Ultra-Wideband Signals for Sensing and Communication:

A Master Plan for Developing Measurement Methods, Characterizing the Signals and Estimating Their Effects on Existing Systems

June 15, 2000

Revision of August 25, 2000 (3:15pm)

1. INTRODUCTION

This plan describes an approach to developing accurate, repeatable, and practical methods for characterizing the very narrow pulses (and pulse trains) of ultra-wideband (UWB) systems. The purpose of this characterization is to provide the information necessary to estimate or measure the potential for UWB systems to interfere with existing (narrowband, channelized, band-limited, and wideband) radio communications or sensing systems. The plan will incorporate tasks undertaken at the National Telecommunications and Information Administration (NTIA's) Office of Spectrum Management (OSM) in Washington, NTIA's Institute of Telecommunication Sciences (ITS) in Boulder, Colorado and the Department of Commerce's National Institute of Science and Technology (NIST) also in Boulder, Colorado, with ITS as the focal point. The plan will be closely coordinated with the spectrum management community, the users' communities associated with the potentially affected systems and the UWB developers.

2. BACKGROUND

The National Telecommunications and Information Administration (NTIA) manages the federal government's use of the radio frequency spectrum. NTIA authorizes operating frequencies for federal government radio stations so the agencies can meet their critical mission requirements without causing interference to either other Federal government users of the spectrum or those authorized by the Federal Communications Commission (FCC).

Recent advances in microcircuit and other technologies have resulted in the development of pulsed radar and communications systems with very narrow pulse widths and very wide bandwidths. These UWB systems have instantaneous bandwidths of at least 25% of the center frequency of the device.⁽¹⁾ UWB systems can perform a number of useful telecommunication functions that make them very appealing for both the commercial and government applications. These systems have very wide information bandwidths, are capable of accurately locating nearby objects, and can use processing technology with UWB pulses to "see through objects" and communicate using multiple propagation paths. However, the bandwidths of UWB devices are so wide that, although their output powers, in many cases, are low enough to be authorized under the unlicensed device regulations of the NTIA and the FCC, some of the systems emit signals in bands in which such transmissions are not permitted because of potential harmful effects on critical radiocommunication services.

The NTIA and the FCC developed rules for unlicensed devices (conventional electronic devices with narrow bandwidths) that did not address the then unknown UWB devices. Thus, NTIA and the FCC must work closely with each other and the users they authorize as well as with the UWB community to develop policies and procedures that will allow the UWB devices to work without interference to the existing systems. The difficulty in measuring both the UWB signal characteristics and their effect on other devices exacerbates the difficulties of this coordination. The pulses are much more narrow, often in the low nanosecond or picosecond range, requiring new measurement techniques and equipment to measure the signal characteristics accurately. Further, the interference effects of very narrow pulses with high repetition rates and aggregations of similar devices, such as could occur in some applications of UWB technology, are not well understood.

Ideally, the NTIA and the FCC need a comprehensive program that fully describes the UWB signals and determines their effects over a wide range of parameters in many potential victims. The program would also determine "acceptable" exposures for existing devices and establish a policy that will allow UWB devices to operate without causing interference to presently and futurely authorized and licensed systems. Because of the speedy development of the technologies involved in both UWB device design and innovative radio services using more conventional technology, we also require speedy policy decisions. Yet, the potential impact of these devices, if they are as successful as envisioned by their developers, could be significant on a number of critical safety-of-life systems. The measurement and analysis program, described below, is a basis for a more comprehensive program.

