Propagation of electromagnetic waves

Transcription

1 Propagation of electromagnetic waves The propagation of waves between a transmitting and a receiving antenna is influenced by many different phenomena. Transmission paths can vary from simple line-of-sight to severely obstructed (buildings, mountains, foliage, etc.). The propagation channel places fundamental limitations on the performance of wireless communication systems. Its modeling is a very complex task. The analysis is typically based on a combination of simplified (statistical) physical models and empirical knowledge. Contents 1. Introduction to radio communication Managing the frequency spectrum 2. The radio communication channel Path loss, fading 3. The different scales of communication systems From mega-cell to pico-cell 4. Propagation models Reflection, diffraction, scattering 5. Propagation media Ground waves, ionospheric and tropospheric effects 1

2 1. Introduction to radio communication Radio or radio communication means any transmission, emission, or reception of signs, signals, writing, images, sounds or intelligence of any nature by means of electromagnetic waves of frequencies lower than three hundred giga cycles per second (300 GHz) propagated in space without artificial guide. Examples of today s common radio communication: Radio broadcasting TV broadcasting Satellite communication (between earth and satellites) Mobile telephony Multimedia communication & mobile internet Safety communication networks Radio determination (GPS, marine & air navigation) Other (amateur radio, astronomy, etc.) and many others 2

3 History of radio communication Physics of radio wave propagation Radio electronics Digital communications Multi-user communication 1873 Maxwell predicts the existence of electromagnetic waves 1888 Hertz demonstrates radio waves 1897 Marconi demonstrates mobile wireless communication to ships 1920 First regular licensed radio broadcast (Pittsburgh ) 1928 US police (Detroit) first use mobile communications 1932 BBC's television public service begins 1945 Arthur C. Clarke proposes geostationary communication satellites 1948 C. Shannon publishes A mathematical theory of communication'' 1957 Soviet Union launches Sputnik 1 communication satellite 1969 Bell Laboratories in the US invent the cellular concept 1979 First generation cellular mobile systems (analog) 1982 Inmarstat satellite mobile communication system 1992 Second generation cellular mobile systems (digital) 1998 Iridium & Globalstar global satellite systems 2002 Third-generation cellular mobile systems Marconi wireless telegraph 3G handset 3

4 Mobile telephony evolution: First generation (1G) 1979 NTT cellular system (Japan) 1981 NMT cellular system(scandinavia) 1983 AMPS cellular frequencies allocated (US) 1985 TACS (Europe) 1988 JTACS cellular system (Japan) 1991 USDC (US) Analog Second generation (2G) 1991 GSM cellular system deployed (Europe) 1991 Digital AMPS, IS54 (US) Generation 2.5 (2.5G) 1993 IS95 (CDMA) (US) 1994 DCS-1800 (worldwide except USA) 1994 DCS-1900 (USA) Digital Third generation (3G) 2002 IMT-2002 (UMTS) World GSM coverage map 2.5G: General packet radio service Data transfer <~144 kbps. 3G: Towards global coverage, personal communication systems Data transfer >384kbps, up to 2 Mbps. 4

5 Fourth generation (4G) Glossary: LTE-Advanced: Long Term Evolution (4G) LTE: Long Term Evolution (3G) HSPA: High Speed Packet Access W-CDMA: Wideband Code Division Multiple Access (IMT-2000 CDMA Direct Spread) EDGE: Enhanced Data rates for GSM Evolution GPRS: General Packet Radio Service GSM: Global System for Mobile Communications IMT-Advanced: International Mobile Telecommunications (4G) WRC: World Radiocommunication Conference J.-P. Bienaime, From HSPA to LTE and Beyond: Mobile Broadband Evolution, Supplement to Microwave Journal, Nov

