Corona Noise in High Voltage Power Line Communication(PLC) using OFDM

Transcription

1 Corona Noise in High Voltage Power Line Communication(PLC) using OFDM B. Priyalakshmi & Abhishruti Bhuyan Dept of Telecommunication Networks, SRM University, Kattankulathur, Tamil Nadu, India b.priyalaksmi@gmail.com,abhishruti29@gmail.com Abstract Power Line communication (PLC) is a popular communication channel for broadband access, multimedia sharing, and as a part of the smart grid since it allows for fault detection, automated meter reading, etc. However, various noise and multipath effects are the main causes of bit errors in power-line communication. PLC has been studied under various communication protocols such as 4D-TCM with additive white Gaussian noise(awgn). PLC can be implemented in high voltage(hv) power lines but HV lines are prone to corona noise. The objective of this research is to investigate the use Orthogonal frequency division multiplexing (OFDM) with convolutional coding and decoding to use a PLC transmission system with corona noise. In this paper, simulation of noise in PLC is by adding AWGN and corona effects in high voltage (HV) transmission lines is explained. The PLC for Zimmermann and Dostert, is taken in to consideration for simulation. Simulation of OFDM with 16-QAM is performed. The bit error rate (BER) in OFDM PLC communication with corona noise has been determined. Keywords Corona noise, Orthogonal Frequency division Multiplexing, Power Line Communication, Gaussian Noise I. INTRODUCTION Nowadays power line communications has become an important subject of research work. PLC appears to be a promising alternative to conventional technologies such as digital subscriber line (DSL) particularly in rural or underdeveloped areas where the conventional telephone line is still not available to a large population at the global level. PLC is an attractive alternative choice for traditional networks due to its ability to offer broadband internet access, cable television, telephone service and home automation. At the same time, the growing demand for multimedia communications provides a good prospect for PLC as a promising transmission technique. However power line communication transmission is affected from many problems such as interference, multipath noise, attenuation delays, presence of echoes, frequency selective fading due to multipath etc [1]. So it is necessary to employ a suitable modulation scheme such as orthogonal frequency division multiplexing (OFDM) to counter its unwanted effects on signal transmission. OFDM is a spectrally efficient multicarrier modulation technique for high speed data transmission over multipath fading channels. It distributes data over a large number of sub-carriers spaced apart at precise frequencies, such that they are orthogonal to each other. Recently, there has been a growing interest towards the possibility of exploiting existing power lines as effective transmission means. Low-voltage (LV) and medium voltage (MV) power lines, below 1 kv and from 1 to 36 kv, respectively, are advantageous because they a potentially convenient and inexpensive communication medium for control signaling and data communication. High-voltage (HV) power lines, typically operating at or above 64 kv, can also be used for communication purposes. In low or medium-voltage power grids, several noises can be found, such as- (i) non stationary colored thermal noise with power spectral density decreasing as the frequency increases, (ii) periodic asynchronous impulse noise related to switching operations of power supplies, (iii) periodic synchronous impulse noise mainly caused by switching actions of rectifier diodes, and (iv) asynchronous impulse noise [5]. On the other hand, performance of the HV power line channel is also limited by disturbances produced by events outside the transmission channel such as, for example, atmospheric phenomena, lightning, or disturbances originating within the system such as network switching, impulse noise, and corona phenomena Appropriate OFDM techniques in PLC has been studied and developed in [2]-[4], however one the most adverse effect of HVPLC- corona noise has not been 116

