Lab Manual Advanced Electronics Measurement 4 (20/21) I. INTRODUCTION
The wide spread of wireless mobile technology has motivated research activities towards the exploration of the millimeter (mm)-waves band (30–300 GHz) as a promising option for Fifth-Generation (5G) wireless communications. In fact, (mm)-waves band has become a promising option for fifth-generation (5G) near- infinite data rate ultra-low latency communications.
In spite of all recent achievements, engineers are still facing technical challenges due to the high susceptibility of extremely short-wavelengths to the propagating medium. Nowadays, small-sized room propagation models are gaining an increasing attention to become among the most important within indoor environments. They can be validated for the characterization of small-sized private offices, small-sized conference rooms, laboratories, residential rooms, etc…
Although the surfaces’ shapes and electrical properties of in-rooms channels are predetermined, however, their characterization is not a straightforward task as experienced at lower bands (<6GHz). In fact, minor dissimilarities in the structure might dramatically change the propagation characteristics of the channel. The disparity among the reported propagation characteristics of in-room sites is obvious, even for close sites dimensions and antenna radiation patterns.
Researchers are performing prior massive measurements to characterize indoor channels. For instance, the Path Loss exponent (n) ranges between (1.32–2.35) and (1.3–1.53) under Directional-Omnidirectional (D- O) and Omnidirectional-Omnidirectional (O-O) scenarios, respectively. The Root Mean Square Delay (RMSD) in nanoseconds (ns) span from 2.9 to 18.7 and from 4 to 35 for D-O and O-O cases respectively.
A recent study has been proposed for modeling beamforming based mm-wave propagation within a 20 m2 room using identical horns antennas having a gain of 20 dBi, and a half power beam width (HPBW) of 15o. The omnidirectional pattern was emulated by rotating the antenna in the azimuth angle. However, the blockage of the millimeter-wave signal by human movement results in an abrupt signal interruption. Millimeter wavelengths have a low diffraction ability around obstacles. A practical solution for ensuring a free movement of mobile users in the channel is to incorporate a donut pattern (omnidirectional) antenna at either or both ends i.e., the base station (TX) and the mobile user (RX).
In this LAB, broadband wireless experimental data are provided. You will learn how to characterise indoor channels and how to extract diverse channel parameters for the design of a 5G wireless link within a room office environment. The propagation measurements have been performed under Line-Of-Sight (LOS) environment. Two radiation patterns configurations have been considered: 1) TX: Directional, RX: Omnidirectional (D-O) and 2) TX: Omnidirectional, RX: Omnidirectional (O-O).
1
II. DESCRIPTION OF THE MEASUREMENT PROCEDURE
A. Measurement Site Description
The experimental environment is described as follow: A typical rectangular shaped conference room with dimensions of 5.83 m (width) × 6.88 m (length) × 2.67 m (height). The walls of this room are typical interior walls made of plasterboard sheets, the floor is concrete and covered with vinyl plastic tiles and the ceiling is also made of concrete and covered with polystyrene tiles and neon lamps positioned all over the ceiling. The room has two large windows made of glass, and a metallic office door. A top view of the considered office room is shown in Fig. 1.
B. Description of the Experimental Procedure
In the considered conference room, the measurement campaign was performed under LOS propagation scenario, where both the receiver and transmitter antennas were placed on mobile carts and kept at a height of 1.7 m above the floor level. Also note that the LOS joining both antennas is 2.5 m away from wall A in a parallel manner. The base station uses either a directional or an omnidirectional antenna that can be pointed towards the mobile to improve reception. The mobile terminal incorporates an omnidirectional antenna to allow the user to move freely around the coverage area. The transmitter was kept stationary; while the receiver was displaced with an increment of 0.5 m along the line joining both the transmitting and receiving antennas as illustrated in Fig. 1.
Fig. 1. Top view of the office room plan identifying the transmitter and the receiver positions in the room.
2
Figure 2. Photography of the measurement site
The measurements were based on the use of two antennas arrangements, Direct-Omni and Omni-Omni at both the remote transmitting and receiving ends, respectively. The omnidirectional antenna has been specifically considered to provide a vertically polarized 360o field in the azimuth plan. Hence two measurement scenarios will be investigated:
1. Omnidirectional antennas are employed at both remote ends of the channel (Omni-Omni).
2. A directional antenna is employed at the transmitting side, while an omnidirectional antenna is used at
the receiving side (Direct-Omni).
a) Directional Horn Antenna
b) Omnidirectional Conical Antenna
Figure 3. Antenna Setup
3
III. STATISTICAL ANALYSIS OF EXPERIMENTAL DATA
A. Frequency Response of the channel
The experimental procedure is described by measuring the frequency response, 𝐻(𝑓)𝑚𝑒𝑎𝑠𝑢𝑟𝑒𝑑 at each position of the receiver as illustrated in Fig. 1. The transmitter output power (as measured at the input to the antenna) was set at -6.44 dBm and -25.8 dBm for the Omni-Omni and Direct–Omni cases, respectively. The relatively low transmitter power was necessary in order to maintain the linear operation of the transmitter power amplifier, as well as to avoid saturating the receiver low noise amplifier at short link distances.
