Review
A survey of active and passive indoor localisation systems
Gabriel Deak ⇑,1,Kevin Curran,Joan Condell
Intelligent Systems Research Centre,University of Ulster,Derry,N.Ireland BT487JL,UK
a r t i c l e i n f o Article history:
Received 25June 2011
Received in revised form 25April 2012Accepted 6June 2012
Available online 26June 2012Keywords:
Indoor active localisation Indoor passive localisation
Location estimation techniques
a b s t r a c t
In recent years the need for indoor localisation has increased.Earlier systems have been deployed in order to demonstrate that indoor localisation can be done.Many researchers are referring to location esti-mation as a crucial component in numerous applications.There is no standard in indoor localisation thus the selection of an existing system needs to be done based on the environment being tracked,the accu-racy and the precision required.
Modern localisation systems use various techniques such as Received Signal Strength Indicator (RSSI),Time of Arrival (TOA),Time Difference of Arrival (TDOA)and Angle of Arrival (AOA).This paper is a survey of various active and passive localisation techniques developed over the years.The majority of the local-isation techniques are part of the active systems class due to the necessity of tags/electron
ic devices car-ried by the person being tracked or mounted on objects in order to estimate their position.The second class called passive localisation represents the estimation of a person’s position without the need for a physical ags or sensors.
The assessment of the localisation systems is based on the wireless technology used,positioning algo-rithm,accuracy and precision,complexity,scalability and costs.In this paper we are comparing various systems presenting their advantages and disadvantages.
Ó2012Elsevier B.V.All rights reserved.
1.Introduction
Indoor location determination has become a crucial component in many applications.Unfortunately a standard for indoor localisa-tion does not exist yet.Various systems can estimate the position of a person or object.The selection of the technique to estimate location is application dependent.One can select the system which offers the accuracy and precision required for a specific application.Location aware systems can be a very important component for many scenarios such as asset tracking,health care,location based network access,games,manufacturing,government,logistics,industry,shopping,security,tour guides,and conferen
ce guides.In this paper we classified the localisation systems into active and passive systems.Location tracking techniques can be classified into two categories:(1)systems requiring tracked persons to par-ticipate actively;and (2)systems using passive localisation.Fig.1presents the two classes which are also known as active and pas-sive tracking systems.By participating actively,we mean that a person carries an electronic device which sends information to a positioning system helping it to infer that person’s position.In some cases the electronic devices can also process recorded data and send the results for further processing to an application server running the localisation algorithm.In the passive localisation case,the position is estimated based on the variance of a measured sig-nal or video process.Thus the tracked person is not carrying any electronic devices to infer the user’s position.
Another classification can be done such as physical location,the place in the ing places,houses,offices,restau-rants,or as a place online known as a virtual location [1].The means by which people interact has changed dramatically.The number of people using social networks,online games or other on-line services increases each year.We refer to a virtual location as a ‘‘location’’online where people can meet,chat or share informa-tion.This is not a physical location such as a GPS coordinate or a measurement that can pinpoint a user’s location on a map,but rather it is represented by a location on the a web site.The physical location class is the focus of this survey thus various location estimation systems will be reviewed in this paper.
The physical location class can be broken down into three subcategories:descriptive locations,spatial locations and network locations [1].A location related to geographic objects such as mountains,lakes,cities,roads,countries or other structures that have a description such as name,identifier,or number is known as a descriptive location.The spatial location represents a point expressed by two-or three-dimensional coordinates in a Euclidean space.Spatial location is used more in professional applications where a descriptive location does not provide enough details.
0140-3664/$-see front matter Ó2012Elsevier B.V.All rights reserved./10.1016/jcom.2012.06.004
⇑Corresponding author.Tel.:+44(0)2871375418;fax:+44(0)2871375470.
E-mail address:Deak-G@email.ulster.ac.uk (G.Deak).
1
School of Computing and Intelligent Systems,Faculty of Computing and Engineering,University of Ulster,Derry,N.Ireland,BT487JL,UK.
