Because many types of users may use SDRs, the design of the radio should be flexible enough to respond to multiple users with varying requirements. Similarly, bands not authorized for a particular purpose in the United States may be appropriate elsewhere in the world. Manufacturing costs can be minimized if the same device can be marketed to the largest possible consumer base.
NTIA and the Commission, in coordination with industry representatives involved in SDR development need to examine security features such as authentication protocols that could be used to prevent unauthorized user access or prevent unauthorized functions to be implemented in the SDR/SR. At the present time it appears that the verification of the authentication protocols should be part the SDR/SR type approval process.
XIV. COULD EQUIPMENT BE DESIGNED TO PREVENT IT FROM TRANSMITTING IN CERTAIN DESIGNATED FREQUENCY BANDS, SUCH AS THOSE ALLOCATED EXCLUSIVELY FOR GOVERNMENT USE, AS A SAFEGUARD AGAINST INTERFERENCE?(40)
NTIA's primary concern regarding the concept of dynamic bandwidth management is the potential for interference to existing (legacy) systems. There are many Government and non-Government frequency bands that are currently allocated to radio services that support safety-of-life and other sensitive operations. Examples of these services include: aeronautical radionavigation, radionavigation-satellite, and radio astronomy.
NTIA believes that the first generation of SDR should be expected to comply with the individual operating and technical requirements of each service/band combination used, whether hardware, firmware, or software handles the function. At the present time NTIA is requiring SDRs used by the Federal agencies include a provision for restricting operation to certain bands allocated on an exclusive basis to the Federal Government. The requirement for SDRs to implement a capability to inhibit access to certain frequency bands is consistent with NTIA's present policy for the Federal use of SDR. Moreover, while SDRs can be programmable to adapt to the surrounding legacy environment systems, the legacy systems cannot. Therefore, in bands shared between Federal and non-Federal services, the legacy systems must be protected from potential interference from SDRs. NTIA anticipates that the experience gained with the first generation SDRs for both Federal and non-Federal applications will provide valuable insight into how additional polices and rules should evolve to obtain the maximum benefit from this technology. NTIA recommends that the Commission work with industry representatives to ensure that SDRs comply with the table of frequency allocations, including having the ability to lockout or prevent access to certain frequency ranges, waveforms, and combinations thereof.
XV. IS THERE A NEED FOR SUCH AN APPROVAL SYSTEM, AND IS IT
FEASIBLE AND PRACTICAL?(41)
Software radio control requires an approval process in which only authenticated software can be implemented in an SDR. The authentication process needs to be standardized so that authentication can be built into the radio. An exception to this might be SDRs that can operate in very limited bands or with low RF power output (i.e., Part 15 type devices). Another exception might be radios under the control of operators that must pass license exams such as amateur radio operators. However, SDRs available for general use must be subject to an approval process, to eliminate the potential to cause serious interference.
For SDRs that can span multiple bands and functions, the approval process needs to consider methods for limiting software operation to certain categories of users. Not every owner of a particular SDR would necessarily be authorized to exercise every waveform designed for that radio. The service provider may not want to allow the waveform to operate unless the appropriate access fee has been paid. Also, the Commission may choose to limit waveforms to those for which a particular user has been granted a license. This type of restriction might be implemented by encrypting the software by using a public key for the radio. Then only the target radio could decrypt (using its private key) and run the software.
If Federal Government SDRs such as JTRS have the capability to operate in non-Government bands, then they must be compatible with whatever authentication process the Commission adopts.
Another area that should be examined is the ability to copy a low-level signal and retransmit it. This would likely be a Part 15 type device used for opening a door, or a code used by cellular telephones. If SDR used its ability to copy the codes and then retransmitted them, it could give an operator the ability to gain access to an area currently controlled by electronic keys.
XVI. THERE ARE INTERFERENCE ISSUES RELATED TO SOFTWARE DEFINED RADIO RECEIVERS THAT SHOULD BE ADDRESSED BY THE COMMISSION
The primary focus of the SDR NOI is on SDR transmitters, however, NTIA believes that the Commission should also address the EMC of SDR receivers and adjacent band interference to SDR receivers.
Methods for analyzing EMC in traditional receivers (such as superheterodyne) are well established. However, the EMC analysis of SDR receivers that utilize digitization of the RF signal at the front-end may be quite different. The digitization process in the SDR receivers can contribute to distortions to the signal that are unique to the digitization process and are not present in traditional receivers. Examples for SDR receivers are that nonlinearities in the analog-to-digital (ADC) converter can cause distortions in the received signal and the process of digitization can introduce the possibility of spectrum overlap, sometimes referred to as aliasing.(42) It is not known at this time what impact these kinds of effects will have on the EMC with existing (legacy) systems. Detailed knowledge of how the SDR receivers operate is necessary in order to help develop appropriate methods for EMC analysis. NTIA and the Commission need to work with the manufacturers in the development of SDRs to ensure these digitization effects will not have an impact on the compatibility of SDRs.
One of the primary features of SDR is to have the ability to transmit and receive in multiple frequency bands. In order to achieve this goal, a typical SDR receiver architecture could include a broadband RF stage, a low noise amplifier (LNA) capable of amplifying signals over a wide frequency range, and a wideband high-speed ADC. It is this broadband SDR receiver architecture that is required for multiple frequency band operation that could make it vulnerable to interference from high-powered adjacent band transmitters such as radars used for national defense and air traffic control. For example, front-end overload is a phenomenon that occurs when an LNA is subjected to a strong signal from an external source. In this condition, the LNA will not only gain compress(43) at the frequency of the overloading signal, but also at all other frequencies in the amplifiers gain response band. Therefore, the desired signal will be lost at frequencies that may be hundreds or even thousands of megahertz away from the frequency of the signal that is causing overload to occur.(44) Thus the technical issues surrounding adjacent band interference to SDR receivers should be addressed by the Commission as part of the rulemaking process.
