Inclusion of the Receiving Antenna
Specify the receiver antenna pattern for a mobile station.
The mobile station is, by default, a single isotropic antenna. With this setting, the receivedpower result with the mobile station included is the same as for a regular propagationonly simulation, since the propagationonly simulation works with exactly that assumption when calculating the received power. A nonisotropic pattern can be specified by doubleleftclicking on Omnidirectional in the table.
Settings of the Mobile Station: Mobile Station / RX Tab
 Single antenna element
 Uniform vertical rectangular antenna array
 Uniform linear antenna array
 Uniform circular antenna array
 Individual location offset for each element
To specify a large antenna array, for example, for massive MIMO, the option Uniform Vertical Rectangular Antenna Array can be used (see Figure 6).
The image in Figure 6 has a mixed perspective. The grid of antenna elements is in the vertical plane and is rectangular, with horizontal rows and vertical columns. The angle $\alpha $ denotes the rotation of the entire vertical array around the vertical axis (East over North), and the angle $\beta $ denotes the azimuth of the individual antenna elements (East over North).
The image also indicates the numbering, which is row by row. The number of Rx elements is defined in the dialog and then the next integer that is a square is used as a basis. For example, in case of 6 elements defined, a 3x3 vertical array is the basis, and only the antenna elements for the two upper lines are considered. The maximum number of antenna elements is 64, for an 8×8 array.
As the computation of the stream results is very time consuming for larger numbers of elements, the stream results are automatically disabled if more than 4 Tx or Rx elements are defined. However, the subchannel results are computed and written if this option is activated. The memory consumption can become large for simulations with large arrays and many result pixels. Also, the disk usage for the result files can become large.
Settings of the Mobile Station: Channel Data Tab
MIMO systems can be evaluated in more detail by calculating the MIMO channel matrix, which describes the radio channel between each transmit and each receive antenna of the system.
There is a complex singleinputsingleoutput (SISO) channel impulse response of length L+1 between every transmit antenna m and every receive antenna n of a MIMO system.
In this equation, L relates to the tapped delay line, and the time dependence accounts for the possibility that antennas or objects are moving.
The linear timevariant MIMO channel is represented by the channel matrix with dimension ${N}_{R}\times {N}_{T}$ :
The MIMO channel matrix can be determined by postprocessing the ray data simulation output of a regular propagation simulation by just calculating the phase differences between the single antenna elements of the MIMO antenna arrays at the base station and at the mobile station.
The ray data give a description of all considered propagation paths between the position of the transmitter and each predicted receiver pixel. Field strength, delay and all interaction points (reflections, diffractions, transmissions, scatterings) are listed for the propagation paths that contribute to the signal level at a specified location. Based on these data and the dimensions of the MIMO antenna arrays, the phase shifts between the single elements can be computed in the following way. The transmitter location given in the ray file is assumed to be the center of the transmitting MIMO antenna array. At the receiver side, the same assumption is made. Each point to be computed in area/trajectory/point mode can be assumed to be the center point of a receiving MIMO antenna array. In order to determine the MIMO channel matrix now, only the phase shifts between the single array elements have to be computed, based on the ray data given in the ray file and on the array settings.
 Propagation Paths
 The ray data file used for the simulation is listed under Propagation Paths. The maximum number of rays, to be considered for each receiver location is set to 128 (can be only changed manually in the .mic file). This number has to be a power of two. If the ray data file contains more path data for a specific receiver location than specified here, the additional rays are skipped.
 Channel Type
 The radio channels can be modeled using the amplitude and phase information extracted from the ray data file (ray tracing simulation). Alternatively, the phases of the simulated radio channels can be derived randomly, for example, Gaussian distributed random variables are used for the phases. The corresponding amplitudes are taken from the ray tracing simulation. The third option can be used to simulate randomly derived Rayleigh fading channels with statistically distributed amplitudes and phases.
 Channel Bandwidth

The bandwidth is considered for the computation of the SNIR (noise power depends on the bandwidth) if the Calculate SNIR map check box is selected.
Furthermore the channel bandwidth is considered for the Frequency Sample Rate (see Results tab).
 Normalization of Channel Matrices