The complex dielectric constant is frequency-dependent, thus, the frequency-variant propagation characteristics must be examined across the whole band of interest prior to other analyses. Fig.4 shows a typical plots of the measured Complex Transfer Function (CTF) obtained at 5 m of antennas separation distance for Direct-Omni scenarios.
Fig.4. CTFs measured using Directional Tx, Omnidirectional Rx
B. Path Loss Gradient
Basically, the path loss, together with the transmit power and the minimum detectable receive power, determine the system coverage, i.e., the maximal distance between both terminals for acceptable communications. Large scale measurements were performed to determine the propagation distance-power law in the environment under consideration. After taking into account the hardware calibration information and antenna gains, the measured complex frequency responses data H(fi, d) were used to estimate the average received power for an arbitrary transmitter-receiver separation distance d. We stress that the averaging of the received power (Pr) over the fading was done on a linear scale (not in decibels).
4
𝑀𝑁
𝑃𝑟 (𝑑)= 10𝐿𝑜𝑔 {1 1∑∑|𝐻(𝑓,𝑑)|2 } −[𝑃𝑟 𝑑𝐵 10𝑀𝑁 𝑗 𝑑𝐵𝑚
𝑎𝑡1𝑚𝑟𝑒𝑓.]
𝑖=1 𝑗=1
𝑑𝐵𝑚
where, fj is the jth frequency tone, d is the TX-RX separation distance and N represents the number of frequency tones measured during a sweep, i.e., 10,001 discrete frequencies ranging from 59.6 GHz to 60.6 GHz and M is the number of frequency sweeps. The averaged PL is
𝑑
̅̅̅̅ ̅̅̅̅̅
𝑃𝐿 (𝑑)=−𝑃𝑟 (𝑑 )+10𝑛𝐿𝑜𝑔 ( )+𝜎
𝑑𝐵 𝑑𝐵𝑂 10𝑑𝑂𝑑𝐵
̅̅̅̅̅
where 𝑃𝑟 (𝑑 ) is the averaged received power in decibels at a reference distance do from TX (do = 1m), “n”
𝑑𝐵 𝑂
is the PL exponent and σdB denotes the location-dependent shadowing parameter. C. Time Dispersion Characteristics
Further analysis should be carried out to examine the dispersive characteristics of the channel, thus, the measured Power Delay Profile (PDP) has to be extracted to ensure the relevancy of the obtained data.
The PDP shows the received signal power as a function of time delay, giving an intuition inspection of the multipath channel.
The channel impulse response for each case is derived by performing the IDFT (Inverse Discrete Fourier Transform) to the measured Channel Transfer Function (CTF) using Hamming/Rectangular window.
Fig. 5 shows the experimental PDPs obtained for both Omni-Omni and Direct-Omni radiation combinations.
C. Root Mean Square (𝜎 ) and Channel Coherence Bandwidth (𝐵 ) 𝜏𝑐
The time dispersive 𝜎𝜏 is computed from the PDP as
∑𝛽2𝜏2 ∑𝛽2𝜏 2
𝜎𝜏=√𝑖 𝑖 𝑖−(𝑖 𝑖 𝑖) ∑𝑖 𝛽𝑖2 ∑𝑖 𝛽𝑖2
where βi and τi are the amplitude and the delay of the ith path
𝜎𝜏 can be extracted using various threshold levels below 0 dB peak of the normalized PDP. Such threshold
ensures the prevention of the noise level.
In many publications a simple relation between Bc and RMS delay spread is
𝐵𝑐50= 1 5.𝜎𝜏
5
which is valid for a correlation coefficient equal to 0.5. In case when the threshold correlation value is set to 0.9, approximate equation for the coherence bandwidth value is as follows:
𝐵𝑐90= 1 50.𝜎𝜏
Fig. 5 Experimental channel responses normalized with respect to 1m reference
6
IV. REQUIRED TASQ
Based on the measured data submitted on DUO and following the information as described in the lab manual, each student has to submit a complete report where the following parameters below has to be plotted and discussed.
1- Channel Frequency Response. 2-Power Delay Profile.
3-RMS Delay Spread.
4-Path Loss
Bonus Question (Not Mandatory)
Validate your measurements through Ray Tracing simulations (using Matlab or any other software) of the Small Office Room Environment taking into consideration all details cited in the lab manual.
7
Dr.Ismail Ben Mabrouk