XVII. THE COMMISSION SHOULD ADOPT A DEFINITION FOR SOFTWARE DEFINED RADIO
There are several definitions that are available for SDR. One definition states that an SDR performs the majority of signal processing in the digital domain using programmable DSP and hardware support, but some signal processing is still done in the analog domain, such as in the RF and IF circuits.(45) Another definition states that SDR is a radio in which the receiver digitization is performed at some stage down stream from the antenna, typically after wideband filtering, low noise amplification, and down conversion to a lower frequency in subsequent stages, with a reverse process occurring for the transmit digitization.(46) There are also several definitions for SR. One definition states that an SR is a device where the antenna is connected directly to an ADC/DAC and all signal processing is done digitally using fully programmable high speed DSP. All functions, modes, and applications can be reconfigurable.(47) Another definition is that SR is where the digitization is at the antenna and all of the processing is performed by software residing in high-speed digital signal processors.(48) In order to avoid confusion regarding the capabilities of SDR and SR, it would be appropriate for the Commission to adopt a single definition for SDR as part of the rulemaking process.
NTIA urges the Commission to consider carefully the issues raised in these comments relating to the spectrum management and regulation of radio systems using SDR technologies. Respectfully submitted,
Gregory L. Rohde Kathy Smith
Assistant Secretary for Chief Counsel
Communications and Information
National Telecommunications and
William Hatch Information Administration
Associate Administrator U.S. Department of Commerce
Office of Spectrum Management Room 4713
1401 Constitution Avenue, N.W.
Edward Drocella Washington, DC 20230
Telecommunications Specialist (202) 482-1816
Office of Spectrum Management
June 16, 2000
ENDNOTES:
1. Inquiry Regarding Software Defined Radios, ET Docket No. 00-47, FCC 00-103 (rel. March 21, 2000) (hereinafter "SDR NOI").
2. See Joint Tactical Radio System (JTRS), available at http://www.jtrs.sarda.army.mil.
3. See Digital Modular Radio (DMR) http://dmr.mot.com/DMR.html.
5. Institute of Telecommunication Sciences (ITS), National Telecommunications and Information Administration, U.S. Department of Commerce, White Paper on Software Radios: Impact on System Review Process (unpublished) (hereinafter "ITS White Paper") at Appendix A.
11. IIT Research Institute, Software Defined Radio (SDR) Spectrum Management and Policy Implications, (hereinafter "IIT White Paper") (Aug. 20, 1999) at 3.
13. J. Mitola III, Technical Challenges in the Globalization of Software Radio, IEEE Communications Magazine (Feb. 1999) at 84.
15. For example, ITU-R Recommendation M.1343 was specifically developed to establish technical characteristics for and facilitate the circulation of Global Mobile Personal Communications by Satellite (GMPCS) mobile satellite service user terminals.
16. Working Party 8A, Draft New Question ITU-R [8A.SDR], Software Defined Radios, Document 8A/TEMP/99 (March 16, 2000).
18. PSWAC, Final Report of the Public Safety Wireless Advisory Committee to the Federal Communications Commission, Reed E. Hundt, Chairman, and the National Telecommunications and Information Administration, Larry Irving, Assistant Secretary of Commerce for Communications and Information (hereinafter "PSWAC Final Report") (Sept. 1996) at 547.
20. Very High Frequency (VHF) and Ultra High Frequency (UHF).
21. Air-interface is a radio to radio signal path defined in terms of access method, modulation scheme, vocoding method, channel data rate, and channel data format.
22. The Open Systems Interconection (OSI) reference model defines seven protocol layers: physical, link, network, transport, session, presentation, and application.
23. PSWAC Final Report at 239.
26. National Telecommunications and Information Administration, U.S. Department of Commerce, NTIA Special Publication 94-27, Spectrum Reallocation Final Report (Feb. 1995) at v; National Telecommunications and Information Administration, U.S. Department of Commerce, NTIA Special Publication 98-36, Spectrum Reallocation Report (Feb. 1998) at iv.
27. Institute of Telecommunication Sciences, National Telecommunications and Information Administration, U.S. Department of Commerce, NTIA Report 96-328, RF and IF Digitization in Radio Receivers: Theory, Concepts, and Examples (March 1996) (hereinafter "ITS Report") at 6.
29. National Telecommunications and Information Administration, U.S. Department of Commerce, Manual of Regulations and Procedures for Federal Radio Frequency Management, (Sept.1995 Edition) (Includes revisions for Sept. 1996, Jan. and May 1997) at § 5.3.5.2.
35. See 47 C.F.R. §§ 2.1046-2.1055.
43. Gain compression is the reduction in the output signal level of the desired signal of an amplifier sufficient to degrade performance.
44. National Telecommunications and Information Administration, U.S. Department of Commerce, NTIA Report 94-313, Analysis of Electromagnetic Compatibility Between Radar Stations and 4 GHz Fixed-Satellite Earth Stations (July 1994) at 25.
46. Bell South Cellular Corporation, A Briefing on Software Defined Radio Applied to Commercial Wireless (hereinafter "Bell South Briefing") (Feb. 1999) at 14.