3. THE MEASUREMENTS PROGRAM - ITS

Most interference analyses are based on frequency-domain calculations or measurements to determine interfering power or power-density levels. At times, when needed, these are followed by time-domain analyses to determine additional, more subtle effects. The UWB signals probably will be most easily (and, most accurately) measured in the time domain. This is due to their inherent nature and the way the UWB pulses are generated in many devices, i.e., with a direct current impulse generator and tuned circuits. Given this, it must be determined how to use the knowledge of the time-domain characteristics in the analyses of interference in the frequency domain. It must also be determined if measuring a UWB signal in the frequency-domain is practical and useful. Then, the most accurate and repeatable way of doing these measurements and analyses must be determined and developed. Statistical metrics may be necessary because identical units from the same manufacturer may not produce identical pulses due to component differences. The approach should be general enough so that future UWB systems can be analyzed using the same approach.

Some UWB signals have been shown to produce a spectrum signature that is similar to noise (at least, when observed on a spectrum analyzer over some fairly narrow bandwidth). The time-domain characteristics that lead to a noise-like emission spectrum must be determined. The same will happen when many UWB systems are deployed near each other. The characteristics of the aggregate signal could be noise-like, but this must be well understood to assess the impact on users of the radio spectrum. Due to the wide variety of possible deployments of such systems and the wide bandwidth of the signals, developing, a priori, a model purely from measurements is not practical (or feasible). Instead one can use analytical methods to obtain an appropriate statistical model to determine spectral characteristics of aggregate systems randomly located in space. Validation measurements will be developed and performed to determine the efficacy of the analytical model under a variety of circumstances.

This Measurement Plan depends upon the services of NTIA's Office of Spectrum Management (OSM) in Washington, D.C., NTIA's Institute of Telecommunication Sciences (ITS) in Boulder, Colorado, and the RF Technology Division of the National Institute of Science and Technology (NIST) also in Boulder, Colorado, with ITS as the focal point. The recently-upgraded NIST facility is an excellent opportunity to obtain measurements of the radiated time-domain waveform from any UWB device. ITS may use other NIST facilities in Boulder, such as an anechoic or reverberation chamber. The intent is to use the NIST measurements as "reference grade" with a goal of developing well understood (and more practical) "laboratory grade" or "field grade" measurement methods. These latter methods must produce results to support analyses that are easily used to determine the effects of UWB signals on existing systems and on spectrum efficiency, in general. Other tasks seek to develop further understanding of the interference effects of actual or simulated UWB signals on several kinds of victim receivers, and the characteristics of an aggregate of UWB signals.

The following paragraphs describe the specific tasks that comprise the full project. The Gantt Chart at the top of the next page illustrates the timing and duration of the tasks. The ITS and OSM Leading support staff members are shown after each bar. Completion dates are based on the project start date of March 13.

Task 1: UWB Investigation and Measurement Plan

From a theoretical or analytical standpoint, examine the stated characteristics typical or known UWB systems to identify UWB parameters that need to be defined and measured for use in interference and spectrum analyses. Investigate UWB signal generation, processing, and radiation methods in use or being developed. Identify the basic or common characteristics of UWB pulses or signals needed to perform interference analyses or measurements, and spectrum efficiency studies. Various applications of UWB systems will be considered, reference-grade and practical measurement approaches will be identified, and a Measurement Plan will be developed.

Task 2: UWB Devices

Based on the results of Task 1 and availability, identify and obtain 2-4 UWB devices (transmitters) that demonstrate the range of characteristics in the parameters to be measured, which will be used in all of the planned measurements. These devices will be obtained through government and industry contacts. Identify and procure an impulse generator for more controlled and repeatable measurements and possibly a fast oscilloscope to make conducted emission measurements of time-domain signal characteristics. Procure several representative incidental or unintentional radiation devices for comparison.

