by Virginie Dégardin IEMN-TELICE (Telecommunications, Interférences et Compatibilité Electromagnétique) Mons, 20/11/2014

Transcription

1 by Virginie Dégardin IEMN-TELICE (Telecommunications, Interférences et Compatibilité Electromagnétique) Mons, 20/11/2014

2 OUTLINE 1. PLC context From indoor environment to transportation systems 2. PLC on vehicular network 4. Conclusion 5. On going and future works Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 2

3 1. PLC context and scientific approach Broadband (BB) PLC for high-speed home networking 2 standards : IEEE P1901 and ITU G.hn in the MHz ( > 500 Mbits/s) in-home or access (for the last mile) network NEW CHALLENGES? Narrow band (NB) or BB PLC for smart grid BB PLC for transportation systems? Need of communication (embedded system for safety and entertainment) A.M. Tonello, J. Song, S. Weiss, and F. Yang, "PLC for the Smart Grid: State-of-the-Art and Challenges," Proceedings of Conference on Mobility and Computing (CMC 2012), Guilin, China, May, 2012 Interest : does not require a new communication bus complexity cable weight vehicle weight fuel Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 3

4 2. PLC on vehicular network Context : need of high-speed communication without adding wires and nodes, and existing 12V supply lines in all cars Objectives : Deduce, from intensive measurements, an accurate and stochastic model of vehicular PLC channel to validate the feasibility of in-vehicle Power Line Communication Main difficulties : EMC aspect : Impact of PLC systems on other services radiated and conducted emission limits limited transmission power for PLC systems; PLC channel Noise : stationary and impulsive Transfer function : Multipath environment and time varying Framework & project : PREDIT contract, in cooperation between VALEO, PSA-Peugeot- Citroën, IETR-INSA and IEMN-TELICE, the Pole Sciences et Technologies pour la Sécurité dans les Transports (Science and Technology for Safety in Transportation -ST2). Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 4

5 2. PLC on vehicular network Architecture of the harness decided with car manufacturers Modeling and characterization of the propagation channel Deterministic propagation Model Experimental approach CRIPTE : Succession of interconnected tubes. Each junction is characterized by its [S] matrix Indirect : D1 D2 Direct : D4 D3 Direct : AB and DE Indirect : AC, AE and AF total length = 260 m ; 116 terminal loads Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 5

6 Probability P(X<x) 2. PLC on vehicular network Modeling and characterization of the propagation channel Comparison between experiments and deterministic modeling Indirect path Direct path direct path scenario, Measurement direct path scenario, Theory indirect path scenario, Measurement indirect path scenario, Theory Good agreement between theory and experiment Average insertion loss : 20dB for direct path scenario 35 db for indirect path scenario Insertion gain (db) Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 6

7 2. PLC on vehicular network Amplitude (V) Highlights: Development of impulsive noise measurement system (4 *200 MHz) 6 Characterization of amplitude, duration, 5 frequency, IAT of the pulses => influence of the 4 driving condition (cruising, braking ) 3 Elaboration of impulsive noise model 2 1 Optimum modem impedance Identification of two scenarios -1 (Direct path and Indirect path) -2 CISPR25 : PSD <-80 dbm/hz Theoretical result : 14 Mbit/s available in direct path scenario References : V. Degardin, M. Lienard, P. Degauque, E. Simon and P. Laly, Impulsive Noise Characterization of In-Vehicle Power Line, IEEE Trans. on Electromagn. Compat., vol. 50, n 4, pp , M. Lienard, M.O. Carrion, V. Degardin, P. Degauque, Modeling and analysis of in-vehicle power line communication channels, IEEE Trans. Veh. Technol., vol. 57, no. 2, , M. Olivas Carrion, Communication sur le réseau d énergie électrique d un véhicule : modélisation et analyse du canal de propagation, thèse de Doctorat, Université des Sciences et Technologies de Lille, Juillet 2006 Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" Time (µs) 7 Signal -60 dbm/hz Noise measurement

