In this section, the simulation method, used to explore the coexistence of digital TV and mobile RF signals under outdoor-to-indoor and indoor-to-indoor conditions, is presented. Furthermore, the proposed measurement testbed and its setup, used in this work, are introduced. The simulation and measurement campaign consists of the following:
-
1.
Simulation (propagation loss) and measurement of LTE performance in different locations (indoor and outdoor environment);
-
2.
Simulation (propagation loss) and measurement of DVB-T2-Lite performance in different locations (indoor and outdoor environment);
-
3.
Simulation and measurement of simultaneous transmission (signal propagation) of both LTE and DVB-T2-Lite RF signals in order to evaluate the influence of coexistence on the performance of both systems (on physical layer (PHY) level); and
-
4.
Identification of the noncritical (both systems can coexist) and critical (partial or full loss of DVB-T2-Lite and LTE signal) coexistence scenarios for both systems.
Simulation setup
The considered coexistence scenario was briefly outlined in the previous section. In this work, we assume that transmitters and receivers are located on the seventh floor (the top floor) in the building of Brno University of Technology (BUT), Faculty of Electrical Engineering and Communications (FEEC) in Brno. Laboratories of Digital TV Technology and Radio Communications, and Mobile Communications of the Department of Radio Electronics (DREL) are located on this floor. The floor plan of the seventh floor is shown in Figure 2. Approximate dimensions of the floor are 50 × 25 m. The HeNB is located in the Laboratory of Mobile Communication Systems (room 7107), and the DVB-T2-Lite transmitter is located outdoor on the terrace.
The whole simulation model is realized in MATLAB. Propagation of the LTE and DVB-T2-Lite RF signals are simulated separately. The simulation of separate propagation loss of LTE and DVB-T2-Lite RF signals will be used as the reference (no coexistence).
The simulation model consists of three main parts for both LTE and DVB-T systems. The first part represents the simulation of a link budget, according to the 3GPP recommendation for system level simulations [28,29] for both coexisting systems. Signal strength in the receiver can be expressed as follows:
$$ {P}_{\mathrm{RX}}={P}_{\mathrm{TX}}-{L}_{\mathrm{TX}\mathrm{C}}+{G}_{\mathrm{TX}\mathrm{A}}- PL+{G}_{\mathrm{RX}\mathrm{A}}-{L}_{\mathrm{RX}\mathrm{C}} $$
(1)
where P
TX is transmitter power, L
TXC are wiring losses, G
TXA is transmitting antenna gain, G
RXA is receiving antenna gain, L
RXC are wiring losses, and finally, P
RX is received signal level. Path losses in wireless transmission are denoted as PL (for details see Equation 3). Value of P
TX is known from the transmitter setup. Values of L
TXC, L
RXC, G
TXA, and G
RXA are constants depending on the used equipment (for details see Subsection 4.2). The second part represents the validation of obtained results from the simulation according to the performed measurement and their interpretation in a map. Details are in Subsections 4.1 and 4.2. The last part compares power imbalance of tested radio channels and computed achievable performance of both systems in certain locations. Details are given in Section 5.
The propagation scenario of the LTE RF signal in femtocell involves indoor-to-indoor line-of-sight (LOS) propagation for the same room where HeNB is located (room 7107) and non-LOS (NLOS) for other indoor locations. Path losses are modeled according to the 3GPP recommendation for indoor LTE femtocell as described in [28], denoted as UE to HeNB, where UE is inside the same building as HeNB. In order to model indoor-to-outdoor propagation from the HeNB to the measurement points on the terrace, the original equation was extended with outdoor wall penetration loss. On the other hand, the recommendations in [28] are generally valid for frequencies around 2 GHz, but we exploit an 800-MHz band in this study. Therefore, it is necessary to perform a correction as described in [29]. This correction defines the correction factor for 800 MHz as follows:
$$ P{L}_{\mathrm{COR}}=20{ \log}_{10}\left({f}_c\right) $$
(2)
where f
c
is carrier frequency in MHz.