6 Fifth generation (5G) Millimeter waves at 28 GHz, 38 GHz, GHz, etc 28 GHz Measured path loss values relative to 5 m free space path loss for 28 GHz outdoor cellular channels. Maximum coverage distance with 119 db max path loss dynamic range without antenna gains and 10 db SNR, as a function of path loss exponent n. Measured path loss for 38 GHz outdoor cellular channels. T.S Rappaport et al, Millimeter Wave Mobile Communications for 5G Cellular: It Will Work! IEEE Access, Vol. 1, pp , May

7 5G Frequencies Bands ITU Nov 2015: , , , , , , 66 76, GHz FCC July 2016: , , , GHz (with smaller sub bands) More research is anticipated at 73 GHz as there is more than 2 GHz contiguous bandwidth available. Though the future of 5G is not yet clear, mmwave will surely be one of the technologies used to define it. While questions remain around the global spectrum allocation, the U.S. is moving directly and decisively towards 28, 37 and 39 GHz. J. Kimery, 5G opens up mmave spectrum: Which frequencies will be adopted, Supplement to Microwave Journal, Nov

8 International Telecommunication Union (ITU) Agency part of the United Nations 193 member countries and over 700 private-sector entities and academic institutions. Headquartered in Geneva, Switzerland, 12 regional/area offices around the world. ITU-R = Radio communication bureau: - ensures rational, equitable, efficient and economical use of the radio frequency spectrum and the GEO satellites orbit. - coordinates efforts to reduce interference between radio stations of neighbouring countries. ITU-T = Telecommunications bureau: - deals with the public telephone network, transmission via cables, optical fibres,... - coordinates developments (standards) The International Telecommunication Union (ITU) manages the frequency spectrum through frequency assignment, standardization, coordination and planning of international telecommunication services, system compatibility issues, research. 8

9 The Radio Spectrum a limited resource Name Frequency range Wavelength Applications Propagation ELF Hz km Navigation, long distance communication with ships VLF 3 30 khz km Navigation, long distance communication LF khz 1 10 km Navigation, long distance communication with ships MF khz m AM broadcasting, radio navigation HF 3 30 MHz m Radio broadcasting, fixed pointto-point (around the world) VHF MHz 1 10 m Radio & TV broadcasting, mobile services UHF MHz cm Cellular telephony (GSM, NMT, AMPS), fixed point-to-point, satellite, radar Wave tube between earth surface and the ionosphere Ground propagation Stable Ground propagation Stable Ground-wave, sky-wave propagation. Fading Large perturbation, reflection in ionosphere Diffraction Shadowing by mountains and buildings SHF (Microwaves) 3 30 GHz 1 10 cm Broadband indoor systems, microwave links, satellite communications Attenuation due to rain, snow and fog EHF GHz 1 10 mm LOS communication (short distance or satellite) Attenuation due to rain, snow and gases 9

10 Managing the Radio Spectrum The frequency spectrum is common to all radio systems and becomes an increasingly scarce resource. Radio propagation is not confined to geo-political boundaries. EM wave propagation has different characteristics in different parts of the radio spectrum. It is necessary to - assign appropriate frequency bands to each application - consider propagation phenomena in the design of radio communication links - derive appropriate mathematical propagation models - derive appropriate mathematical channel models 10

11 2. The radio communication channel A radio communication system consists of a transmitter, a channel, and a receiver Source of information Transmitter Channel Receiver User of information The purpose of the transmitting antenna is to efficiently transform the electrical signal into radiation energy. The purpose of the receiving antenna is to efficiently accept the radiated energy and convert it to an electrical signal for processing by the receiver. 11

12 Radio Communication System Design Analog Communication System Digital Communication System 12

13 The communication Channel Noise source Channel modifies transmitted information Receiver designed to recover information from channel Source Transmitter Receiver Destination The channel Noise types affecting the system SNR X + Multiplicative noise Additive noise Multiplicative effects Antenna directionality Reflection (smooth surfaces) Absorption (walls, trees, atmosphere) Scattering (rough surfaces) Diffraction (edges of buildings and hills) Refraction (atmosphere) Additive noise Thermal and shot noise in receiver Atmospheric & cosmic noise Interference (intentional & otherwise) 13