2 addressed for OFDM system. In[5], an optimal 4-D TCM detector has been proposed considering coloured channel noise in HVPLC where corona noise models has been developed based on the work of [6] on digital simulation of corona noise for various HV power lines. This paper discusses the effect of i) Gaussian noise in low voltage and medium voltage PLC and ii) corona noise in HVPLC. The bit error rate (BER) in OFDM PLC has been determined. II. OFDM AND PLC SYSTEM The OFDM based receiver structure proposed [1] (Fig.1) consists of an encoder, a BPSK modulator, Inverse Fast Fourier Transform (IFFT) and Cyclic prefix (CP) insertion respectively in the transmitter side and vice-versa in the receiver side. The coder adds the redundant information to the sequence of bits. If there is an error in bit chain the redundant information could be used for an error detection and correction by the help of detection and correction coders. The coded bits are then mapped to binary phase shift-keyed (BPSK) symbols. Inverse FFT (IFFT) of the BPSK symbols is taken and a cyclic prefix of length is added to construct the transmitted symbols. The main focus of the algorithm is that-the transmitted signal s(n) passes through the power-line channel h(n), and white non-gaussian noise w(n) and corona noise is added, giving rise to the received signal- y (n)=s(n).h(n) + w(n) (1) the multipath model proposed by Zimmermann and Dostert[8]. The power line can be defined as a multipath channel caused by reflections generated at the cable branches through the impedance mismatch. The Zimmermann model works in the frequency range from 500 khz to 20 MHz. The identity of this model is that it is based on physical parameters developed from propagation effects and not from properties of the components used in the network, as in the case of other models. A simple 4-path PLC is shown in Fig. 2. The node-a,b,c and D/E are the path-1,path-2,path3 and path-4 topology. Fig. 2 : A Multipath PLC topology Following this model, the frequency response of the channel may be expressed, in the frequency range from 500 khz to 20MHz, as [1]- (2) where N is the number of relevant propagation paths, a 0 and a 1 are link attenuation parameters, k is an exponent with typical values ranging from 0.5 to 1, g i is the weighting factor for path i, d i is the length of the i th path, and v p is the phase velocity. In this work we consider a PLC channel modeled in [8] with parameters shown in Table I. III. PROPOSED CORONA NOISE MODEL Fig. 1 : OFDM transmitter-receiver system A. PLC Model One of the most important elements in the PLC system model is power line channel model. Several approaches for modeling the transfer function of power lines can be found in the literature [7]. The models are represented in a multipath propagation environment. The most widely known model for the PLC channels is Corona noise is a common noise source for HV transmission lines due to discharge of electric field between two conductors. The corona noise intensity depends on (i) the transmission voltage, (ii) the size and spacing of the conductors (iii) the type of conductors involved in the line and (iv) the weather conditions [5]. Corona noise is caused by partial discharges on insulators and in air surrounding electrical conductors of power lines.when HV power lines are in operation, the voltage produces a strong electric field around the 117

3 H(f)in db H(f)in db ITSI Transactions on Electrical and Electronics Engineering (ITSI-TEEE) conductor. This electric field accelerates free electrons present in the air produces collision among the molecules o the air, generating free electrons and positive ion couple. This process continues forming an avalanche phenomenon called corona discharge. Corona noise models have been developed by many researchers and available in the literature [9]-[11]. In this paper we have considered a corona noise model proposed in [10]-[11]. Here corona noise is considered as a random signal characterized equivalently through its autocorrelation function or its power spectrum. Hence the corona noise spectrum is generated by a method that takes into account the generation phenomena of corona currents injected in the conductors and the propagation along the line. This spectrum generates an autoregressive (AR) digital filter whose output is described by the expression- (3) where w k is a sequence of independent zero-mean Gaussian random variables and v l is the set of coefficients modeling the corona noise process. The synthesis of the digital filter essentially calls for the identification of the coefficients v l proposed in [11]. Considering an order-4 of the corona filter proposed, we have taken the following set of values of v l for a power transmission voltage of 225 kv [11]- v 1 = , v 2 =1.052, v 3 = , v 4 =0.217 The Gaussian noise (w k ) used in the digital filter of Eqn (3) are taken as -29.8dB for fair weather and -15.7dB for fair weather for a central frequency of 280kHz and 199kHz respectively proposed in Table 1 in[9]. K 0.7 (single path) 1.0 (multipath) Table I PLC model parameters Channel Values parameter g 1 a 0-2.3x10-3 a x10-7 s/m g g g g a x10-3 a x10-7 s/m IV. NUMERICAL RESULTS Taking N to be the number of paths, the frequency response was obtained for N=1 and N=4 over the entire range of frequencies for which the model is suitable i.e. 500 khz to 20MHz. The frequency responses were obtained for the channel parameters used for simulation of the PLC channel as shown in Table I. We have considered line length for the four paths as shown in Table II- Line type Low voltage[8] Medium voltage High voltage Table II: PLC channel lengths Line distance d 1 =200m, d 2 =222.4m, d 3 =244.8m, d 4 =267.5m d 1 =1 KM, d 2 =2KM, d 3 =3KM, d 4 =4KM d 1 =10KM, d 2 =20KM, d 3 =30KM, d 4 =40KM First we have simulated a low voltage and short distance single path and 4-path PLC in Zimmerman model using equation (2) with the channel length shown in Table II. Short distance PLC ( m) with only a few branches (1 to 4) mostly exhibit typical attenuation values starting from a few decibels at 500 khz going up to db at 20 MHz [8].The frequency response of the simulated short distance single path (N=1) PLC transfer function is shown in Fig.3 while Fig.4 shows the simulation for multi path (N=4) frequency,mhz frequency,mhz Fig.3 Frequency Response of short distance PLC N=1 N=4 118