Network location refers to a location based on the topology of a communications network.A user’s devi
ce position in a network is achieved based on its Internet Protocol(IP)address.However in mobile networks a network location is achieved from the base stations used by the mobile terminal.
The localisation technologies proposed in the literature are also classified as indoor and outdoor localisation systems.The GPS [1–5]is widely used for outdoor position determination and this technology is currently implemented in many mobile devices. GPS however cannot estimate location in indoor environments due to the technology request for Line-of-Sight(LoS)when con-necting to satellites.Thus systems were developed based on the proprieties of radio waves which can be used indoors.This class of localisation systems is the focus of this survey paper.
The well known GPS(Global Positioning System)is now used as a component in most of the mobile devices on the market.There are still many devices with no GPS support.In those cases network providers make use of existent base stations to estimate location [6].Cellular networks such as GSM(Global System for Mobile Com-munications)or UTMS(Universal Mobile Telecommunications Sys-tem)use tower cells as base stations.The techniques used to estimate location in a cellular network have been successfully used in Wireless Local Area Networks(WLANs)and Ultra-wideband (UWB)positioning systems.In the last two cases,access points (APs)or dedicated base stations are used to infer location.Existent systems include dedicated positioning systems where the tech-niques u
sed are based on infrared,Radio Frequency Identification (RFID)or ultrasound technologies.Many location estimation tech-niques have been proposed over the years such as infrared,ultra-sonic,bluetooth,radio frequency and also hybrid technologies [2,7,8].
Recently wireless indoor localisation has become a popular re-search subject.Many have focused on hybrid location systems combining two or more techniques in order to improve the accu-racy and precision of the location estimation[2,9,10].
Thefirst applications developed using location-aware systems were able to route a phone call to a phone located near to the user’s location,to use the nearest printer or to display information based on the position of the user.Location information is also helpful for monitoring daily activities of a person[7].
The vast majority of these technologies have the requirement that the tracked person carries a physical electronic device,which in some cases can process some information and send the results for further processing to an application server running the localisa-tion algorithm.Multiple wireless technologies can be used for wireless indoor location determination.These technologies may be classified according to the location positioning algorithm,the physical layer or location sensor infrastructure[8].Location sens-ing approaches can be classified as follows:locationfingerprinting (scen
e analysis)[11],triangulation,trilateration,hyperbolic latera-tion,proximity,and dead reckoning[2,7,8,12].The metrics used in most of the approaches are:Received Signal Strength Indicator (RSSI),Time of Arrival(TOA)/Time Difference of Arrival(TDOA), and Angle of Arrival(AOA)or Direction of Arrival(DOA).The fol-lowing list represents a range of possible positioning technologies: (1)wireless local area networks(WLANs),(2)Ultra-wideband, (UWB),(3)Field strength systems,(4)Radio Frequency Identification(RFID),and(5)Next-generation indoor positioning systems[13].
The RSSI-based localisation techniques are considered more attractive because of their simplicity and robustness in environ-ments affected by multipath compared to the techniques based on metrics like time or angle[14].The RSSI-based position estima-tion can be classified as follows:terminal assisted,terminal based and network based[1,15].The terminal assisted mode is based on RSSI measurements taken by the target and sent to a server which is managing the radio map and is running the localisation algo-rithm.For the terminal based mode,the radio map is built on the terminal and used to determine the target’s position.The network based method uses measurements taken in the environment by ac-cess points(APs)or Basestations(BSs).Indoor localisation systems based on signal strength have the advantage of using the existing WLAN infrastructure,and therefore do not have any extra deploy-ment costs[16].
UWB has some advantages compared to the WLAN location estimation such as:not affected by other RF signals,easy to distin-guish correct signals from those generated from multipath,the UWB signals pass through walls.However interferences can be caused by materials like metal or liquid.The effects of metallic and liquid materials can be reduced with a good placement of the UWB readers.UWB technology is suitable for2D and3D loca-tion estimation.The most common methods used in a UWB local-isation system are TDOA and AOA.A combination of TDOA and AOA can reduce the number of sensor required for a system using just TDOA.
