RTLS communications

RTLS communications

Articleissues
essay=July 2008
copyedit=May 2008
refimprove=August 2008

RTLS communications is a special breed of wireless communicating with the aim to support locating wireless nodes not only topologically, but merely topographically outdoors or indoors.

This page on communicating for real-time locating systems refers to RTLS according to ISO/IEC 247340. Please get informed also with the ranging section of the RTLS page.

The operation of the wireless functions requires well defined approaches to usability. The basis for success is a sound concept for metering. Currently the approaches in IEEE 802.15.4a show various options. Additionally the concepts of mathematical calculation and for operational networking with or without administering are crucial for performance.

Ad hoc networking versus conventional administration with RTLS

The nodes in a wireless network are the vertices (pl. of vertex) in a mesh. The edges of this mesh are the distances between the nodes to be measured. To enable adequate wireless communication between these nodes, the properties of these nodes must be adjusted to become members in the network and then published to support recognition though the other nodes. However, not all of the nodes will see each other node and not all nodes must know all other nodes. But to disseminate information at least about direct neighbors and some of other next nodes in the vicinity, there must be a concept of administering the nodes and their membership data.

Compared to traditional administration, ad hoc networks work out this task by some concepts of autonomous pushing and pulling with subsequent memorizing to keep the network largely free of this administrative load. The degree of autonomy may be regarded as a feature of qualification of the networking concept.

Proactive and reactive exploration when ad hoc networking an RTLS

In mobile ad hoc networking (IETF MANET), the main categories of routing are proactive and reactive. Grouping vertices in an RTLS shall take these classes into account:

* There are residing vertices with known positions in most RTLS set ups. For these exploration will be very effective being performed advance to any locating of moving nodes. Such is proactive routing.

* There are temporarily moving nodes that might contribute to networking the RTLS set up. For theses nodes there should be a group membership determined by motion.

* There are simply moving nodes knowing none about there own location. For such condition, reactive routing is the best choice.

Coexistence criteria for RTLS platforms

As with any communications system the coexistence of various wireless systems is affected by any new participant in the same band. Many systems the same licensed ISM bands and therefore interfere with each other. There was an understanding, that deterministic continuous wave communications are more harmful to each other. That led to the quasi-stochastic pulse coded wide band communications, that macroscopically just add some noise in the used bands thus pretending better tolerance and coexistence.

With increase of such population the noise level rises also significantly, hence producing a threat of ubiquitous stressy burden for yet existing systems. New strategies will enable sound balancing of stress to bearable levels. Such strategies may apply
* neglecting any problem (UWB solutions in the 2.45 GHz range)
* conformance with existing frequency and channel schemes (e.g. Nanotron, 2.45 GHz WLAN scheme)
* using fewer populated frequency bands (e.g. SYMEO, 5.8 GHz)
* escaping to newly released frequency bands (e.g. UBISENSE, 5.9 GHz, released in UK)
* escaping to a wider range of UWB bandwidth (3.1 GHz to 7.0 GHz in US)Increasing usage of such approaches will result in new findings about real interference.

Challenges

Some inevitable effects have a strong impact on performance of RTLS communications. There are first hand parasitic reflections and second hand shortage of resources. Designing for both qualitites defines qualifications of the systems and its performance.

Antenna characteristics

Antennae may be designed for directed emission. non directional antennae have a specific antenna characteristic: Pins have a dominantly spheric footprint, rods have a dominantly cylindric footprint, but mostly the geometric shape does not simply correlate with the footprint.

The footptint may be depicted in an antenna disgram, that shows the main beacons and some side lobes. So weird antenna patterns are required. These serve the emission with more or less isotropic antenna diagrams. However, the major effect comes with reflecting surfaces or deflecting edges. These flections generate signal paths beyond line of sight and may vary in intensity with the conditions of radiation in the atmosphere as well as with the conditions of reflection at surfaces.

Reducing the antenna diagram to one single main lobe focuses the transmitted energy to one direction. However, reflection from a passive target has always more or less speheric characteristics, whereas the active reponse from active responders follows the antenna characteristic of that responder. A dedicated antenna characteristics shall improve the communications between the involved wireless nodes that perform laterations.