4 The attenuation of the single path is found to be about 70dB over a frequency range of 20MHz which conforms to the earlier literature. The multi path frequency response exhibits two notches at 5MHz and 15 MHz which also conforms to the earlier literature. This simulation ensures that the model can be used for OFDM transmission and simulation of corona noise in HV transmission. The PLC model explained above were used to transmit the OFDM signal in 16-QAM with the following modulation parameters- No. of Carriers : 64 Coding used Single frame size : Convolutional coding : 96 bits Total no. of Frames : 100 Modulation : 16-QAM No. of Pilots : 4 Cyclic Extension : 25%( 16) We have simulated generation of corona noise in HV PLC in digital IR filter domain using equation (3). The corona noise was simulated for fair weather condition while severe corona noise was achieved in foul weather condition. To obtain the performance of PLC channel there is a need to obtain the BER plot for OFDM under PLC channel. This is done by obtaining the channel response (in the frequency domain) of the PLC channel with different number of multipath (N=1, 2 and 4) and plugging these values into the channel model which is the general OFDM model with AWGN. The BER vs SNR plots obtained for different number of paths for a low voltage and medium voltage power lines are shown in Fig.4 (c) Fig.4 BER for OFDM of PLC Model with N = 1,2 and 4 for low voltage N = 1,2 and 4 for medium voltage with AWGN noise (c) N = 1,2 and 4 for medium voltage with AWGN noise and 50 Hz It is evident from Fig.4 that the BER have pretty differences in the two types of PLC- low voltage and medium voltage, and also between number of paths. The BER shows considerable increase in medium voltage compared to low voltage, in higher path compared to single path and in AWGN noise environment than without noise. Further the PLC in OFDM was simulated to generate corona noise in long distance (i.e. high voltage). Fig.5( a) shows the BER in HV PLC based OFDM with the 4-path length having corona noise. The BER of the corona induced path is quite high for the least SNR condition. Moreover, we have simulated corona noise for two different weather conditions as shown in Fig

5 VI. REFERENCES Fig.5: BER for corona noise in HV PLC Corona noise in path-4 Corona noise in different weather conditions It is observed from Fig.5 that the corona noise in 4- path in foul wearher condition is more severe than corona noise in fair weather condition in path-2. V. CONCLUSION Power line communication can be considered as an attractive alternative to traditional networks as it offers broadband internet access, telephone service and cable television. However, in addition to AWGN, corona noise is found to be a major interfering noise in HV PLC in OFDM. This paper shows that the performance of OFDM can be affected by corona noise with the variation of the branch lengths in the multipath and different weather condition. The BER prominently shows that the corona noise is a major factor compared to AWGN. [1] Tayyar Güzel, Eser Ustünel, Hasan Basri Celebi, Hakan Deliç, senior member, IEEE, and kivanç mihçak, member,ieee,. Noise Modeling and OFDM Receiver Design in Power-line Communication. IEEE Transactions on Power Delivery, Vol. 26, no. 4, October [2] Sakshi Chawla and Abhijeet Kumar, Performance analysis of OFDM under PLC channel, International Journal of Research and Innovation in Computer Engineering, Volume 1, issue 2, issn , (46-52). [3] Opatija Croatia and Mario, Bogdanovi, Power line communication system and modeling based on coded OFDM, MIPRO 2012, Opatija, Croatia. [4] P. Mlynek etal, OFDM model for power line communication, Latest trends on communications and information technology, May 21-25,2012, [5] Riccardo Pighi and Riccardo Raheli,Linear, Predictive Detection for Power Line Communications impaired by Colored Noise, Hindawi Publishing Corporation Eurasip Journal on Advances in Signal Processing Volume 2007, Italy [6] P Burrascano etal, Digital Generator of corona Noise on Power line carrier channel, IEEE Transc. On Power Delivery, Vol 3, No.3, pp , July, 1988 [7] Ferreira, H.C., Grove, H.M., Hooijen, O. Vinck, A.J.. Power line communications: an overview, Proc. of IEEE ISPLC, 1996, pp [8] Manfred Zimmermann and Klaus Dostert, A multipath model for the powerline channel, IEEE Transactions on Communications, vol. 50, no. 4, April 2002 [9] N. Suljanovi c, A. Mujˇci c, M. Zajc, and J. F. Tasiˇc, Computation of high-frequency and time characteristics of corona noise on HV power line, IEEE Transactions on Power Delivery, vol. 20, no. 1, pp , 2005 [10] P. Burrascano, S. Cristina, and M. D Amore, Performance evaluation of digital signal transmission channels on coronating power lines, in Proceedings of IEEE International Symposium on Circuits and Systems (ISCAS 88), vol. 1, pp , Espoo, Finland, June 1988 [11] P. Burrascano, S. Cristina, and M.D Amore, Digital generator of corona noise on power line carrier channels, IEEE Transactions on Power Delivery, vol. 3, no. 3, pp ,