14 Additive noise A T c Reference plane for T sys T A T m Receiver T r Feeder L f, T f T A T (, ) G(, )sin d d B G(, )sin d d T T 1 10 T 10 A/10 A/10 B m c Lf /10 Lf /10 sys r f 1 10 A 10 T T T T T A : Antenna noise temperature (see Ch. 2, p. 44) T B (q,f): Brightness temperature T m : Physical temperature of medium (typically 280 K for clouds and 260 K for rain) T f : Physical Temperature of feeder T sys : Total system noise temperature A(q,f): Total absorption due to rain and gaseous attenuation (db) L f : Feeder loss (db) 14

15 Multiplicative noise Three scales of multiplicative noise X X X X X + Transmit antenna Path loss Shadowing Fast fading Receive antenna Additive noise Fading processes Path loss Shadowing (slow fading) Fast fading (multipath fading) Time varying processes appear between antennas 15

16 Overall signal strength (db) Three scales of mobile signal variations Total signal Distance transmitter-receiver (m) Path loss (-db) Shadowing (db) Fast fading (db) Path loss Shadowing Fast fading Distance transmitter-receiver (m) Dependent on distance Transmitter- Receiver Variation scale: In the order of 100 m (depends on large obstacles) Variation scale: In the orderl/2 (Multipath interferences) 16

17 Path Loss Antenna gain G T Antenna gain G R Transmitter Transmit Power P T P TA Feeder loss L T r Path loss L P RA Received power P R Feeder loss L R Receiver Link Budget: P R PG G LLL T T R = T T R T R Path Loss (Line of Sight - LOS) L PG G = PLL R T R Remarks: Propagation loss changes if shape of antenna radiation patterns (i.e. G T, G R ) changes. Path loss does not take multipath propagation into account. Maximum range r max of system is reached when received power drops below receiver sensitivity. Value of L for this level is maximum acceptable path loss PG T TGR L = r max P L L max Rmin T R 17

18 Path loss as a function of transmission distance (free-space propagation) Friis Transmission Formula: Pr P t 2 æ l ö = ç GG T è 4pr ø R ( L = L = 1) T R A = 4pr 2 With increasing distance r between T-R, the received power decreases. Since power is spread over surface area of sphere, which increases with r 2, the available power at the receiver is proportional to 1/r 2 in the far-field. r Friis formula can be rearranged into the same form as the propagation loss formula. Assuming feeder losses to be zero or defining the propagation loss in free space only, we can write the path loss L F as L F 2 2 PG G æ t T R 4prö æ4prf ö = = P = ç l c è ø çè ø r 18

19 Free-space path loss in db: L F[dB] = 10 log L F Reference: L = (i.e. no path loss) F,0 1 Expressing L F in db, with frequency f in MHz and distance r in kilometers L = logr + 20logf F [db] [km] [MHz] Free space loss increases by 6 db for each doubling in frequency or distance (or 20 db per decade). Example: A mobile user transmits and receives on a 1.8 GHz channel. He is 1 km away from the nearest cell tower. Determine the path loss. Solution: L F rf 4 rf log 20log 20log 97.5dB 8 c c 310 alternatively: L log(1) 20log(1800) 97.5dB F How realistic is this number? 19

20 Fading: Excess loss Unfortunately, the free-space path loss calculations are too optimistic. Additional loss factors will contribute in practice ( Sect.3,4&5).Thefreespace value of path loss is only used as basic reference. The loss experienced in excess of this value(indb) is referredtoas excess loss L ex L = LF + Lex Fading: physical situation Relative delays >> symbol duration Wideband Fading Each beam is composed of several individual waves 6 db Example: Two-path channel Transfer function of received signal LOS R 0.1 ms delay (30 m) (will be treated later in Sect. 4) 1 ms delay (300 m) 20

21 3. The different scales of communication systems Global Rural & suburban Urban In-building Pico-Cell Typical scale Traffic density Mega-cell 1000 km Macro-Cell Micro-Cell Pico-Cell 10 km 1 km 100 m 21