The RFID position estimation is based on electromagnetic com-munication between RFID readers and RFID tags.The RFID tags can be passive or active.The range of the passive tags is limited to approximately1–2m and another drawback is the high cost of compatible readers[2].The active tags have a much longer range around tens of meters which makes them suitable for larger environments.
The next-generation systems can be considered as hybrid sys-tem which will rely on mobile platforms to estimate location.Such systems are implemented today for mapping,vehicle navigation and robot navigation.Currently it is not possible to present advan-tages or disadvantages of these systems as the systems are known only as future work.
In[3–5,17,18]various functions that could be implemented for DfP(Device-free Passive)localisation systems are presented.These functions can be classified as follows:tracking,identification,mul-ti-person and automatic construction of a passive radio map.
Fig.1.Localisation techniques taxonomy.
1940G.Deak et al./Computer Communications35(2012)1939–1954
Our paper focuses on indoor active and passive localisation sys-tems.Previous survey papers such as[8,19–23]classified indoor active positioning systems based on the technology and algorithms used.Our purpose is to provide key features of recent active and passive localisation systems.To the best of our knowledge passive localisation was not covered nor specified in any of the previous survey papers.We have describedfive passive localisation systems. The passive location estimation systems were classified based on techniques such as Device-free Passive(DfP),Ultra-wideband (UWB),Physical Contact,Differential Air Pressure and Computer Vision,used to estimate location information.
The remainder of this paper is organised as follows:Section2 presents the comparison of methods used for deploying Active Localisation Systems,in Section3various passive localisation ap-proaches ar
e presented with examples.Section4concludes the paper.
2.Active localisation systems
The vast majority of indoor location estimation systems have the requirement that the tracked person carries an electronic de-vice or tag,which in some cases can process information and send the results to an application server running the positioning algo-rithms for further evaluation and/or processing.Various metrics used in indoor localisation approaches are Received Signal Strength Indicator(RSSI),Time of Arrival(TOA),Time Difference of Arrival (TDOA)or Direction of Arrival(Angle of Arrival,AOA).Positioning lower cost than the Pinpoint system.SpotON is actually based on
a product called AIR ID sold by RFIDeas,a company from Illinois
[26].One of AIR ID’s limitations is the RS232basestation protocol. RS232cabling is not ubiquitous,has a limited run length,and a limited number of serial ports existent on the server.In order to have moreflexibility the system uses the Hydra microweb server (see Fig.2).Hydra was developed at Xerox PARC but it is no longer an active project[27].
SpotON is based on multiple base stations measuring the signal strength.A central server has the role active下载
of aggregating the values and estimating the position of the tracked object using triangulation. The estimated location is then available to a client application. The system presented in[25]is a prototype which only offers the immediate location of an object and does not store any data on the server.A visualisation client offers the possibility to view the estimated location.The limitations of the system are significant. The precision and accuracy are comparable with a rudimentary motion sensor thus the real-life applications are limited to the user’s requirements[25].SpotON can be used to deploy applica-tions such as light control or similar applications which do not re-quire a high accuracy.The authors in[25]mentioned that an object can befixed to a position with an accuracy of around3m.
The measurement frequency for each location can take between 10and20s.Significant data can be lost because of the long time required to get one measurement.Another aspect that can further limit the system is the power supply,with two lithium coin cell batteries providing10h of normal operation[25].Although the system has significant limitations,the idea behind SpotON is that low cost off-the-shelf technology can help gain more experience in location estimation and the required parameters.
2.2.Ultra-wideband(UWB)
Ubisense is a commercial location estimation system.Ubisense has very high precision,approximately15cm for95%of the read-ings due to the use of the active tags signal triangulation[28,29]. The main components of a Ubisense systems are:the sensors,the tags to be tracked and the software platform.The Ubitags use RF Radio(2.4GHz)to coordinate the UWB(6–8GHz)pulse transmis-sion time.The system uses both TDOA and AOA in order to esti-mate the location of a specific tag.The3D location of an Ubitag can be estimated using at least two receivers.
