This provides evidence for time-varying conditional second-order moment of the noise time series. Therefore, PLC technology can provide the maximum range of high-speed telecommunication services throughout the house.
PLC Ecosystem
Standards
European Telecommunication Standards Institute (ETSI)
TR Quality of Service (QoS) Requirements for Internal Systems": This document defines the classification of home applications, home devices, service categories, traffic categories and QoS requirements. TR Technical Requirements for Internal PLC Modems": Currently there are There is virtually no technical specification available for physical (PHY) and medium access control (MAC) layers.
IEEE 1901
International Telecommunications Union (ITU) G.hn
Major Players
- HomePlug Power Alliance (HPA)
- Universal Powerline Association (UPA)
- Consumers Electronics Powerline Communication Alliance (CEPCA) . 6
- OMEGA
- Powernet
- OPERA
OMEGA is an integrated project in the information and communication technology domain (ICT) financially supported by the European Commission under EU Framework Program 7 (FP7). The new cognitive BPL technology was recently developed based on the considerable knowledge of the shortcomings encountered in the previous field trials conducted.
Electromagnetic Compatibility
This represents a remarkable opportunity for the development, competitiveness and authority of the European broadband industry which will stimulate the creation of jobs, to the improvement of the information society technology in Europe and also the socio-economic prosperity in the EU [14].
Research Objectives
Cyclic spectral analysis is used to identify and analyze the second-order periodicity (SOP) of the measured noise time sequences. The cyclostationarity analysis mainly consists of estimating random aspects and periodic behavior of the impulsive noise via the spectral correlation density.
Thesis organisation
The cyclostationarity analysis mainly consists of estimating random aspects and periodic behavior of the impulsive noise via the spectral correlation density. An autoregressive GMM-driven model is finally used to model a short-lived cyclostationary impulse noise.
Contributions in terms of Journals and Conference Proceedings
Cyclic spectral analysis is used to identify and analyze the second-order periodicity (SOP) of the measured noise time series. This chapter aims to provide insight into some of the contributions in literature relating to power line communication.
Properties of Electrical Power Networks
High Voltage Networks
The main losses present in high-voltage lines are heat losses due to the resistance of the transmission line material and leakage losses [16]. Data communication over high-voltage lines at high frequencies is also subject to interference, which reduces the reliability of the system.
Medium Voltage Networks
Low Voltage Networks
Characteristics of Power Lines
Capacitance and Inductance
Appliances connected to the electricity grid are characterized by a certain inductance (L) and capacitance (C), each of which depends on the amount of current flowing through the circuit of the appliance. The inductance of an electrical circuit defines the amount of magnetic flux due to the current flowing in the circuit.
Impedance
PLC Noise
Periodic impulsive noise asynchronous to the mains frequency: this periodic interference occurs with a repetition rate in the range of 50 - 200 kHz. An overview of some of the prevailing approaches to modeling impulsive noise will be given in Section 2.6.
PLC Channel Modelling
Time Domain Approach: The Multipath Model
The magnitude of the channel frequency response can be used in determining the attenuation parameters (i.e. π0, π1 and π). The path parameters (ππ,ππ andππ) can be obtained from the channel impulse response of the channel.
Frequency Domain Approach: Transmission Line Theory Models
- Two-Conductor Transmission Line Models
- Multi-Conductor Transmission Line Models
In addition, the model neglected the effect of electromagnetic compatibility issues in the estimation of the common mode currents. Therefore, the modeling of the power line channel in the presence of a third or fourth conductor should instead consider multiple conductor transmission line theory.
Impulsive Noise Modelling
- Middleton Class A Noise Model
- Bernoulli-Gaussian Model
- Poisson-Gaussian Model
- Autoregressive Moving Average (ARMA) Models
Parameters In the case of determining the parameters of the ARMA(π, π) model, the process is usually not as easy as it is with the AR(π) and MA(π) models due to numerical stability issues [56, 57] .
Multi-Path PLC Channels
- Measurement Description & Instrumentation
- Channel frequency Response Measurements
- Practical PLC Channel Modeling
- Channel Response Characteristics
- Channel Model Parameters
- Channel Impulse Response
- Power Delay Profile
- Time-Delay Spread Parameters
- First-Arrival Delay (π π΄ )
- Mean Excess Delay (π π )
- RMS Delay Spread (π πππ )
- Maximum Excess Delay (π π )
- Relationship between RMS Delay Spread and Mean Delay
- Results Analysis of Dispersion
- Coherence Bandwidth
The dependence of the RMS delay distribution on the mean excess delay was analyzed in the measured CFR. The multipath nature of the PLC channel results in deep fading (nulls) in the channel frequency response. The coherent bandwidth of the channel is particularly relevant for frequency-hopping spread spectrum (FHSS) scheme and multi-carrier schemes such as OFDM [84].
Deterministic Channel Modeling
- Powerline Cable Parameters
- Determination of Propagation Parameters
- Channel Model Description
- Derivation of Model Parameters
- Model Description
- Simulation Results
These parameters, as shown in Figure 3-10, are affected by the electrical characteristics and physical dimensions of the transmission line in question. The power line as a network of distributed parameters can be seen as a cascade of sections such as shown in Figure 3-10 for a limited line length. Because of the different branch lengths, the individual parallel resonant sections will have different resonant frequencies, π0 in the frequency domain.
Summary and Conclusions
It is observed that the notch positions in the frequency response are related to the branch lengths in the electrical power network. In general, channel attenuation increases with the size of the power network in terms of the number of branches. Although this approach has a direct relationship with signal propagation in the network, determining such matrices is usually not easy and the difficulty increases with the size of the given power network.