The calculated channel matrices can be normalized in various ways using different normalization algorithms either in time or frequency domain. The chosen normalization mode can be applied for all computation steps of the simulation or for the calculation of the MIMO channel capacity and Keysight PROPSIM output only.
The channel capacity of a nonfrequency selective MIMO channel can be written as
$$C={\mathrm{log}}_{2}\left(\mathrm{det}\left[{I}_{{N}_{R}}+\frac{P}{{N}_{T}\cdot {\sigma}_{n}^{2}}\cdot {H}_{F}\cdot {H}_{F}{}^{H}\right]\right)[\text{bit/s/Hz}]$$with the unity matrix I, the overall transmit power P and the noise power. The channel matrices H_{F} have to be determined by N_{F} point Fast Fourier Transformation. The FFT needs to be done for a certain number of sampling points (the number needs to be a power of two) which is by default 128. Frequency domain  Same path loss
 In order to compare different MIMO channels based on the
same path loss, the system has to be normalized to
fulfil the following condition:$$\sum _{i=1}^{{N}_{M}}{\displaystyle \sum _{m=1}^{{N}_{T}}{\displaystyle \sum _{n=1}^{{N}_{R}}{\displaystyle \sum _{l=0}^{{N}_{F}1}{H{}_{n,m}(l,i)}^{2}={N}_{M}\cdot {N}_{T}\cdot {N}_{R}\cdot N{}_{F}}}}$$where l is the loop index over the sample points in the frequency domain.
 Frequency domain  Same SNIR
 For comparison of different MIMO channels based on the
same SNR, the system has to be normalized to fulfil the
following condition:$$\sum _{m=1}^{{N}_{T}}{\displaystyle \sum _{n=1}^{{N}_{R}}{\displaystyle \sum _{l=0}^{{N}_{F}1}{H{}_{n,m}(l)}^{2}={N}_{T}\cdot {N}_{R}\cdot N{}_{F}}}$$
 SignaltoNoiseandInterferenceRatio
The signaltonoiseandinterferenceratio (SNIR) at the receiver locations is used for the calculation of the channel capacity. The SNIR can be specified to be a mean value for all receiver locations. This means the same mean value is considered for all specified receiver locations, independent of the actual received power at the receiver.
The Calculate SNIR map option calculates the SNIR for each receiver location separately taking into account the actual received power as well as the thermal noise (depending on the bandwidth) and an addition interference level, which can be also specified.The Load SNIR map from file option enables the user to load a previously computed SNIR map from the ProMan ray tracing simulation.
Settings of the Mobile Station: Results Tab
 Channel Matrices
An overall MIMO channel matrix can be computed per receiver point (Channel matrices per point which is activated by default). This result is derived by summing up coherently all ray contributions impinging on the receiving antenna element. The complex power values of the channel matrices can be written either as real and imaginary part or in amplitude and phase notation.
Additionally, the channel condition number can be computed as well by selecting the Channel Condition Number check box. The MIMO channel condition number is a MIMO performance indicator and computed as ratio of the largest and smallest eigenvalues of the MIMO channel matrix, with lower condition numbers allowing higher data rate transmissions. Results for all Modes (stationary and nonstationary scenarios)

 Generate per stream and subchannel power results

The subchannel results include the power for a single subchannel (between one Tx antenna element and one Rx antenna element).
The stream result includes always the best value from the subchannels (the higher value is always considered without adding anything, as a result, corresponding to selection diversity).
The
normal
RunMS power result includes the superposition of the subchannels (for example, power addition in case incoherent superposition has been selected).  Frequency Sample Rate
 This value is only considered for the ASCII subchannel power results. Defined frequency sample rate is considered together with the defined frequency channel bandwidth (under Channel Data tab) to compute the coherent power superposition for a set of frequency bins, which is done by recomputing the phase of each ray for the new frequency (loop over the frequency bins) and then summing the ray contributions coherently for this frequency bin.
 Results for Trajectories