22 a) Megacells Satellite communication systems Constellations of satellites provide global coverage Frequency range: 2-40 GHz Propagation issues: High elevation angles, so only environmental features close to the mobile user contribute significantly Atmospheric effects are significant in systems operated at SHF and EHF Propagation prediction techniques must combine shadowing and fast (multipath) fading since both occur on similar scales and cannot easily be separated 22

23 Orbit types: LEO: Low earth orbit ( km) Several dozens of satellites necessary for full earth coverage Fast-moving satellites (orbit period ~100 minutes) MEO: Medium earth orbit ( km) Fewer satellite necessary (8-12) Slower moving satellites (orbit period ~ 6 hours) GEO: Geostationary orbit (36000 km) 4 satellites (service areas duplicated) Looks stationary from the ground satellite constellation GEO km 23

24 Examples of systems: Inmarsat Established 1979 for ship communication 98% coverage GEO satellites Iridium* Established 1998, stopped in 2000 Relaunched in LEO satellites (778 km) 100% coverage Phone satellite at 1.6 GHz Satellite-gateway at 19.5 GHz Satellite-satellite at 23.3 GHz Globalstar Commercial service since LEO satellites (1400 km) 80% coverage Gallileo Launched in of 30 MEO satellites (23,222 km) by Dec % coverage by 2020 (1m accuracy) * The name Iridium was established based on 77 satellites (77 is the atomic number for Iridium). With the current 66 satellites, the system should be renamed Dysprosium. 24

25 Free-Space path loss for satellite communication: Free-space loss large but constant for GEO systems Free-space loss time-varying for LEO and MEO systems elevation 90 user elevation km Iridium: 778 km (LEO) GHz 7.5 minutes pass 778 km These values would be encountered over a period of around 7.5 minutes 25

26 Doppler shift 0 -Df Doppler +DfDoppler Doppler shift results from motion of the satellite relative to location of mobile Can be compensated for by retuning transmitter or receiver Example: Iridium ±37 khz Doppler shift Local sources of satellite transmission impairment: Trees Building Terrain These interact with wave propagation via the following mechanisms: Diffraction Scattering Reflection Multipath Signal amplitude (db) Example of channel variation Measured in suburban area Note the rapid, frequent transitions from LOS to Non-LOS conditions Time (s) (mobile speed = 30 km/h) 26

27 Local Shadowing Effects: Roadside buildings produce significant attenuation when more than 0.6 times the first Fresnel zone is blocked (Diffraction, see Sect. 4) Satellite 0.6 times First Fresnel zone (see Sect. 4) Satellite Shadowing by trees produces attenuation of around 1.7 db/m at 900 MHz Satellite Reflection and rough surface scattering produce multipath and hence fast fading. The path length differences are small, so the wideband effects are modest. Multiple scattering is typically weak (e.g. satellite A B receiver). B A 27

28 b) Macrocells Terrestrial communication in rural and suburban environment Frequency range: GHz Antenna typically considerably higher (typically m) than surrounding buildings and local terrain Propagation issues: Designers are usually not interested in particular locations but in the extent of the coverage area ( non-overlapping hexagonal cells). Empirical models gained from intensive measurements are available for particular conditions. Physical models such as diffraction and reflection can be coupled to these empirical models to take into account hills, large buildings, etc. 28

29 Empirical models for macrocells: Loss models based on extensive sets of measurements with parameters derived from - the environment - the frequency - the antenna height Simplest form: Power law model r L[dB] = 10nlog + Lref r ref Received signal level (dbm) Distance from base station (m) n: path loss exponent (typically ~4) Other models: CCIR, Clutter factor, Okumura-Hata, COST-Hata, Walfisch Ikegami Environments are classified subjectively in categories (urban, rural, hilly, wooded, ) 29