Multipath response

Basic requirement for reception and response is line of sight or some deviated path between transmitter ans responder in both directions. But only direct line of sight serves for lateration. All other kinked paths may contribute to network cohesion, but not for proper lateration. However, when there is no match with line of sight, emission might not meet the receiver antenna at all.

As always direct and kinked responses arrive in short temporal distance, the directly received signals must be properly segregated from secondary reception. This appears easy, as direct path must be faster than secondary path, but energy in direct reception may be much lower than with any of the other signals. Besides, stochastic effects in data processing may even provide virtually shorter travel times.

One easy solution to overcome detached kinked responses is repetition of measurement, thus stressing the communications time budget and this may not apply properly with faster moving targets. Another easy solution to overcome detached kinked responses is over-determining the respective system of equations. More sophisticated solutions combine motion measures to perpetuate a track, however this does not work with resident wireless nodes. Multi-frequency approaches may support as well better reesolution to fractions of wave cycles as well as for eliminating secondary path responses, but this approach puts additional burden on designing the resolution for the measurement unit. Further approaches may take physical qualities into account, again challenging the measurement circuit designs.

Any issues on modulation concepts are wrong that tell there is no multipath response, caused by the misunderstanding of relative immunity to multipath fading as similar to non existence of multipath reflection. Multipath is inevitable. But signal processing may be supported through reception an various frequencies, allowing for phased computing and elimination of multipath signals due to stochastic conditions.

Anti-collision concepts

Despite the fact that UWB and CSS modulation concepts operate transmission with short pulses, the group of nodesoperating under this paradigm still require quiet times to transmit and receive between pairs of vertices. At least the two of each pair simply do not operate in full duplex mode without colliding. This defines the demand for anti- collision and some higher sophisticated coordinated resource allocation:
* The easiest but most expensive approach is the use of separate channels in a crowded ambience
* The other approach is the usage of statically interleaving pauses, which is contra dictionary to real time
* The skilled approach is some proven frequency resource sharing and anti-collision concept as CDMA, CSMA, CSMA/CD, CSMA/CARP, etc. (read e.g. [ [https://www.auto.tuwien.ac.at/LVA/HGA/stp/hartl.pdf Low Rate WPAN ZigBee] ] ).
* specially tailored sequencing, coordination and code sharing schemes as TDMA, OFDM, OFDMA, QFDM, QPSK and others, chosen coherent with the modulation concept (read e.g. [http://www1.alcatel-lucent.com/com/en/appcontent/apl/S0506-Wireless_Mobile-EN_tcm172-262261635.pdf] ).However, anti collision concepts as Aloha, slotted Aloha and CSMA/CA are not supportive for real time requirements, as known with older RFID systems designs, because these concepts do not determine access to slots.If there is no concept applied, the system approach simply will fail, latest under heavy load in dense populations.

The interested party shall not believe in demos with few resident or single slowly moving nodes.

Methods of time synchronization or illumination

Basically for time measurement between two points either the clocks at these points shall be synchronous or the difference from synchronism must be known.

Regarding TOA and TDOA, synchronization of time is conditionally required. This may be obtained in different ways
* explicitly with synchronized clocks in the metering locations
* with measuring supported from third reference pointsFor details please refer to respective system design manuals or patent applications.

Hence, when pairs of radio signals for angulation must be precisely synchronized between distant units, the better results may be obtained with lesser precise lateration compared to angulation.

Asynchronous metering

To overcome the synchronization problem, already GPS applies a multi frequency transmission and as well makes use of precisely synchronized clocks in conjunction with relativistic correction calculus.

As for terrestrial and cooperative locating systems the distances are very short compared to satellite metering, such sophistication is not required in hardware. With sufficient accuracy, the metering may be based on the concepts of non synchronised metering with just stable generators for transmitted frequencies.

The easier approach is:
* either to reflect a signal from the receiver to the first transmitter and save the synchronizing at all - or
* to measure angles with two close positioned receivers on the very same basis - or
* to measure phase differences for at least two frequencies traveling the very same path.

Hence the synchronization problem is solved by different method to implicitly compensate for asynchronism
* with two propagation concepts and highly different travel speed measured with different equipment
* with pulsed multi-frequency propagation and metering the pulse phase difference in one channel
* with phased single-frequency propagation and metering the phase time difference between two channelsThe latter two solutions are state of the art and basis for commonly available products.