The resulting path loss equation is:
$$ \mathrm{PL}=38.46+20 \log (d)+0.7{d}_{\mathrm{in}}+\mathrm{L}{\mathrm{P}}_{\mathrm{floor}}+q\;{L}_{\mathrm{INwall}}+n{L}_{\mathrm{OUTwall}}+\mathrm{P}{\mathrm{L}}_{\mathrm{COR}} $$
(3)
where d is the distance between the HeNB and the UE, d
in is indoor propagation distance, LP
floor is penetration loss due to propagation through the floor (it is equal to zero because a single-floor propagation scenario is assumed), parameter q is the number of indoor walls separating the transmitter and receiver, L
INwall is the penetration loss due to walls inside the building, n is the number of outside walls, L
OUTwall is the penetration loss of the exterior wall, and PL
COR is the frequency correction factor as defined in [29] and shown in Equation 2. In our case d
in = d, LP
floor = 0 (single floor), q > 0 (in the case of the NLOS scenario), L
INwall = 5 dB, n is between 0 and 5 (depending on the concrete position on the floor) and L
OUTwall = 10 dB.
The propagation scenario between the DVB-T2-Lite transmitter and TV receiver is considered as outdoor-to-indoor urban femtocell propagation, where the UE is outside as described in [29]. The DVB-T2-Lite RF signal attenuation with frequency correction can be calculated similarly to Equation 3 as:
$$ \begin{array}{l}\mathrm{PL}\kern0.5em = \max \left(15.3+37.6{ \log}_{10}(d),38.46+20{ \log}_{10}(d)\right)+\\ {}0.7{d}_{\mathrm{in}}+L{P}_{\mathrm{floor}}+q\;{L}_{\mathrm{INwall}}+n{L}_{\mathrm{OUTwall}}+\mathrm{P}{\mathrm{L}}_{\mathrm{COR}}\kern0.5em \end{array} $$
(4)
where all variables have the same meaning as in Equation 3.
No fading was included in the data displayed in Figures 3 and 4, however, both fast fading and shadowing were computed according to recommendations in [29].
Figures 3 and 4 show the results of LTE and DVB-T2-Lite radio signal propagation obtained from simulation, respectively. System parameters determined according to the simulation results prove accessibility of wireless services in all tested locations.
Path loss model data provides the basis for coexistence simulations. We have provided a detailed description of LTE and DVB-T2-Lite coexistence in our previous works [30] and [31]. Based on data collected from the mentioned measurements, we made a dense description (linear model) of coexistence. There are two types of input parameters for the models: global and local. The global parameters are mainly represented by the settings of both systems’ PHY, such as modulation used in DVB-T2-Lite, inverse fast Fourier transform (IFFT) size, and the Forward Error Correction (FEC) code rate of both systems. Obviously, the overlapping bandwidth is also a global parameter. Local parameters, used as the model input, are mainly local power levels of signals, background noise, and the local fading model employed. These parameters are input into the linear model, which maps them to the Quality-of-Service (QoS) parameters. More details can be found in Subsection 5.1.
Measurement setup
For evaluating the interaction of the described coexistence scenarios between DVB-T2-Lite and LTE RF signals, the same measurement testbed was used as described in our previous works ([30] and [31]). The whole measurement campaign was implemented on the seventh floor of the building of BUT, FEEC, DREL (see Figure 2). The measurement campaign and the basic principle of our measurement method are as follows.
Firstly, the parameters and performance of the 3GPP LTE network are measured in different locations on the seventh floor. At the time of LTE measurement, T2-Lite services were not broadcasted. The HeNB is located in room 7107, and its antennas are placed on top of a table (approximately 1 m above the floor). The HeNB consists of two main hardware components, namely a PC with the Fedora Linux operating system and universal software radio peripheral (USRP) N210 from Ettus, equipped with an SBX daughter card. The PC runs the commercial software package Amari LTE [32], implementing functions of LTE Mobile Management Entity (MME) and eNB (both are 3GPP LTE Release 9 compliant). A detailed configuration of the LTE network is summarized in Table 2. The receiving UE is Huawei e398-u15 (Huawei, Shenzhen, China) (LTE UE Cat. 3) [33], connected via USB port to a laptop equipped with the Rohde & Schwarz drive test software ROMES4. For receiving LTE services, the TechniSat Digiflex TT1 mobile antenna (TechniSat, Vulkaneifel, Germany) was used (G < 2 dBi). The length of its feed line is 3 m. The UE is connected to an external antenna placed on a wooden cart approximately 1.0 m above the floor. We set up the connection between UE and HeNB and performed simultaneous full buffer transmissions in uplink and downlink. The measurement was carried out in fixed points distributed on the seventh floor as shown in Figure 2. The receiving antenna was kept still for 2 min at each measurement point and in each location we have collected approximately 100 samples of each network parameter of interest (including RSS, Channel Quality Indicator (CQI), Error Vector Magnitude (EVM), etc.).