 Doppler shift
 The option Doppler shift gives the Doppler shift per ray due to the velocity along the trajectory in [Hz]. Doppler shift results are only available if the receiver array is moving on a trajectory.
 Doppler spread
 Doppler spread is computed based on the power weighted mean
Doppler shift and represents the square root of the Doppler
variance for all ray contributions (squared difference of
ray Doppler shift minus mean Doppler shift, multiplied with
ray power, normalized with total power (same principle
formulas as for delay spread)).$${T}_{M}=\frac{{\displaystyle \underset{0}{\overset{\infty}{\int}}t\cdot {\lfloor \underset{\_}{a}(t)\rfloor}^{2}dt}}{{\displaystyle \underset{0}{\overset{\infty}{\int}}{\lfloor \underset{\_}{a}(t)\rfloor}^{2}dt}}$$$${T}_{\sigma}{}^{2}=\frac{{\displaystyle \underset{0}{\overset{\infty}{\int}}{(t{T}_{M})}^{2}\cdot {\lfloor \underset{\_}{a}(t)\rfloor}^{2}dt}}{{\displaystyle \underset{0}{\overset{\infty}{\int}}{\lfloor \underset{\_}{a}(t)\rfloor}^{2}dt}}$$Doppler spread results are only available if the receiver array is moving on a trajectory.
 Propsim Channel Sounder

WinProp radio channel data can be converted into the Keysight PROPSIM channel format (.asc file per subchannel and a .shd file).
In order to generate channel data, which can be directly imported into the Keysight PROPSIM F8 channel emulator, some requirements have to be fulfilled: Calculating the channel data for the channel emulator is only possible for a receiver moving along a trajectory. Therefore a corresponding receiver trajectory must be defined in the ProMan project file and considered in RunPRO. In order to guarantee proper emulation results, the distance resolution should be chosen in such a way, that the resulting evaluation points along the trajectory are located not more than half a wavelength apart from each other. Besides this, the velocity assigned to the defined trajectory points should be constant.
Generating channel data for the Keysight PROPSIM F8 emulator requires channel normalization of type Time Domain  Strongest ray.
Settings of the Mobile Station: Results – Optional Tab
 Mutual Antenna Coupling
 This option is disabled by default, as such an analysis is better handled using Altair Feko.
 Diversity Combining

Receive diversity can be evaluated using different combining techniques at the receiver array. The resulting diversity gain can be written for each receiver location as well as in terms of probability density functions and cumulative probability density functions.
The diversitycombining gain is added to the reported total received power under the RunMS results in the project tree. It will thus automatically affect the networkplanning results through an improved signaltonoiseandinterference ratio. The diversitycombining results are also provided as separate .txt files, where the computed diversity gain is given for each computed point (for example, along trajectory). Additionally, there is some information about the expected theoretical diversity gain mentioned in the header of the .txt file, for example: Theoretical diversity gain for Rayleigh fading channel: 3.01 dB
 Mean diversity gain of evaluated area: 2.99 dB
In order to improve the signal reception in mobile radio various diversity techniques can be applied, for example, by using multiple receiving antenna elements. Based on this, different diversity combining techniques can be utilized to combine the multiple received signals into a single improved signal. Selection combining (SC)
 For this option, the strongest signal is selected of the N received signals. When the N signals are independent and Rayleigh distributed, the expected diversity gain has been shown to be $\sum _{k=1}^{N}\frac{1}{k}$ , expressed as a power ratio. Accordingly the additional gain for increasing numbers of received signals is limited.
 Equal gain combining (EGC)
 For this option, all the received signals are summed up coherently.
 Maximum ratio combining (MRC)
 For this option, the received signals are weighted with respect to their SNR and then summed up. The resulting SNR yields $\sum _{k=1}^{N}SN{R}_{k}$ where SNR_{k} is SNR of the received signal k. This combining technique is often used for large phasedarray systems.
 Channel Capacity
 The ergodic and / or outage MIMO channel capacity is computed by calculating the eigenvalues for equal power or water filling power allocation mode. The results are available for each specified receiver location as well as in terms of probability density functions and cumulated probability density functions. The chosen normalization mode for the channel matrices and the considered SNIR (defined on the Channel Configuration dialog) are displayed here as these settings have a strong influence on the channel capacity.
After Specifying the Mobile Station
 Click .
 Click the RunMS icon in the project toolbar.
 Press F6 to use the keyboard shortcut.
 RunPro
 RunMS
 RunNet
Of the three simulation types, the propagation simulation (RunPro) tends to be the most time consuming. Sometimes you need to modify the properties of the mobile station, or modify related output requests, without losing the results of the regular propagation simulation. That can be achieved by accessing
.This option is not available (grayed out) in a new project. It becomes available when the Consider Antenna of MS check box was selected (  Propagation tab.