Ubisense does not require Line-of-Sight due to the UWB tech-nology.The signal can befiltered and the multipath effect mini-mised.This is huge advantage because the multipath effect represents the main cause for low accuracy indoors.Ubisense can cover large areas and offers the possibility of tracking a large num-ber of user in real-time.The spatial coverage is ensured using clus-ter methods running a large number of services with a fair usage of bandwidth.The system works in similar manner with a cellular network where the environment is organised into cells with at least four sensors/readers.
A drawback of Ubisense is the timing cable required for every Ubitag which can become challenging in some environments.Ubi-sense can estimate the location of the Ubitags with very high accu-racy.However the price to deploy such a system is high,about $16.875for the research package[21].
2.3.IEEE802.11a/b/g/n WLANs
2.3.1.Ekahau real time location system(RTLS)
Ekahau is a commercial positioning system that uses WLANs and tracks electronic devices such as tags,personal digital assis-tants(PDAs),PCs,handsets equipped with wireless network cards. The estimated location is computed correlating space information with the signal strength measured when the device is connected to
Fig.2.Hydra microweb server[25].
G.Deak et al./Computer Communications35(2012)1939–19541941
various access points(APs).In order to achieve this it is necessary to consider the propagation characteristics of the recorded signal and to use advanced probabilistic mathematics.
Various location estimation methods such as using the signal strength of standard Wi-Fi infrastructures are used.This method performs better in comparison with time-based methods which can be expensive considering the proprietary infrastructure re-quired[2,8].Another drawback is the poor performance of the time-based estimation methods indoors.The signal’s short travel time and also common obstructions such as human presence,walls or other objects represent the main reasons for a
poor perfor-mance.The Ekahau RTLS is actually a combination of tools(see Fig.3)such as the Ekahau Client,Ekahau Positioning Engine,Eka-hau Manager,Ekahau Planner,Ekahau Application Framework and SDK.
The application running on the server in charge of computing the device position is the Ekahau Positioning Engine(EPE).The Positioning Engine is a Java-based software that can estimate the position in two ways:as client location(x,y),floor,speed,direction or as logical information(‘‘mail room’’,‘‘meeting room’’).
Location maps can be created as positioning models using the Ekahau Manager.This is a stand-alone tool for drawing logical areas,testing positioning information and analysing the accuracy of the estimated location.After the positioning models are created, the models can be saved in the Positioning Engine.The Manager of-fers functions such as location maps management,performing site calibration,permissions for networks and APs,tracking rails man-agement(provide improved positioning accuracy),displaying de-vice locations and properties for administrative purposes,logical areas management(used to determine whether a client device is
Fig.3.Ekahau RTLS overview[30].
1942G.Deak et al./Computer Communications35(2012)1939–1954
This method can be used only if the tracked users repeat a task reg-ularly.A similar method was implemented for outdoor location estimation using GPS.The location of a car entering a tunnel can be predicted based on a map,the direction of movement and the speed.
2.3.2.Microsoft research radar
Radar is probably thefirst example of a positioning system using IEEE802.11networks[32,33].The main issue to be ad-dressed in this case is the noisy characteristic of the wireless radio which can be affected by multipath fading,reflections,and obstructions.In order to address the noise problem,Radar uses thefingerprinting method which is an offline method that involves recording a radio map of the environment.This map is built based on the radio strength of the wireless signal in specific locations in the environment.Thus the radio characteristic is linked to a phys-ical location in the environment.The Radar system computes the location by monitoring the signal strength of the tracked devices and comparing the value with the database entries used to create the radio map.This method represents the empirical method used by Radar.A second method is a mathematical model based on measuring the radio propagation properties of the environment to be monitored.In the mathematical method the ra
dio map was replaced with the indoor propagation properties and layout infor-mation of the environment.
