NIST 700 MHz Band Channel Propagation Model

Objective

• C.L. Holloway, W.F. Young, G. Koepke, K.A. Remley, D. Camell, Y. Becquet, "Attenuation of Radio Wave Signals Coupled Into Twelve Large Building Structures", Natl. Inst. Stand. Technol. Note 1545, Apr. 2008.
• K.A. Remley, G. Koepke, C.L. Holloway, C. Grosvenor, D. Camell, J. Ladbury, D. Novotny, W.F. Young, G. Hough, M.D. McKinley, Y. Becquet, J. Korsnes, "Measurements to Support Broadband Modulated-Signal Radio Transmissions for the Public-Safety Sector", Natl. Inst. Stand. Technol. Note 1546, Apr. 2008.

Measurements to date

Environments

The measurements were conducted in a large number of structures of varying dimension, shape, and function throughout the United States. The following table summarizes the environments:

A single measurement in a given environment consists of measuring the frequency response of the channel between a single transmitter (TX) located outside the structure and a single receiver (RX) located inside, typically 10-100 meters from the structure (where an incident command station may be located). The following figures depict a representative measurements environment, showing the façade of the Republic Plaza Building in Denver, Colorado and the placements of the transmitters and receivers.

Measurement system

The complex frequency response of the channel is measured using a vector network analyzer (VNA). The VNA operates by generating at Port 1 a continuous-wave tone at a given frequency, sending it through a device under test (DUT), and then measuring the attenuation and phase-shift the DUT imparts on the tone at Port 2, relative to Port 1. Sweeping a particular frequency band then allows characterization of the normalized response of the DUT. In our case, the DUT is the channel between the TX and RX antennas and the band is 698-806 MHz.

Omnidirectional antennas were used for the TX and RX antennas for frequencies in the 700 MHz band. A figure of the measurement system is shown below. The responses of the antennas and the measurement instrumentation were calibrated out of the measurements to render them completely independent from the test system. The reports cited in the Objective section provide specifics on the measurement system.

Channel model to date

Path-loss model

Path-loss is defined as the average signal power across the band of interest. It is conventionally modeled through the power law (single or dual slope) as a function of TX-RX distance. We have extended this model as a function of the number of walls between the transmitter and receiver to ensure a better fit. The frequency parameter has been introduced recently in UWB channel modeling (UWB frequency-dependence model) in relation to the strong presence of frequency fading across such a wide band. While the 700-800 MHz band is relatively flat by all measures, modeling any observed phenomena, such as the peak around 780 MHz observed in the frequency response below, enables a better characterization of any sub-band.

Temporal model

The temporal response lends readily to channel characterization and parameterization as it exhibits more descriptive features than the frequency response. In order to generate it from the latter, we opted to use super-resolution techniques over conventional Fourier methods due to their accuracy and reliably in isolating the arrivals for a100 MHz bandwidth. The figure in the sequel shows the temporal response generated from the measured frequency response above.

Our temporal model is based upon the Saleh-Valenzuela model (S-V model), where the arrivals are grouped into clusters. Once the arrivals are indexed according to delay and amplitude, we visually decomposed them into clusters. We observed one to five clusters in the temporal responses: the first cluster typically exhibits an exponential rise due to the non line-of-sight conditions, while the subsequent clusters exhibit exponential decays. In all, the model features four main parameters:

1. cluster decay parameter: this parameter models the exponential decay of the cluster envelope;
2. cluster delay parameter: this parameter models the temporal separation between neighboring clusters;
3. arrival decay parameter: this parameter models the exponential decay of the arrival envelope within a single cluster;
4. arrival decay parameter: this parameter models the temporal separation between neighboring arrivals within a single cluster.

Forthcoming measurements

In the near future, we have considered enhancing the measurements to date with a richer database, with focus on the urban-canyon environment. We invite feedback on the following factors:

Environment type

• Building type:
  • residential / high-rise / stadium / oil refinery / grocery store
  • construction material
  • number of floors
• metropolitan / rural

Channel conditions

• indoor – indoor / outdoor – outdoor / indoor – outdoor
• line-of-sight / non line-of-sight
• TX height, location (building / roof)
• TX mobile / stationary
• vacant / non-vacant

Antenna type

• directional / omni-directional
• single element / array (size, shape, number of elements)
• TX power, RX sensitivity
• polarization

Forthcoming channel model

MIMO applications

Replacing the single element of the transmitter and/or receiver with an antenna array could feed into the forthcoming model for Multiple-Input Multiple-Output (MIMO) applications. This would enhance the number of channels available between elements of the TX-RX taken pairwise to investigate channel diversity or small-scale fading. Otherwise the responses from the elements could be synthesized collectively to generate a spatial-temporal response which indexes the arrivals according to both time and angle in a two-dimensional profile.

Mobile applications

Doppler spread is a feature that could be included in a forthcoming channel model to use for mobile applications. In particular, it would be interesting to know the sorts of applications and the maximum speed of the transmitter (mobile station).

Location applications

Besides time-of-arrival extraction, it may be interesting to generate statistics on angle-of-arrival as well, in particular for location systems based on angle, or in combination with time-of-arrival for location systems based on joint time and angle (Joint range-angle system).

Contacts