Task 3: Full Bandwidth Time Domain Measurements

Obtain the high-accuracy measurements, defined in the Measurement Plan of Task 1, for radiated (open) and conducted (closed) system tests, if possible, of the UWB and unintentional or incidental radiation devices and the impulse generator obtained in Task 2. High accuracy measurements are required in the time domain along with digitized sampling of the time domain information for Fast Fourier Transforms (FFTs) in the frequency domain. We will compare the UWB pulse generator characteristics with identical UWB device characteristics for non-dithered, dithered and gated PRF formats to ensure that they accurately simulate the devices. The full bandwidth measurements will also provide the bases for determining and understanding the limitations and the accuracy for laboratory or field measurements using readily-available band limited test equipment also called commercial off-the-shelf (COTS) equipment as will be performed in Task 4. Measurements made with both full bandwidth and bandwidth limited COTS equipment will be compared to confirm the accuracy and reliability of the two separate measurements.

Task 4: Development of COTS Techniques and Bandwidth Limited Measurements

These measurements will investigate procedures using COTS (bandwidth limited) equipment to quantify UWB system characteristics (e.g., RF emission spectra, total peak and total average power, pulse widths, PRF, percent dither, and gating). The results of these COTS measurements along with theoretical calculations to quantify UWB system characteristics will be compared with the high-accuracy results of Task 3 to determine limitations and accuracy of these practical COTS measurements.

Task 5: IF Envelope Time Waveform and Bandwidth Correction Factor Measurements

Perform frequency and time domain measurements on UWB devices obtained in Task 2, according to the Measurement Plan in Task 1. Perform analyses to reduce measurements to standardized descriptions, if possible, of the interference potential from the UWB signals. This will include transfer properties of receivers to repetition frequency (PRF), non-dithered and dithered, gating, and receiver IF bandwidth. Spectrum analyzer measurement bandwidths of 10 kHz to 3 MHz (current RSMS) and to 50 MHz (FY-2000 RSMS upgrade) will be used. Receiver transfer properties will be described in terms of receiver IF output time waveforms (noise-like, CW-like, or pulse like), amplitude probability distributions (APDs) and bandwidth correction factors (BWCFs) as a function of average and peak power.

Task 6: Victim Systems (Receivers) and Interference Measurements.

Identify and obtain access to 3 to 4 victim systems (receivers). The following receivers are candidates (measurements with GPS receivers will be made under separate test plans at ITS or elsewhere):

Obtain or develop interference protection criteria (I_pk/N and I_av/N), and through measurements determine the minimum separation distance or maximum EIRP necessary to ensure that UWB emissions will not exceed the interference protection criteria of telecommunication systems. The UWB device or simulator power emission limits used in determining the minimum separation distance or maximum EIRP will include but not be limited to those defined by 47 C.F.R. Section 15.209. We will perform both conducted and radiated measurements. We will also test the interference effects of several incidental or unintentional radiators on the selected systems. We will compare the results of these measurements with the one-on-one interference analysis procedure (See Task 7).

Task 7: One-On-One Interference Analysis

Develop the capability to determine analytically the minimum separation distance or maximum EIRP necessary to ensure that UWB emissions will not exceed the interference protection criteria of telecommunication systems. The one-on-one interference analysis procedure will include hypothetical UWB devices whose average power emissions meet the limits defined by 47 C.F.R. Section 15.209 and an analysis of what UWB output powers or they require separation distances to meet the interference criteria established for the victims' receivers. The analysis procedure will include bandwidth correction factors (curves and equations) to permit conversion from UWB parameters to receiver IF output peak and average power (and I_pk/N and I_av/N) in a one-on-one situation.

The receiver transfer properties to UWB emissions will be based on ITS measurements in connection with Task 4, and simulations conducted by ITS to investigate peak and average power at a receiver IF output for various IF bandwidths as a function of UWB emission parameters such as, PRF, PW, and dithering. In addition, ITS measurements conducted under Task 5 will be used to validate the one-on-one interference analysis procedure.

Task 8: Aggregate Effects

A. Aggregate Model Assessment. Complete the review, comparison, and identification of key features of the 6 working aggregate models and 5 additional aggregate methodologies currently under study to determine their applicability to UWB issues.