30 Physical models for macrocells: Complementary to empirical models for situations where there is no LOS path Permit to improve accuracy of empirical models by taking into account shadowing (slow fading) Rooftop diffraction Increased complexity of prediction Diffraction models (simple and multiple diffraction) Ikegami model Flat edge model Other models covering different scenarios: Okumura Model, Hata Model, COST231-Hata model, COST231-Walfish-Ikegami Model, Okumura-Hata Model, Walsch-Bertoni Model, Waveguide Model, 30

31 c) Microcells Communication in high-density urban environment Smaller cells increase the capacity (fewer subscribers per cell) Frequency range: GHz Base station antenna mounted only 3-6 m above ground or on the side of a building Propagation issues: Accurate path loss models are needed for correct placement and optimization of microcells (unorthodox shapes of cells possible). Planning exploits the environment to minimize interference between cells so that frequencies can be reused over short distances. Dominant propagation mechanisms: free-space plus multiple reflection and scattering, diffraction around edges. 31

32 Multipath Propagation in street canyons: Extend two-ray model to four, six or more rays 2 rays: Open environment 4 rays: Paths up to 1 reflection 6 rays: Paths up to 2 reflections 32

33 Two ray model results Four ray model results Path Loss (-db) Path Loss (-db) Vertical polarization Horizontal polarization -90 Vertical polarization Horizontal polarization Distance (m) Distance (m) Multipath causes fast fading 33

34 Non-Line-of-sight models: Reflection and scattering from walls and the ground Diffraction over building rooftops Diffraction around vertical building edges Rooftop diffraction as an interference mechanism B D A Various propagation mechanisms for NLOS models: A: Diffraction B: Reflection C: Multiple reflection D: Rooftop diffraction C Diamond shape coverage area 34

35 d) Picocells Communication inside buildings Cellular telephony for high-traffic areas: Shopping centers, airports, office buildings High-data rate applications (WLAN) Frequency range: /94 GHz Propagation issues: Effects even more dependent on geometry than in microcell. Simple path loss models are inadequate. Transmission through walls and floors, reflection and multipath. In-building propagation is also relevant for macro/microcell penetration. Empirical and physical models in a complex 3D environment. 35

36 Empirical models of propagation within buildings Wall and floor factor models Loss at r = 1 m # of floors Attenuation per floor (db) Path loss in db: L = L log r + nfaf + nwaw Straight line distance between T-R (m) # of walls Attenuation per wall (db) 1/r 2 a w a f a w 36

37 ITU-R model: A similar model is suggested by ITU-R model [ITU, 1238], except that only floor loss is accounted for explicitly, and the loss between points on the same floor is included by changing the path loss exponent n. Carrier frequency (MHz) Floor penetration loss (db) Path loss in db: L = 20 log f + 10nlog r + L ( n ) -28 T c f f Path loss exponent Distance in m # of floors 1/ r n a f 37

38 Empirical values for the ITU-R model are dependent on environment and frequency Path loss exponent n Floor penetration factors L f (n f ) Frequency (GHz) Environment Frequency (GHz) Environment Residential Office Commercial Residential Office Commercial (1 floor) - 19 (2 floors) 24 (3 floors) n f (n f -1) 6 +3 (n f -1) The penetration losses may be overestimated for a large number of floors since other paths might become dominant (e.g. diffraction through windows, reflection on neighbour building, etc.). Example: A 900 MHz signal propagates a distance of 30 m through two floors in an office building. Find the path loss. Solution: L = 20 log f + 10nlog r + L ( n ) -28 T c f f = ( 20 log log )dB L» 99dB T 38

39 Physical models of indoor propagation: Deterministic approaches for specific site prediction: - Ray tracing - Geometrical theory of diffraction -FDTD Indoor models depend highly on the composition and position of walls, floors, soft partitions, furniture and people. Model factors typically include: - T-R separation distance - Floor attenuation factor - Wall attenuation factor - Soft partition attenuation factor Use floorplan drawings to make per-floor predictions. Issues: - Highly geometry dependent - Quality of data? - Computationally expensive 39