Secondly, we have measured the performance of the DVB-T2-Lite signal in different locations on the seventh floor. At the time of T2-Lite measurement, LTE services were not provided. By using the R&S single frequency unit (SFU) broadcast test system, an appropriate video transport stream for portable TV scenarios was generated. Then, the DVB-T2-Lite complete system configuration was set up, and the output signal was RF modulated (to the frequency of 794 MHz). For its amplification, a custom-built RF power amplifier (PA), based on hybrid module Mitsubishi RA20H8087M (Mitsubishi Electric, Tokyo, Japan) [34], was applied. This RF three-stage module is primarily destined for transmitters using FM modulation that operate in the range 806 up to 870 MHz, but it may also be applied in linear systems by setting the proper drain quiescent current with externally settable gate voltage. The PA was assembled according to the recommendations of the producers and thoroughly tested. The comprehensive measurement demonstrates that this PA can be used in a wider band, circa from 650 to 900 MHz, and can be used in the presented coexistence test. The gain of the PA strongly varies in the introduced frequency range from 36 to 50 dB, but in a narrow band, the gain is quite stable (max. 1.5 dB in 10 MHz bandwidth). The maximum output power of this amplifier is around 30 W. However, we practically used only 1 W (5 W was used for the scenario where power imbalances were equal to 20 dB) with quiescent drain current 4 A, gate bias voltage 4.3 V, and supply voltage 13.8 V to achieve high linearity for reliable application in the mentioned setup. Accordingly, the power efficiency in this setting is only 2%. On the other hand, reaching linearity is the fundamental parameter which needs to be set for minimizing any nonlinear distortion. For the used testing DVB-T2-Lite frequency (794 MHz), the measured 1 dB compression of this PA is 37.9 dBm (6.2 W), two-tone third-order intermodulation distortion (IMD3) (tone offset 1 MHz) is better than −38 dBc at the output power of 1 W, which corresponds to output intercept point (OIP3) 49 dBm. Between the PA and the antenna, there is an attenuator in the signal path. It serves as PA protection in the case of antenna switch-off or strong reflections in the antenna near the field. The JFW Industries 50BR-104 N attenuator (JFW Industries, Indianapolis, IN, USA) was used which was set to 0 dB during measurement. The mentioned nonlinear distortions caused by PA are not considered in our simulation model.
The used antenna is a multi-element Yagi antenna (G
max = 15.4 dBi) whose horizontal radiation pattern is shown in Figure 2. The feed line for the TV transmitter chain is a coaxial cable RG58 C/U which has a power loss of approximately 0.35 dB/m on the tested bandwidth. Attenuation of the auxiliary connection between ‘N’ and ‘BNC’ connectors is approximately 0.5 dB/m.
For the LTE system (HeNB), the Sectron AO-ALTE-MG5S antenna (Sectron Inc., Ormond Beach, FL, USA) was used. In our case, it was used as an omnidirectional antenna in vertical polarization (G < 3 dBi). After setting up the testbed, we moved with the Sefram 7866HD-T2 analyzer (Sefram Instruments and Systems, Saint-Étienne, France) to measure the received TV signal through all measuring points. The same antenna setup was used as is outlined above for LTE downlink. Once again, we spent 2 min at each measurement point for correctly evaluating the performance of the received DVB-T2-Lite RF signal (to avoid fast fading by averaging).
Figures 5 and 6 show measured and extrapolated values of RSS. Figure 5 shows the results of LTE radio signal propagation while Figure 6 shows the results of T2-Lite radio signal propagation obtained from measurement. System parameters determined according to the simulation results proves accessibility of wireless service in all tested locations. As we can see, results from measurement, shown in Figures 5 and 6, correspond with simulation results shown in Figures 3 and 4. This experimental result proves our simulation technique valid for coexistence applications.
Afterwards, the whole measurement campaign was repeated, but now both wireless services (DVB-T2-Lite and LTE) were provided together at the same time. The above outlined QoS parameters of both services, caused by coexistence between them, were measured separately with Rohde & Schwarz devices.