The first concept is known from roughly determining the distance to a thunderstorm. Applications are known not for better precision of distance metering, but for better segregation of residence in neighbored confinements, as with infrared transponders combined with radio transponders or Ultrasonic Identification transponders in such combination.

Multi-frequency metering

A completely different calculus may be obtained from application of two or more different frequencies transmitted from one position. The travel time varies with frequency and the distance from transmitter to receiver again is a linear function of time difference of arrival, after the two signal were transmitted at the same time. For this method the synchronization task is inherently performed with the coincidence of sending the two signals and some reliable stability of the used frequencies, especially the frequency ratios of the two carriers.This concept may be extended to phase metering. The correlation of two frequencies is not only frequency sensitive, but also phase sensitive.

Multi-structure metering

More complex designs may be used to escape existing patents claims by means of new patent claims. This may appear interesting, but normally comes with additional burdens. The idea e.g. to provide general illumination first requires direct line of sight with this common transmitter for any ranging receiver.Such is e.g. shown in [http://www.freepatentsonline.com/7026992.html). This approach my be not applicable in environments, where obstacles hinder line-of-sight.

Centralized triggering and synchronization

Systems layouts may stipulate coordinated unilateral measurement with several anchor node receivers and hence cover one portion of time synchronizing. These systems provide an easy approach to time synchronisms with the clock rates of the wireless nodes involved. Initial triggers may be sent out from one central trigger node or just from the node to be located to the anchor nodes with known location.

The other necessary portion of time synchronizing applies with the uplink to the anchor node receivers. These will receive the lateration signal independently, thus require synchronization of clock rates via a second link. This is normally provided with wire or fibre links, that send the synchronization information in an downlink. Respective systems concepts are patented.

Multilateral asynchronism compensation

Another systems layout compensates for clock rate asynchronisms with bidirectional, symmetric measurement. This approach stabilizes the results against stochastic oscillations and typical variations and and thermal drifts of the clocks involved. This systems concept requires no hierarchy of wireless nodes, but also uses a set of anchor nodes with known location. Respective systems concepts are patented.

Central clock synchronization

This approach is common with satellite navogation systems. However none of the existing and coming satellite systems is subordinated to ISO 19762-5 RTLS definitions. With global navigation satellite systems (GNSS), a widely common illumination of the global surface is achieved. For secure and reliable operation, not only the clock rates are synchronized, but the absolute time information used is coordinated with all satellites of each system (e.g. NAVSTAR, GLONASS, Galileo). This centralistic solution is required to compensate for relativistic effects occurring with satellites, flying at high speed at large distances to the observer. Other methods of communicating lateration signals based just on stabilized clock rates would not provide sufficient accuracy.

Non-ranging approaches

Ranging as the basic principle of metering is mainly reserved for the term of travel time measurement (time of arrival, TOA). This may be extended to differential time measurement (time difference of arrival, TDOA), serving for determination of angles. But, angulation must include lateration, as similar triangles with equal angles may have various size with different length of edges. Similarly, any phase determination must be accompanied by travel time measurement, as phases only repeat per each cycle.

Other approaches than using that equivalence of wavelength and travel time take other but the strong relation of time and light speed in to account, are basically better guessing. The reasons are the strong impact of multiple reflections, atmospheric attenuation and material permittivity effects.

Various approaches are considered by the market at large to fall under the category of RTLS, though they are not included in the ISO 24730 standard. These other include e.g.:

* RSSI Received Signal Strength Indication is a procedure for utilizing the strength of a signal.RSSI, named "indication" may be applied successfully, but just to determine a best estimate for current location. No such solution is published for moving targets (2008Q2). This procedure can either use the strength of a single signal arriving at multiple readers, or the strength of multiple signals arriving at a single reader. As with any RTLS solution, accuracy can be variable based on the environment and number of RF reading devices. This method is most common in systems based on the IEEE 802.11 standard for wireless networks, and is most frequently used in tightly enclosed indoor areas (hospitals, office buildings, etc.)

* Index position relations with coarse fixed location incrementing, e.g. using RFtag indexing.The index relations alternative makes use of components as with RFID. However the RF tags are fixed and the RF readers are moving. That reverse configuration rules why this systems approach shall not be understood as a trade off from any RFID implementing, where the RF tag identifies an object and the RF reader reads the identity en passant.

