In recent years tracking system become too much popular. There are different kinds of tracking system, some for indoor and some for outdoor. Different tracking system has different characteristics like deployment difficulty, cost, target usage field, equipment size, response time, accuracy, and different tracking technology and location determination algorithms.
Tracking services are now used in our mobile handset, PDAs and those have great impact in near future in our every day life. We can use this system for various purposes like searching for an unknown location, if we lost the direction this tracking system help us to reach our destination without any difficulty. Global Positioning System (GPS) works well in outdoor environment and operate in real time application. There are lots of tracking methods for indoor situations like cellular based, Wireless LAN (WLAN), home RF, RFID, Bluetooth, IrDA. Most of those tracking methods are not able to provide acceptable accuracy and the cost of those systems is too high.
For short range application we use Bluetooth and ZigBee technology. Bluetooth deployed for low bandwidth communication scenarios and for low cost. We use it as like a cable replacement. Normally it is being used in mobile phones, portable computer and handsets, many modern digital cameras, televisions and printers. Bluetooth form a personal area network (PAN) which connect all the personal electronic equipments which are very close to each other.
Now most of the handset has Bluetooth and it could be used for indoor tracking scenarios. Precision is too much important in tracking and positioning system. Different technology has different accuracy and precision. Dedicated sensor network have great precision and accuracy, on the other hand infrared and ultrasonic signal have about 15 cm accuracy levels. [1] For all the tracking system, we need to build various positioning infrastructure, high configurable mobile device. There are another tracking system based on IEEE 802.11 which is called WLAN (also called Wi-Fi) and WLAN enabled devices are available in the market [2]. For location estimation this system monitor signal strength from surrounding base stations simultaneously and calculate the distance to those base stations. In here triangular method is being used.
Bluetooth and Wi-Fi have some similar characteristics like both of them operate in the same frequency band and it is 2.4 GHz. In a tracking system Bluetooth is more advantageous than Wi-Fi. First, the cost of Bluetooth is lower than Wi-Fi only exception for client side because they have to use high configured mobile devices. But the cost for the base station is too low because a self-power Bluetooth dongle is enough for to generate positioning beacons. Secondly, Bluetooth consume lees power than Wi-Fi. It requires less power to transmit and has an automatic transmission power control mechanism. Wi-Fi uses on demand power mechanism but Bluetooth uses always on power mechanism because BT uses less power. So less battery power is being consumed. The using of Bluetooth technology is increasing day by day and the growth rate is just behind than Internet and cellular phones [3]. Bluetooth enabled mobile phone is increasing rapidly but Wi-Fi does not reach such kind of penetration.
Bluetooth based positioning and tracking system can be used in many sectors like a system which give direction in a large massive building and corridor like airport, museums, hospital, shopping mall etc and also provide context based information for departmental store and museums. Availability is too much important than accuracy for mass application for low cost positioning technology.
First step for tracking process in Bluetooth is stand on a sporadic investigate for Bluetooth equipments in the neighborhood at diverse positions (query process). As a outcome the queerer gets a list with addresses of detectable BT equipments. Identified equipments are marked by a first-seen/last-seen timestamp and a location-stamp. All outcomes then front warded to a fundamental tracking record and concatenated stand on the distinctive Bluetooth equipment address (BDADDR).The scheme consists of disseminated Bluetooth query scanners or sensors linked to a innermost tracking database. For time harmonization a NTP server is used and for analyzing and visualizing front-end based on an Apache web server is used [4].
The 1st Chapter presents the objectives, endeavors and introduction of this project.
The 2nd Chapter gives the background knowledge of Bluetooth technology, different technique and algorithms which is normally used in tracking system and different versions of Bluetooth which is frequently used in positioning system
The 3rd Chapter gives the information of the tracking system and which is needed for the test and how to do this. Here some constraints are discussed very briefly. Some equation and data and coding which is needed to perform this test.
The 4th Chapter gives the information about legalization of the test results and discuss about the interference between Bluetooth and WLAN especially about Wi-Fi.
The 5th Chapter discusses about the future work and the perspective of Bluetooth and concludes the project.