8 Context : The More (or even the All) Electric Aircraft Replacement of hydraulic and pneumatic energy sources by electrical ones : Higher electrical power Changes in the voltage levels Increase in the wires mass Increase data communication exchanges on A/C which leads to: Increase in the number of wires Increase system complexity Framework & project : TAUPE Project with European Community's Seventh Framework Programme (FP7/ ) under Grant agreement number ; DPCA contract ISS Power and Control in collaboration with Airbus and Safran Engineering Services ; PhD. Thesis FIAC in collaboration with Sagem Défense, Safran Engineering Services and IETR from INSA Rennes International Campus on Safety and Intermodality in transportation systems (CISIT). Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 8

9 3 possible applications in a commercial aircraft DO160 EMC constraints => CM current limit I CM a complex tree-shaped architecture of the harness Signal crosstalk between adjacent systems in bundles Cabin Lighting System Flight control system (Spoiler, aileron) Landing gear Between a motor and an PWM inverter on a 3 phase AC Power cable PWM impulsive noise PP and Point-to-Multipoint topology HVDC network Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 9

10 A. PLC on Cabin Lighting System (CLS) Typical architecture of a CLS : representative of tree-shaped of many aircraft harness configurations SPDB IBU IBU CIDS DEU Typically: 1 power line for 8 24 IBUs Power: Secondary Power Distribution Box (SPDB) Illumination Ballast Units (IBUs). Typically, 1 power line feeds 8 to 24 IBUs - Each power line runs within a cable bundle Data (control command): Remote control (cabin crew) through the Cabin Interconnection Data System (CIDS) Decoder Encoder Unit (DEU) IBU idea : Dedicated transmission line can be removed by using PLC Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 10

11 A. PLC on Cabin Lighting System (CLS) STEP 1 : Laboratory Test bench Simulation Architecture Statistical channel properties for the various links based on the theoretical modeling of the propagation on the harness Modeling the PLC link expected throughput or BER SPDB : Distribution Box IBU : Illumination Unit Tree network architecture Maximum length SPDB-IBU: 43 m VT : multiplying connector PLC lines inside a cable bundle Number of wires in the bundle: 2 to 30 Total length of the wires: 706 m Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 11

12 STEP 1 : Laboratory Test bench Simulation Architecture Statistical channel properties Modeling the PLC link expected throughput or BER A. PLC on Cabin Lighting System (CLS) Important local fading due to multipath + coupling to the other wires PLC frequency band Statistical aspects on: Path loss Coherence bandwidth (band in which H(f) does not vary appreciably ) Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 12

13 Probability P(X<x) STEP 1 : Laboratory Test bench Simulation Architecture Statistical channel properties Modeling the PLC link expected throughput or BER A. PLC on Cabin Lighting System (CLS) 10 0 Comparison statistical behavior of the channel Predicted results vs Measurements => Good agreement 10-1 Exp MHz Theory MHz Input for the simulation of the data transmission Insertion gain(db) Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 13

14 STEP 1 : Laboratory Test bench Simulation Architecture Statistical channel properties Modeling the PLC link expected throughput or BER A. PLC on Cabin Lighting System (CLS) HPAV specifications subcarriers on [1.8-30] MHz (only 917 sc with spectrum mask) - ½ Turbo convolutional code and channel interleaving - compatible with the CLS channel characteristics What s about injection and noise power? Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 14

15 STEP 1 : Laboratory Test bench Simulation A. PLC on Cabin Lighting System (CLS) Amplitude of noise and signal current on the PLC line? Noise due to systems directly connected to the PLC line is assumed to be well filtered Noise is due to the coupling of the disturbing currents flowing on the other wires Norme DO160 => I CM 20 dbµa/khz Modem Tx I DM signal I CM signal I CM dist Bruit blanc I DM noise Modem Rx signal Amplitude? I DM signal = I CM signal + CF DM-CM Noise Amplitude? I DM noise = I CM dist + CF I dist I DM noise CF = I CM_signal / I DM_signal CT = I dm_noise / I cm_dist Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 15