* Symbolic locating takes names of locations instead of positions described in coordinates for locations.Symbolic locations as a reference to meet a certain target or to approach a defined locality. This is the standard method for human intellect, as with room numbers, seat numbers, platform numbers and so on. Nobody should try to name such approach as real time locating, as currently no such approach may be safely automated based on cognitive capabilities of machines, computers ar vision systems.

Received signal strength indicator (RSSI)

RSSI is not a metering solutions with RTLS but a necessary and an auxiliary means. Those who had set this notion "RSSI" where closer to physics than those who market this as an RTLS concept. Any indicator may not become a proper metrics with any defined accuracy. This does not affect the basic importance of RSSI to build an ad hoc network.

If there are any contradictions: Test it, once is enough. Discussion with suppliers will show: The limitations are not to be neglected. Many approaches take this into account with hybrid solutions including LASER ranging (LIDAR. But remember, light has just another frequency, but generally adhere to the rules applicable of all types of radio waves.

Line-of-sight (LoS) propagation as a prerequisite for RTLS

Travel time of radio waves between transmitters and receivers can be measured disregarding the type of propagation. But, generally, travel time only then represents the distance between transmitter and receiver, when "line of sight" (LoS) propagation is the basis for the measurement. This applies as well to RADAR, to Real Time Locating and to LIDAR. The term "line of sight" not necessarily refers to visual sight, but to chosen transmission and for the used frequency. This rules: Travel time measurements for determining the distance between pairs of transmitters and receivers generally require line of sight propagation for the applied transmission to obtain proper results.

Whereas for communicating the desire to have just any type of propagation to enable communication may suffice, this does never coincide with the requirement to have strictly line of sight at least temporarily as the means to obtain properly measured distances.

The travel time measurement may be always biased by multi-path propagation including line of sight propagation as well as non line of sight propagation with single or multiple reflections in any random share.

A qualified system for measuring the distance between transmitters and receivers must take this phenomenon into account. Thus filtering signals traveling along various paths makes the approach either operationally sound or just tediously irritating.

Fairy tales

Several publications on the Internet report (e.g. [http://research.microsoft.com/~jws/pubfiles/locationaware-ieeecomputer2004.pdf Invisible computing page 96] ): [quote] "In addition, ultrawideband technology does not require a direct line of sight between tags and sensors" [unquote] , i.e. that Ultrawideband measurement does not require line of sight (LoS) conditions.

But, two things are fact:
* Propagation without direct line of sight does not deliver measures for direct distance.
* Propagation through other material than air delivers other measures for travel time than through airPotentially the other writers reporting the very same phenomenon copied the referred unintelligible nonsense from Microsoft(R) sources. Current research by several university teams confirms that the speed of light defined for vacuum did not change yet.

Line of sight (LoS)

Any of the known locating approaches, either RTLS, GNSS or other system require
* either direct line of sight between the nodes forming the network, or
* common illumination of the operational area.If neither one of the two conditions applies, locating will not work. Especially RTLS is bound to the line-of-sight (LoS) condition. To anybody who is in doubt about this message, feel invited to prove the contrary.However, in terms of wireless transmission, LoS does not require visual sight for humans to enable proper links in radio frequent communication bands. The longer waves of radio transmission may pass visually dense obstacles. But, there is no escape: Electromagnetic waves do not travel around corners or through electromagnetically interfering material.

Non-line-of-sight (nLoS)

Generally the wireless link with any non-line-of-sight propagation (nLoS) enables some propagation between two vertices via a couple of unknown paths. Hence, only the straight wireless propagation path along the line of sight (LoS) between pairs of vertices delivers the appropriate measurement, despite possibly missing optical visibility along this line. Practically such through-the-wall propagation is again LoS metering and may contribute to locating as far as sufficient signal discrimination and filtering is performed. This requires sufficient power to transmit and receive between pairs of vertices through walls. Hence any single metering through walls serves results, but never the correct geometric distance. Errors add to the inherent metering errors in the systems performing the metering.

Consequently, the result of any "wireless nLoS" between pairs of vertices will contribute to build an ad hoc network, but only such "visual nLoS but wireless LoS" may contribute to distance metering. The escape is not as easy: It would require the metering with composite straight line LoS segments between a set of pairs of vertices to form the entire distance. Hence the well defined speed "c= 299,792,458 m/s" of light in vacuum does not apply for transmission through materials. The error on short distances may easily exceed the resolution.