B. Aggregate UWB Signal Into the Victim Receiver. We will develop an analytical, statistical model to predict the spectral characteristics of the aggregate signal emitted from UWB systems located randomly in space. The initial model will assume idealized pulses and an evenly distributed array of UWB emitters. The next step will be to vary the locations of the emitters and use actual measured pulse shapes. We will further refine the model, as required, to represent the expected deployment of operational systems realistically. Use the model to determine how many and what kinds (time-domain characteristics) of UWB signals must be present to allow the aggregate signal to be treated as noise. We will make measurements using two or more devices to validate the aggregate model. All activities on this task will be coordinated with related activities at OSM (aggregate studies) and ITS (simulation studies).

ITS will develop an analytical model that describes the aggregate signal. This analytical model will allow one to determine the characteristics of an aggregate of UWB signals from the characteristics of the individual signals. Our goal is to express the nature and characteristics of an aggregate of UWB signals - both "in space" and "in the receiver." Although the aggregate signal in the time domain can be determined in an infinitely wide bandwidth, the receiver's narrower bandwidth affects the UWB signal, and one must consider the effects of the receiver on each of the interfering signals.

In the case of one UWB signal getting into one receiver, the UWB signal has a nanosecond or narrower pulse and an emission spectrum of up to several GHz. That very wide emission exists only for the duration of the pulse. If the receiver response time is much slower than the pulse's width, the IF circuits could ring for some time after the pulse passes at frequencies related to the receiver IF bandwidths. Thus, the effects in the receiver are as much due to the receiver response as to the interfering signal characteristics and will be different for different receiver characteristics. We need to understand how a receiver responds to an impulse, not to the averaged (over time) power spectral density, and, how and under what conditions we can develop and use a bandwidth correction factor. These are the key issues we intend to explore under the "aggregate" task in the plan.

The "aggregate" task in the plan is to develop an analytical model, coupled with simulations ITS is doing on another task. The analytical work here will help confirm our simulations, and some measurements of effects on receivers will further confirm or refute either or both of the above.

C. Define and Implement UWB Aggregate Model (Average Power). Based on results of the above and confirmations in Task 6, complete an aggregate UWB model to compute the average aggregate power and I_av/N at a victim receiver based on various fixed and random geographic distributions of emitters. The model will consider the methodology in ITU-R Recommendation M.1316. For random emitter distributions, the statistics of the power will be computed as well. In connection with Task 7, propose several measurements to conclusively confirm the fundamental aggregate premise that the average powers received from multiple UWB emitters add linearly.

D. Define Methodology for Peak Aggregate Power. Research to date has not identified any methodology to determine the aggregate peak power (or aggregate I_pk/N) in a victim receiver. However, we can define several limiting cases for very narrow receiver bandwidths and very wide receiver bandwidths based on combinations or aggregation of I_av and I_pk from individual emitters. Study will continue to define more general methods to determine aggregate peak power.

E. Upgrade Aggregate Model to Compute Peak Power. Based on results of D, upgrade the aggregate model to add an aggregate peak power calculation.

F. Validate Model. In connection with ITS task 7, define several tests to confirm the results of the aggregate model for two and possibly three interferers in various combinations of powers, distances, parameters, etc.

G. Extend Results. Based on the results of the model validation completed in F, exercise the model to extend the results to a wider set of cases including airborne, high density, indoor/outdoor situations, and other cases of interest by the OSM/ITS UWB team.

Task 9: Interim and Final Reports

Prepare a comprehensive formal report to include recommendations for accommodating UWB systems.

Project Management:

Office of Spectrum Management - Paul C. Roosa, Jr.

Institute of Telecommunication Sciences - Dr. William A. Kissick

1. There are several ways of generating very wide signals including spread spectrum and frequency hopping techniques. The UWB signals for the devices of concern in this plan are generated by direct current impulse responses fired into a tuned circuit. This generates a burst of energy of ideally one positive going cycle shaped by the tuned circuit to a specific portion of the spectrum.