- TDMA versus CDMA
- OFDM for wireless multimedia communications
- Universal Wireless Personal Communications / Edition 1
- Full text of "OFDM For Wireless Communications Systems"
- Duplicate citations
Figure 10 Reflectogram of S 1. Figure 11 Reflectogram of S 2. Figure 12 Reflectogram of S 3. After interference mitigation in distributed reflectometry, we propose now to integrate communication between sensors via the transmitted part of the test signal which has never been done with conventional methods [ 9 , 10 ]. For this reason, the test signal must be capable of carrying information which is possible thanks to the OMTDR method [ 11 ]. In this section, we propose to integrate communication between sensors to enable data fusion in the context of distributed diagnosis.
For this reason, we propose to use not only the reflected part of the diagnosis signal, but also the transmitted part. A signal carrying information is then used as test signal to enable reflectometry measurement and communication through the OMTDR technique. To do so, let us begin with the structure of the test signal.
As the test signal is carrying information, the data is formatted into frames themselves subdivided into 9 fields. After having described the frame structure, we propose now to classify the distributed sensor into two groups: master and slave. To do so, we propose to assign a weight of eligibility to each sensor for sensor classification. In fact, the reflectogram's quality depends strongly on the network topology in terms of distance and number of junctions [ 1 ].
The same remark holds for the communication quality. We propose now to study the impact of network topology on communication quality. We focus only on the number of junctions in the network. Recall that a junction causes the reflection of a part of the energy of the transmitted signal. Figure 14 shows the different topologies considered in order to calculate the BER. Figure 15 shows the evolution of the BER versus the number of junctions in the network.
TDMA versus CDMA
It may be noted that the BER depends on the complexity of the network topology in terms of junctions number. Indeed, the increase of the number of junctions causes the increase of the attenuation of the signal during its propagation. Figure 15 Evolution of bit error rate in terms of junctions number. Based on these findings, the weight of eligibility may be calculated by the following parameters.
- How to Drop Your Misery and Suffering!
- PAPR Reduction Technique for High Speed Application.
- A Fire in the Sun.
- Refine your search.
The minimization of this value reduces the propagation attenuation and hence the bit error rate. The minimization of this value reduces the bit error rate due to multiple reflections as shown by Figure In fact, the minimization of the weight of eligibility reduces firstly the bit error rate and increases the diagnosis accuracy since it minimizes the attenuation of the test signal. Then, the sensor with the lowest weight of eligibility is designated as the master while other sensors are considered as slaves. Besides network diagnosis signal injection, received signal processing, fault detection, etc.
For their part, slaves must do their diagnosis, identify the fault position, and send it to their master. In this section, we propose to develop an algorithm to automate the detection and location of a fault. We propose firstly to generate a reference measurement obtained when the network is healthy. We propose to save in sensor memory only the position of the local extrema of the corresponding reflectogram to avoid the saturation of the embedded memory.
The number of extrema in the reference is noted N ref. Figure 16 describes the proposed algorithm for detecting and locating automatically a possible fault. After the construction of the reflectogram, we extract local extrema noted e curr p curr i , a curr i. Then, we compare it in terms of position with those stored in memory reference. This indicates whether there has been an evolution of the state of the network or not. If there is no change, we must ensure that all local extrema are treated. Figure 16 Algorithm for detecting and locating faults in a single measurement. The algorithm described above allows automatic detection and location of a fault in a single reflectometry measurement.
Indeed, saving only local extrema permits to optimize both processing time and memory capacity. Thereafter, the position of the detected fault is encapsulated in the field data of the frame to be sent to the master if the actual sensor is a slave.
The soft fault position is stored in the memory of sensor S m. This position is also stored in its memory.
OFDM for wireless multimedia communications
Note that the processing of the measurement is done locally. For this, the slave must have a good memory and processing capacity. The sensor must, every time, analyze the new reflectogram and compare it with that obtained at the previous time to check if the fault persists, if it has evolved amplitude variation, increasing the length, etc. At the reception, the master S m extracts the data sent by its slave and stores it in its memory. After receiving data sent by all its slaves, the master analyzes this data and makes the decision about the fault location in the network. In this example, the fault is located on branch B 1 as shown by Figure Moreover, it may provide information about the state i.
We propose to verify the efficiency of data fusion strategy in a CAN bus system. In this section, we consider the CAN bus system described in Figure We consider the presence of a soft fault with length of 0.
Universal Wireless Personal Communications / Edition 1
Here, the master manages 5 slaves. Firstly, we calculate the weight of each reflectometer using Table 2 shows the weight of eligibility of each sensor. It may be noted that both sensors S 3 and S 4 have the lowest weight. If we were in a heterogeneous case, we could differentiate between the two sensors by another metric such as reliability, computing, or memory capacity and so forth. However, we have assumed a homogeneous case in this paper. As a result, we can choose either sensor S 3 or S 4. In this case, we will consider the sensor S 4 as the master.
Using the strategy described above, each slave must detect and locate the soft fault and send it to its master S 4. Figures 19 and 20 show reflectograms obtained by salves S 5 and S 6 , respectively. The positions of the fault are then sent to master S 4. After receiving all data of its slaves, the master makes the decision on the location of the fault in the whole network. Table 3 shows the available data at master S 4.
Full text of "OFDM For Wireless Communications Systems"
Given that the network topology is already known by the master, it is able to locate the fault on branch B 3. It is noted that the amount of information depends heavily on the complexity of the network topology and the number of sensors. This directly affects the decision-making time. Table 3 Fault location on branch B 3.
Sensor fusion is an innovative solution in the field of reflectometry. The sensor fusion allows the centralization of information and facilitates decision-making about the fault location in the whole network. Figures 21 and 22 show reflectograms of slaves S 5 and S 6 , respectively. Note that the soft fault can not be detected either by sensor S 5 or by sensor S 6 because of signal attenuation after 5 or 6 junctions. Thus, both sensors always send information about the fault previously detected on branch B 3.
- Citations per year.
- Technology Trends in Wireless Communications.
- ACCUPLACER®: Bob Millers Math Prep (College Placement Test Preparation).
In this case, there is a fault location ambiguity relative to the master S 4 as shown in Table 4. In the context of complex wiring networks, data fusion strategies suffer from signal propagation phenomena attenuation and dispersion which affect the diagnosis reliability for reflectometry measurement and data credibility for communication.
In addition, the increase of complexity of the network topology comes with the increase of the amount of information, the time of information analysis and decision making. When a hard fault open circuit or short circuit appears, the master may be unreachable. As a solution, we propose a sensor clustering strategy.