However, despite all statements on possibilities crossing walls with appropriate transmission power and suitable choice of frequencies, the travel time or time of flight (TOF) of radio waves is biased for materials, because any material will affect the speed of travel.

The proof is easy: Test it. Is the offered solution simply operating and makes it expedient use of nLoS e.g. in buildings for distance metering or not??

Peak pulse forming

The signal flow from transmitter to receiver is much better performing, when with digital modulation schemes the generating of one or two signals may use special pulse forming to improve the determining of direct wave propagation signals more easily. Time of arrival of any single sharp pulses is always better determined even in noisy environments than any continuous wave carrier modulated with switches signals. For reconstructing the received signals from ambient noise overlays, such signal types like sinc pulses (sinus cardinalis) offer special characteristics.

Various systems are offered that make used of advanced signal procession technologies. However, even sharp pulses are not immune against multi path propagation and will come of several paths. Publications that neglect this fact might refer to systems not capable to filter properly the direct path signals from clutter.

Multi-frequency approaches

Using more than one signal frequency does not change the physical constraints of communications. However the statistical qualities allow for eliminating the influence of stochastics by correlation across the frequencies. This requires the higher time measurement resolution as with phase measurements.

Time synchronization versus asynchronous oscillators

To measure a distance via the travel time of a wave requires an exact conversion of time to distance. Hence stability of the time reference is an asset requiring extensive effort. There are three concepts in the market, which differ in cost and precision:

* local oscillators with phase locked loop control but without synchronizationThis solution may be found with the ISO/IEC 24730-5 standard proposal. It appears strange, that such approach will provide sound metering, but it works at a very interesting balancing of benefit and cost.
* local oscillators with central synchronization and local computingThis solution is the classical approach with GNSS systems, as with GPS and GALILEO. It requires a very expensive infrastructure, which was initially paid just with military requirements. The scope of applications is vastly beyond military, however private equity funding did not discover this operation as too much interesting. Due to military non disclosure requirements, such solutions are yet not set out for publishing details as standards, as earlier foreseen with ISO/IEC 24730-4.
* centralized systems with broadcast time informationThis solution defines the lowest requirements in hardware for the nodes, but either puts a heavy burden on the wireless communication or invests largely in cabling. Such solutions are not set out for standardization.Each party deciding on implementing RTLS must tale all the constraints into account.

= See also =

References with Wikipedia

All additional references within Wikipedia and external references on the Internet may be found on the RTLS page and with the references to special RTLS topics.
* RTLS implementing
* Real-time locating standards

References with the text

External links and references

* [http://www.iec.ch/ International Standardization in Engineering]
* [http://www.iso.org/ International Standardization in Industry and Science]
* [http://www.iso.org/iso/search.htm Search Page for ISO standards concerning RTLS]
* [http://www.ieee.org/ International Standardization in Electrical Engineering]
* [http://www.ietf.org/ International Standardization in Communications]
* [http://www.aimglobal.org/ International Lobbying in Automatic Identification]
* [http://www.aimusa.org/ US American Lobbying in Automatic Identification]
* [http://www.aimglobal.org/technologies/rtls/ Statement of International Lobbying on Real Time Locating]
* [http://www.autoid.org/SC31/wg5.htm/ ISO JTC1 (TC122) SC31 WG5 Homepage on RTLS]
* [http://www.autoid.org/sc31/MIIM/MIIMn0061_MtgRpt_20071030-31.doc/ ISO JTC1 (TC122) SC31 WG5 Inaugural Meeting on MIIM]
* [http://www.qed.org/ Company driving international standardization on e.g. RFID and RTLS in US]
* [http://www.autoid.org/SC31/wg5.htm Schedule and actual scope of RTLS standardisation in ISO JTC1, TC122 and SC31]
* [http://www.car-to-car.org/ Car-to_Car communications programme funding of the European Union]

