Indoor Dominant Path Model

The dominant path model (DPM) determines the dominant path between the transmitter and each receiver pixel. So the computation time compared to ray-tracing is significantly reduced and the accuracy is nearly identical to ray-tracing.

Ray-optical propagation models are still time-consuming – even with accelerations like preprocessing. And what is even more important, they rely on an accurate vector database. Small errors in the database influence the accuracy of the prediction. On the other hand, empirical models are rely on dedicated propagation effects like the over-rooftop propagation (for example the direct ray COST 231). A comparison of prediction results computed with three types of prediction models is presented in the following figure:



Figure 1. Prediction results with three types of prediction models. Empirical COST 231 (on the left), intelligent ray-tracing (in the middle) and indoor dominant path (to the right).

Analyzing typical propagation scenarios shows that in most cases one propagation path contributes more than 90% of the total energy. The dominant path model (DPM) determines exactly this dominant path between the transmitter and each receiver pixel. So the computation time compared to ray-tracing is significantly reduced and the accuracy is nearly identical to ray-tracing.



Figure 2. A typical channel impulse response where one path dominates.

Empirical models (like COST 231) consider only the direct path between a transmitter and a receiver pixel. Ray tracing models (like IRT) determine numerous paths. DPM determines only the most relevant path, which leads to short computation times.



Figure 3. Comparison of different approaches (COST 231 on the left, ray-tracing in the middle and DPM to the right).

Advantages of the Dominant Path Model

As a consequence of the properties and restrictions of the available prediction models mentioned above, the dominant path prediction model (DPM) has been developed. The main characteristics of this model are:
  • The dependency on the accuracy of the vector database is reduced (compared to ray-tracing).
  • Only the most important propagation path is considered, because this path delivers the main part of the energy.
  • No time-consuming preprocessing is required (in contrast to IRT).
  • Short computation times.
  • Accuracy reaches or exceeds the accuracy of ray-optical models.

Typical Application of the Dominant Path Model

As the dominant path model does not require a preprocessing of the building vector database, it is ideally suited for large indoor areas. Additionally the approach to compute only the dominate ray emphasizes this operational area. The model does not compute the complete channel impulse response, as a result, if you are interested in the channel impulse response, the delay spread or the angular spread, a ray-tracing model is recommended. The DPM is the ideal approach to compute coverage predictions in large multi-story indoor environments.

Algorithm of the Dominant Path Model

The DPM determines the dominant path between transmitter and each receiver pixel. The computation of the path loss is based on the following equation:

(1) L = 20 log ( 4 π λ ) + 10 p log ( l ) + i = 1 n f ( φ , i ) + j = 1 m t j Ω
where L is the path loss computed for a specific receiver location. The following parameters are considered by the model:
  • Distance from transmitter to receiver ( l )
  • Path loss exponent ( p )
  • Wave length ( λ )
  • Individual interaction losses due to diffractions ( f )
  • Individual transmission losses of all penetrations ( t )
  • Empirical determined wave guiding ( ω )
  • Gain of transmitting antenna ( g t )

As described above, l is length of the path between transmitter and current receiver location. p is the path loss exponent. The value of p depends on the current propagation situation. In buildings with a lot of furniture (which is not included in the vector modeling) p = 2.3 is suggested, whereas in empty buildings p = 2.0 is reasonable. The function f yields the loss (in dB) which is caused by diffractions. The diffraction losses are accumulated along one propagation path. Reflections and scattering are included empirically. For the consideration of reflections (and scattering), an empirically determined wave guiding factor is introduced.

This wave guiding factor takes into account, that a wave propagating in a long close corridor will be reflected on the walls leading to less attenuation compared to free space. Thus, wave guiding effects can be expressed as an additional gain in dB. The transmission (penetration) losses are also accumulated along the one propagation path. The directional gain of the antenna (in direction of the propagation path) is also considered.

Configuration of the Dominant Path Model



Figure 4. The Configuration of Dominant Path Model dialog.
Path Loss Exponents
The path loss exponents influence the propagation result computed by the DPM significantly. The path loss exponents describe the attenuation with distance. A higher path loss exponent leads to a higher attenuation in same distance.


Figure 5. Three predictions using different path loss values. LOS exponent 2.0 (on the left), LOS exponent 2.3 (in the middle) and LOS exponent 2.6 (to the right).

The following table shows recommended path loss exponents. The more objects that are missing in the vector database, the higher the path loss exponent should be.

Table 1. Recommended path loss exponents.
Environment Empty buildings Filled buildings
LOS 2.0 2.1
OLOS 2.1 2.3
NLOS 2.2 2.5
Losses
Each change in the direction of propagation due to an interaction (diffraction, transmission/penetration) along a propagation path causes an additional attenuation. The maximum attenuation can be defined. The effective interaction loss depends on the angle of the diffraction. It is recommended to use the default value.
Waveguiding effects
As mentioned before, the DPM includes waveguiding effects to achieve accurate results. Two parameters are available to configure the wave guiding module. The first parameter is used to define the maximum distance to walls to be included in the determination of the wave guiding factor. With the second parameter the weight of the wave guiding effects in the computation of the path loss is defined (1.0 is suggested. Values below 1.0 reduce the influence of the wave guiding factor and values above 1.0 increase it). In typical scenarios the wave guiding effects can be turned off, this reduces the prediction time.
Breakpoint distance
The breakpoint distance describes the distance from the transmitter when destructive superposition of the direct ray and the ground reflected ray occurs. Because of that the attenuation is much higher and thus larger path loss exponents are often used. The default distance for the breakpoint distance in indoor environments is 500m.

Additional Features

The DPM supports additional features:
  • Combined network planning: The DPM offers the CNP mode which allows the combination urban and indoor predictions. In the urban environment the prediction is computed with the urban dominant path model (UDP), its settings and the resolution selected for the urban domain. For the predictions in the indoor area the indoor dominant path model (IDP) with a higher resolution is used. The settings of the dominant path model, such as path loss exponents and interaction losses, can be defined for both environments (urban and indoor) individually.
  • Multi-story prediction: The DPM is able to compute wave propagation predictions on several heights (layers) at the same time. This is useful for network planning in multi-story buildings.

Auto Calibration

The model can be calibrated automatically.