Radar has an accuracy of2–3m(approximately the size of an office room)with a probability of50%using scene analysis.A sec-ond deployment uses lateration with an accuracy of4.3m and a similar probability as the scene analysis.Another drawback of the Radar system is the change of the indoors radio propagation whenever significant changes of the environment occur[33].The only method to overcome this problem is reconstructing or creat-ing a new signal strength database.Radar can be used to detect the direction and orientation of the user.The system also allows the 2.3.3.AeroScout
AeroScout is a Wi-Fi-based localisation technique[34].An exis-tent wireless infrastructure is used to compute the location of any mobile devices using the802.11b/g standards and can also track the AeroScout tags.The system uses Time Difference of Arrival (TDOA)for larger indoor environments and outdoors,and Received Signal Strength Indicator(RSSI)for smaller size indoor environ-ments.The AeroScout receivers are used differently based on the localisation technique implemented.Thus for TDOA the receivers are long-range Wi-Fi readers and for RSSI the system uses the same receivers or in some cases it uses Cisco access points as readers.
The AeroScout system presented in Fig.4can be deployed using AeroScout Wi-Fi tags,location receivers,exciters and an AeroScout Engine.
AeroScout tags are battery-based tags with a long battery life which can be extended if the tags are switched off whenever their location is not within the monitored area.The tags use standard wireless communications enabling the possibility to compute the location of people and assets which otherwise are not connected to a802.11wireless network.They are equipped with wireless or serial interfaces which can be used for remote programming.Tra-ditional RFID require dedicated readers in comparison to AeroScout which uses standard wireless access points as long-range readers in order to estimate location.Long-range tracking is deployed using standard wireless access points as readers where for short range the tags can be triggered by the AeroScout Exciters.The range covered by the exciters is approximately6.5m.In order to extend the coverage area multiple exciters can be linked together behaving as a single exciter.
AeroScout requires TDOA measurements in order to compute the position of the tracked tag or device.AeroScout receivers have the main role to record the TDOA measurements of standard 802.11messages and send this information to the AeroScout Engine.Each receiver has the capability of processing approxi-
Fig.4.AeroScout architecture[34].
G.Deak et al./Computer Communications35(2012)1939–19541943
connection there is no need for extra hardware transforming the Intel Place Lab into an inexpensive localisation system.Nowadays the number of devices with integrated Wi-Fi has increased,thus the user can download the localisation software for free from the Place Lab web site.This localisation system uses the unique ID broadcast by the access points whenever a user is trying to con-nect.Even though this unique identifier seems useless for many users it can be used as a parameter connecting the device to a spe-cific location where the AP is deployed.
The majority of the localisation systems using Wi-Fi require line-of-sight from the device to the APs.Place Lab does not need line-of-sight to determine the location and offers an accuracy of 20m.Fig.5presents an overview of the Intel Place Lab compo-nents.A client running on the user’s device which does a survey of the area to be monitored searching for APs and recording the un-ique IDs of the APs found.The unique IDs are linked to the GPS po-sition of the AP,building a map of the available APs.The method to compute the location of the user’s device is triangulation by com-paring the unique ID sent by the APs with the entries recorded in using Global System for Mobile Communica
tions(GSM)and Bluetooth devices in parallel with the802.11APs providing an accuracy of20–30m.
The key components of a Place Lab system are:radio recordings also known as beacons,databases containing the location informa-tion of the beacons and clients estimating current location based on the data from recorded in the databases.
The user’s device is listening for beacons sent by Wi-Fi, Bluetooth,and GSM and compares the unique ID received with the database entries of a precomputed map where the ID is directly linked to a physical location.
The Place Lab does not encounter privacy issues as the location of user’s device is computed entirely on the device itself using seg-ments of the database of known beacons from that specific loca-tion.Whenever a new client starts up or an existing client is moving to a new location the device downloads segments of the database representing beacons from the APs in the surroundings. The user can receive relevant information such as closest restau-rant,hotel,and post office based on the location of the device.
Fig.5.Intel Place Lab Overview[37].
1944G.Deak et al./Computer Communications35(2012)1939–1954
版权声明：本站内容均来自互联网，仅供演示用，请勿用于商业和其他非法用途。如果侵犯了您的权益请与我们联系QQ:729038198，我们将在24小时内删除。
发表评论