16 STEP 2 : Lab. Test bench Experiments A. PLC on Cabin Lighting System Characteristics of the Univ. Lille modems Design and development of versatile modems (based on FPGAs) to be able to change: Signal processing, modulation scheme, etc. In the following HPAV standards Principle of the experiments (DM) White noise generator Modem I CM noise 1: Adjust the power of the modem I CM signal < 20 dbµa in the whole bandwidth 2: Adjust the power of the noise generator I CM noise < 20 dbµa in the whole bandwidth I DM signal Current probe I CM signal 3: Send 100 PHY Block (520 bytes) of OFDM frames, Store the Rx frames BER Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 16

17 STEP 2 : Lab. Test bench Experiments A. PLC on Cabin Lighting System Throughput : Maximum bit rate to guarantee a BER <= 10-3 maximum bit rate : 98.5 Mbits/s. Throughput : Maximum bit rate to guarantee a BER <= 10-3 Chosen values for CF and CT for a percentile to 80% IBU 1 IBU 2 4 IBU 5 7 IBU 8 11 IBU Distance (m) Number of VT Throughput Exp. (Mbits/s) Throughput Th. (Mbits/s) V. Degardin et al., "Investigation on power line communication in aircrafts", IET Commun., vol. 8, no. 10, , V. Degardin et al., Theoretical approach to the feasibility of power line communication in aircrafts, IEEE Trans. vehicular tech, March Report of the weekly magazine "Air &Cosmos on Taupe project, Vers une architecture électrique de l avion mieux optimisée, March Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 17

18 3. Performance on the avionic network B. PLC on a 3 phase AC Power cable between a motor and a PWM inverter Objectives: Feasibility of a PLC communication on a 3 phase AC power cable Decreasing the number of wires to be used and simplifying the architecture Highlights: Measurements of PWM impulsive noise => influence of the cable length, motor speed, inverter voltage, and equipment. Measurement of insertion gain Optimize the PLC link with noise processing at 20 Mbit/s, a CSD > 56 dbµa/khz is required to obtain BER < 10-4 Study and optimize the MIMO PLC link (SFBC code) bit rate of 5 Mbits/s : Gain of 7 to 10 db References: V. Degardin, K. Kilani, L. Kone, M. Lienard, P. Degauque, "Feasibility of a high bit rate power line communication between an inverter and a motor", IEEE Trans. Ind. Electron., vol. 61, no. 9, , K. Kilani, V. Degardin, P. Laly, M. Lienard, Transmission on aircraft power line between an inverter and a motor : impulsive noise characterization, IEEE International Symposium on Power Line Communications and its Applications, ISPLC 2011, Udine, Italy, April 3-6, pp , Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 18

19 4. Conclusion Analysis of the feasibility of the PLC communication in vehicular and avionic environments The scientific approach : Characterizing and modeling the networks (Insertion gain and noise) Impulsive noise measurement system (4 inputs 200 MHz) vehicular noise models deterministic model of avionic and vehicular representative harnesses validated by measurements Modeling and optimizing the PLC link to predict BER and throughputs PLC simulation tools based on OPERA and HPAV specifications Validating the theoretical results with configurable modems versatile modems Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 19

20 5. Future works Applicability of PLC to aircraft: Need to fulfill regulatory requirements for reliability, susceptibility and robustness EMC susceptibility standards After an interruption of the link, PLC communication must be reestablished t<1ms Low latency => HPAV specifications not adequate. Simplify the transmission scheme while guaranteeing the prescribed maximum value of the BER. Optimize channel coding and modulation New application : default detection, cable monitoring, arc tracking avoidance Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 20

21 Thank You. Journée d'études "Will we be Smart by 2020?" - "Chaire ORES Smart Grids - Smart Metering" 21