mmWave 28, 40 & 60GHz RF Modelling for Small Cells, 5G Mobile Networks & WiGig Applications

The deployment of high capacity links for 4G small cell backhauls and 5G urban dense cells, as well as the expanding utilization of the microwave and mm-wave spectrum for point-to-point and point-to-multipoint fixed wireless access systems have prompted the development of analytical and modeling tools for the characterization of the electromagnetic (EM) propagation in different environments, complementary to broad- and narrow-band measurements of the channel. In a recent study, we proposed a Ricean model for the received field in fixed links, comprising random scattering from irregular surfaces, identified from a coarse definition of the propagation topography as important contributors to the scatter.

Our publications formulates and discusses a propagation model for high-frequency (20, 40 & 60GHz) links operating in urban and sub-urban environments founded on an EM analysis of the environmental scatter and a realistic, three-dimensional (3D) depiction of the geometry. It is considered as more accurate than the standard Fresnel zone approach and of adequate computational efficiency for use in link design, with respect to antenna selection and positioning, and the estimation of the outage probability and cross-polar discrimination (XPD).

Key Features of our mmWave model:

-Supports point-to-point and point-to-multi-point LOS and NLOS indoor and outdoor links
-Comprehensive channel modeing
-Fast and accurate RF link predictions including RSSI, Coverage, Channel Impulse Response, Power Delay Profiles, etc
-Uses Lidar 3D representation of environment
-EM field calculations are based of meshed surfaces
-Incorporates realistic 3D antenna patterns
-Can be easily integrated with air-interface technologies

Details of the developed mm-wave propagation model:

A statistical model based on a detailed LiDAR 3D model (click here to see an example) representation of urban wireless channel topographies and the prediction of electromagnetic (EM) propagation in fixed point-to-point and point-to-multipoint (LMDS) wireless links operating at millimeter wavelengths (specifically targeting 28, 40, 60 & 80GHz bands) has been developed, with potential applications to the characterization, design and deployment of point-to-point and point-to-multi-point fixed wireless networks. The model uses realistic antenna radiation patterns. A specific goal of this model is the parameterisation of the main physical attributes of the propagation mechanisms, which at millimeter waves predominantly incorporate scatter from building and ground surfaces, both non-uniformly illuminated by a directive transmitting antenna. The line-of-sight (LOS) received signal is presented in terms of universal probabilistic distribution functions (PDF) and quantified in terms of their first and second order moments. Building, rooftop and ground scatter are modeled using Physical Optics (PO) approximations which have been validated against field measurements and experiments. Model derivations and details can be found here.

All results have been verified against extensive field experimental measurements (see figure below). Average of accuracies achieved are within 5dB. Figure below shows the meshing and the predictions from surface contributions to the received signal power for a P2P 40GHz link in a 600×900m suburban area (Trefforest Industrial Estate near the University of South Wales). The model presents facilitates specific site planning and coverage calculations. The computing running time on a typical PC (i3/i5) computer depends on the resolution of the LiDAR used, the link range and the antenna’s beamwidth. For example, for a horizontal resolution of 5m, a P2P distance of 650m and a 20dBi antenna, the code takes no longer than 20 seconds (in Matlab) to run through all the triangulated surfaces. Where shadowing is implemented the running time is increased (this of course depends on many factors).

Figure below shows coverage at 40GHz

See a high-image resolution example here

The K-factor is the ratio between the coherent and non-coherent fields, is an important parameter in link characterization and can be directly related to outage and interference probabilities. To learn more about K-Factors please our brief report here. Figure below shows the measured K-Factor values for a 680m link vs modeling using varies geometrical resolutions

Figure below shows predictions (with no shadowing) from a P2P 60GHz link in typical street canyon environment using 1m LiDar data resolution.

Figures below show the geometry of a NLOS 40GHz link, along with the measured and modeled results.

Site and geometry

Measured 360 degree pattern.

Measurement vs. predictions

The model can also predict coverage, received power distributions and Rician K-factors. Progress is still ongoing in this model to include: shadowing, MIMO & capacity calculations and channel’s impulse response plus others typical parameters.

For more information, please contact Dr. Zaid Al-Daher

[1] Al-Daher, Z.; Ivrissimtzis, L.P.; Hammoudeh, A. “Electromagnetic Modelling of High Frequency Links with High Resolution Terrain Data”, IEEE Antennas and Wireless Propagation Letters, 10.1109/LAWP.2012.2226135, Oct 2012.

[2] Muhi-Eldeen, Z.; Ivrissimtzis, L.P.; Al-Nuaimi, M. “Modelling and Measurements of Millimetre Wavelength Propagation in Urban Environments”, Microwaves, Antennas & Propagation, IET, Volume: 4 , Issue: 9, Publication Year: 2010 , Page(s): 1300 – 13.

[3] Z. Muhi-Eldeen, L.P. Ivrissimtzis, M.O. Al-Nuaimi, “Measurements and Modelling of Cellular Interference in Local Point-to-Multipoint Distribution Systems”, Muhi-Eldeen, Z.; Ivrissimtzis, L.P.; Al-Nuaimi, M.O.; Microwaves, Antennas & Propagation, IET, Volume: 3 , Issue: 2, Publication Year: 2009 , Page(s): 250 – 259.