Standards

* Real-time locating standards
* [http://www.iso.org/iso/home.htm International Organization for Standardization]
* [http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=50718 ISO/IEC FDIS 19762-5 Information technology -- Automatic identification and data capture (AIDC) techniques -- Harmonized vocabulary -- Part 5: Locating systems]
* [http://www.iso.org/iso/search.htm?qt=iso+24730-1&searchSubmit=Search&sort=rel&type=simple&published=on Real-Time Locating with ISO/IEC 24730]
* [http://www.autoid.org/sc31/wg5.htm Standardization for Real-Time Locating in ISO JTC1 / TC122 SC31 WG5 with www.autoid.org]
* [http://www.autoid.org/sc31/sc31.htm Standardization for Automatic Identification in ISO JTC1 / TC122 SC31 WG5 with www.autoid.org]
* [http://www.autoid.org/2002_Documents/TC%20122_WG%207/N026_RFIDStds_Updt_20010123.ppt www.AutoID.org on Real-Time Locating]

Further literature

* Indoor Geolocation Using Wireless Local Area Networks (Berichte aus der Informatik), Michael Wallbaum, Shaker Verlag , Aachen (2006), ISBN-10 3-8322-4838-2
* Location-Based Services: Fundamentals and Operation, Axel Küpper, Wiley, New York, 2005, ISBN-13 978-0470092316
* Local Positioning Systems: LBS applications and services, Krzysztof Kolodziej & Hjelm Johan, CRC Press Inc (2006), ISBN-13 978-0849333491
* Ortsbezogene Anwendungen und Dienste, 4. GI/ITG KuVS Fachgespräch, Roth, Küpper Linnhoff-Popien (Ed.), Verlag Dr. Hut, München 2007, ISBN-13 978-3-89963-591-1
* Praxis der GPS-Navigation, Martin Friedrichs, Delius Klasing Verlag, Bielefeld (2006), ISBN-13 978-3-7688-1773-8


Wikimedia Foundation. 2010.

Игры ⚽ Поможем написать курсовую

Look at other dictionaries:

  • RTLS implementing — Articleissues essay=August 2008 copyedit=May 2008Implementing a real time locating system into operational environment is the challenging task to prepare some standardized equipment for deployment and tailoring the applications to the special… …   Wikipedia

  • Real-time locating — Articleissues confusing=July 2008 refimprove=May 2008 essay=May 2008: This page specifically concerns operational aspects of RTLS. For methodology issues see locating engine. For technology issues see wireless. According to ISO/IEC JTC1 SC31 and… …   Wikipedia

  • Real-time locating standards — Articleissues copyedit=May 2008According to ISO/IEC FDIS 19762 5 [ [http://www.autoid.org/SC31/sc31.htm] (ISO Subcommittee SC31)] an RTLS (real time locating system) is a combination of hardware and software that is used to continuously determine …   Wikipedia

  • Real-time location services — Real Time Locating Systems (RTLS), also misspelled real time location systems, sometimes named local positioning systems, are electronic systems that are intended to locate people or things by means of small electronic devices carried on these… …   Wikipedia

  • Locating engine — Articleissues refimprove=May 2008 essay=May 2008A locating engine performs the computing of coordinates of locations of objects and of persons based on methods of multilateration or triangulation. This applies mainly with real time locating… …   Wikipedia

  • Intergiciel pour étiquettes électroniques — Un intergiciel pour étiquettes électroniques est un logiciel tiers destiné à simplifier l accès et l exploitation des informations stockées dans des étiquettes RFID issues de l industrie électronique. Les étiquettes électroniques sont le plus… …   Wikipédia en Français

  • Fuzzy locating system — Fuzzy locating is a rough but reliable method based on appropriate measuring technology for estimating a location of an object. The concept of precise or ‘’crisp locating’’ is replaced with respect to the operational requirements and the economic …   Wikipedia

  • Space Shuttle abort modes — A Space Shuttle abort is an emergency procedure due to equipment failure on NASA s Space Shuttle, most commonly during ascent. A main engine failure is a typical abort scenario. There are fewer abort options during reentry and descent. For… …   Wikipedia

  • Automatic vehicle location — Automatic vehicle locating or AVL is a concept for determining the geographic location of a vehicle and transmitting this information to a point from where it was requested. Telelocating using location based services is a common term in the… …   Wikipedia

  • Unilateration — Articleissues essay=May 2008Unilateration is a special breed of RTLS, i.e. locating or positioning systems. Unilateration is a special simplification applied to Real time locating as a technology to observe a process from any distance. Unless… …   Wikipedia

Share the article and excerpts

Direct link
Do a right-click on the link above
and select “Copy Link”