**3. Case study 2: indoor channel modeling**

The second case concerns the evaluation of the MIMO channel capacity in a deterministic indoor channel model and comparison against a stochastic indoor channel model. The indoor floor plan shown in **Figure 3** is provided by Remcom. It has a width of 66 m, a length of 35 m and a height of 3 m and consisting of 23 offices, one main lobby and two big office areas containing 22 desks. Wood, concrete, glass, drywall and metallic materials were used to build the model. It provides a multipath-rich environment which is essential for MIMO applications. A corresponding SCM is created in MIMObit with similar dimensions using TGn 802.11n channel model B which is used to represent indoor office environments with NLOS conditions. A two-element MIMO antenna array is used as a transmitter and two-element half-wavelength dipole arrays are used as the receiving antennas in both models.

**Figure 3.** *RT indoor model.*

The transmitting antenna array used in both the RT and SCM tools is a modification of the design reported in [4] which is a miniaturized two-element monopole antenna array decoupled using a frequency selective structure, mounted on a grounded dielectric substrate and fed by two coaxial cables. The array operates at 3.7 GHz with a bandwidth of 160 MHz extending from 3.62 GHz to 3.78 GHz as determined by the S-parameters presented in **Figure 4**. The antenna has a minimum S11 of �30.7 dB and a reduced mutual coupling below �20 dB with a minimum of �38 dB over the operating bandwidth. The array achieves orthogonality between the main lobes of the 3D radiation patterns of the antenna elements as shown in **Figures 5** and **6**, which is favorable for spatial diversity and multiplexing. Each element pattern has a maximum gain of 7.01 dBi at boresight.

The array is placed 36 m from the west main wall and 6.5 m below the north wall of the floor plan (blue spot in **Figure 3**) at a height of 2.5 m and is rotated 180<sup>o</sup> across the length as an Access Point (AP) with an input power of 1 W. 1056 twoelement half-wavelength dipole antenna arrays operating at 3.7 GHz are uniformly distributed over the model at a height of 1.5 m and are used as the MT receiving antennas. The MT dipole elements are separated by 2λ where λ is the free-space wavelength at the operating frequency. The maximum number of ray reflection,

**Figure 4.** *Tx/Rx antenna S-parameters.*

**Figure 5.** *3D antenna radiation pattern.*

*Stochastic versus Ray Tracing Wireless Channel Modeling for 5G and V2X Applications… DOI: http://dx.doi.org/10.5772/intechopen.101625*

#### **Figure 6.**

*2D antenna radiation pattern (θ is the complement of the elevation angle and* ∅ *is the azimuth angle).*

transmission and diffraction per path is seven, one and two, respectively. These numbers are chosen after a trade-off between simulation time and accuracy. The space between the transmitted rays is chosen as 0.25° . MIMO open-loop scheme with no channel-state information is chosen as the MIMO scheme with no precoding or beamforming using 20 MHz bandwidth. Equal gain combining is used as the combining method operated at the receivers. With the absence of interference in the channel, each MT will experience a unique multipath from the AP and hence resulting in different MIMO channel capacities as shown in **Figure 7**. The average capacity over the 1056 locations considered in the simulation is displayed in **Figure 9** for different SNRs ranging from 5 to 30 dB as the deterministic channel model MIMO capacity results.

A similar scenario is created using MIMObit's TGn 108.11n model B SCM where the AP antenna is used at the same coordinates. However, only one receiving MT antenna is placed at (�10 m, �5 m, 1.5 m) as shown in **Figure 8** and the average capacity is computed over a certain number of channel realizations in the time domain. The model is simulated over 1000 channel realizations where the channel environment changes at each realization resulting in a different set of multipath experienced by the signal traveling from the AP to the MT and hence resulting in a different MIMO channel capacity. The number of realizations has been chosen to achieve statistically reliable results. Comparison of the average MIMO channel capacity is shown versus SNR ranging from 5 to 30 dB in **Figure 9**.

The MIMO channel capacity in the SCM is higher than the capacity obtained from the deterministic channel model at SNR values larger than 20 dB. A

**Figure 7.** *Indoor model channel capacity.*

**Figure 8.** *SCM indoor model.*

**Figure 9.** *Indoor MIMO channel capacity.*