Channel Description
The Lattice Approach
There are (π+ 1)networks at each branching node, where π is the total number of branches in the network. At the branch node, there are replicas of the signal that reach the node after bouncing off the left grid (πΏπ), bouncing off the right grid (πΏπ), passing through the right grid (πΏπ), and passing through the left grid (πΏπ ).
Model Formulation
Variables π0, π1 and πΎ are determined from measurements by a Mins-Absolute Residual (LAR) robust analysis of the attenuation constant. In our method, a homogeneous electric power network is considered as it corresponds to many indoor wiring practices, so the characteristic impedance of the main line is the same as that of the branches and does not differ between branches. The expected voltage level at a receiver placed at any of the branch elements is then determined.
Multipath Model
The number of paths depends on the size of the network and the position in the network where the receiver is placed. With the simplicity of the network diagram, it is possible to quantify the number of paths to each node in the network. Therefore, the results of Table 4.1 are modeled as a mathematical function, through pattern recognition techniques that can fit seamlessly to any network size for a receiver located anywhere in it.
Model Validation
Discussion of Results
The time period of the first notch (π‘0= 133ns) must correspond to twice the difference. measured data PRC model grid model. The structure of the topology in Figure 4-6 is such that the second branch has the same length as the first. If it is desired to determine the time period for the first notch due to the 7 m branch, we can follow the same procedure as before.
Conclussion
Noise Measurements
- Measurement Setup
- Measurement Instrumentation
The noise generated by numerous electrical appliances in the electricity network can be obtained from an electrical socket in an indoor environment. Communications equipment such as modems generally allow minimal voltage on their ports, which is entirely the contradiction in the electrical grid. To circumvent this shortcoming, we used the Tektronix RSA5126A Real-Time Signal Analyzer in the frequency range 1-30 MHz.
Measurement Results
The bitmap image, unlike a line trace, provides greater insight, allowing one to distinguish numerous versions of the same signal as it varies over time. The same figure also shows the instantaneous graph of noise, which essentially varies continuously around the average peak value. Considering Figure 5.2, while there is the advantage of identifying all the noise over time, it is clear that such a randomly varying process must be classified statistically over time.
Noise Distribution
As an example: the Gaussian distribution suggests that the probability of a noise component being 7.5 dB above the average power is 0.5, whereas the measured probability of the same power exceedance of 7.5 dB is 0.15. As an example: the Gaussian distribution predicts that the probability that a noise component is 2.5 dB above the average power is 0.7. According to the measurements, the probability that a noise component is 2.5 dB more than the average noise power is approximately 0.3.
Prediction of Asynchronous Impulsive Noise Volatility
- Time Series Data Acquisition & Treatment
- Generalized ARCH (GARCH)
- Determination of Conditional Means and Variances
- Simulation and Experimental results
This observation comes from the residuals of the time series after performing an OLS regression of the noise data sequence. In our approach, time-varying volatility is predicted using past time-varying variances in the error terms of the noise data set. To test for heteroskedasticity, an OLS regression of the measured noise time series is performed.
Conclusion
- PLC Noise Sequence Acquisition
- Cyclostationarity Analysis
- Cyclic Spectral Analysis: Application to PLC Noise
- Cyclic Coherence Function for Cyclostationarity Testing
- Simulation and results
As can be seen from the two figures, the cyclic coherence function reveals the underlying periodicities. a) Cyclic spectral coherence of the short-term noise. To clarify the presence of periodicities in the noise data, a cyclic spectral density (CSD) is calculated and presented in Figures 6-4(a) and 6-4(b). a) Cyclic spectral density of the short-term noise. The squared magnitude of the cyclic spectral coherence is shown along with its 1% significance level.
Parameter Estimation for Linear Regression Models
Data Acquistion & Treatment
To develop accurate regressive models for powerline time series, it is imperative that we transform the periodic noise sequence into a stationary one. It is still practical to develop periodic (non-stationary) time series models, only that they are suitable for small samples. It is common practice to use differentiation techniques to achieve stationarity, but unfortunately periodicity cannot be eliminated with this approach.
Generalized Method of Moments (GMM)
This is because the underlying theoretical model imposes constraints on the distribution of the data, while not fully specifying its form. That is, a population moment is a statement that some function of the data and parameters has a null expectation under the actual parameter value evaluation. ππ can (and generally will) depend on the data, but it is required to converge in probability to a positive definite matrix to properly define the estimator.
Simulation and Measurement Results
To operationalize the GMM estimator, π , the number of moments, will be required to be greater than or equal to π, the number of unknown parameters.
Conclusion
During this work, the characterization and modeling of internal PLC channels was carried out. In the case of the bottom-up strategy, two other channel models were developed. The simplicity of the models was outlined and their validation in all cases was shown.
Conclusions
- Chapter 1
- Chapter 2
- Chapter 3
- Chapter 4
- Chapter 5
- Chapter 6
Cyclic spectral analysis was used to identify and analyze the second-order periodicity of the noise time series. The chapter further discusses the estimation of asynchronous impulsive noise volatility in power line communication (PLC) by examining the conditional variance of the residuals of the noise time series. Numerical results prove that the proposed model is able to provide an accurate stochastic representation of impulsive noise in the frequency band of 1β30 MHz.
Possible future prospects
Galli, βA new approach to modeling the indoor power line duct part 1: circuit analysis and companion model,β IEEE Trans. Entrambasaguas, βModeling and Evaluation of the Indoor Power Line Duct,β IEEE Communications Magazine, vol. Gerald, βDealing with Unknown Impedance and Impulsive Noise in the Power Line Communication Channel,β IEEE Trans.