CHAPTER 2: Background:
2.1 Bluetooth:
2.1.1 System Architecture:
Bluetooth operates in the 2.4-GHz frequency band. This band is dedicated for industrials, scientific and medical (ISM) use for wireless communication as small distances and is unlicensed on a universal basis. In place of modulating data at a precise frequency, Bluetooth occupies frequency hopping in 79 hops dislodged by 1 MHz, starting at 2.402 GHz and ending at 2.480GHz. Frequency hopping is employed to lessen the effect from signal intervention and fading, with the anticipated assortment of a typical Bluetooth apparatus being roughly 10 or 100 m and genuine range (10 or 100 m) depending on the power level. Bluetooth was developed with the universal community in brain, some exceptions to the basic use of frequency materialized as a result of different frequency consumption policies all over the world. Two most important exceptions to the use of the 2400 to 2483.5 MHz frequency band are France, where the 2483.5 MHZ band is used, and Spain, where the 2445-to 2475 MHz band is used. Products developed for those markets will have reduced frequency band and utilize a dissimilar frequency-hopping algorithm. 79 frequency channels spaced 1 MHz apart result in channels spaced from 2402 to 2480 MHz. Under the frequency-hopping formats employed by Bluetooth, the 1600 hopes per second translate into duration of 625 μs prior to hopping to another frequency within the previously mentioned frequency range. [5]
2.1.2 Power requirement:
All the Bluetooth handheld devices is dependent on batteries for power. They are three different class devices available.
Table 1 : Power level of different BT classes [6]
Class
|
Power required
|
Range
|
1
|
0 dBm(1 mW)
|
10 m
|
2
|
4 dBm(2.5 mW)
|
10m – 20m
|
3
|
20 dBm(100 Mw)
|
100m
|
2.1.3 Power operation modes:
Bluetooth device can be classified according to three power classes. Bluetooth also defines requirements for three low-power modes as an apparatus to reverse battery life. The three power modes, which surpass the presented power classes, include sniff mode, hold mode, and park mode. When in the sniff mode, a device listens to the network at a concentrated procedure rate that results in a lessening in its power recruitments. In the hold mode, the device’s clock continues to operate and remains in synchronization with the master; however, the device is not a participant in a Bluetooth network. In the hold mode, the device remains its active member address. The third inferior power mode, which is the park mode, is similar to the hold mode, but the appliance does not hold its dynamic member address because that address is abandoned when it goes into the park mode. The power obligation diminishes as a device goes from sniff to hold to park mode. It should be noted that the time to recover contribution in a Bluetooth network is normally inversely proportional to the power operation mode of a device. That is, when in sniff mode, the device is listening for transmission and can react quicker than when it is a dissimilar power mode. [5]
2.2 Tracking methods:
Normally tracking method was built on basis of the triangulation algorithms. In triangulation method a circles is drawn centered a radio frequency access point, and the radius of the circle directly depends on measured signal strength or the time to transmit the signal to the mobile station. If there is three terminals intersect each other in a point, then the point indicates the position of the mobile station. Sometimes several intersecting point may exist, in this scenario the intersecting make a common area. By calculating the average of the coordinates of the intersecting points which make the common area we can determine the location of the device.
Figure 1: Triangulation method [7]
Here there is three fixed point to identify the unknown location. Measuring this signal strength a circle is drawn from every access point. The three circles intersect a point and it is the position of the mobile device.
Several techniques already been established on the basis of TN method by utilizing time, time difference, angle of arriving signal and signal strength which for determining the distance between base station and mobile terminal. I will mention some of those techniques:
2.2.1TOA/RTOF:
By measuring time it is easy to determine the distance between the terminal and mobile unit. [8]
D= c*t
Where, D= distance (m)
c= velocity of light (300m/µs)
t= time (µs)
Figure 2: Time of Arrival [9]
2.2.2RTOF:
It is an improvisation of TOA approach where moderate relative clock synchronization takes place instead of absolute synchronization. Measuring unit sends scanning signal which act like a radar. For measuring the complete round trip propagation time, MU response to the signal of the measuring unit. Measuring unit know the exact procession/delay time caused by MU.
2.2.3TODA:
In this system mobile unit sends positioning signals to its surrounding measuring units and time difference of the received signal is evaluated. In this system every tower or measuring unit measure the time of its receiving a phone's signal. Then they translate this information to measuring the distance of the MU from the tower. Its need to do the synchronization of the measuring units which is considering the main benefit of the TODA system and this is done by the backbone network.
Figure 3: time difference of arrival [10]
2.2.4 AOA:
In this system position is calculated by using goniometry. In here directional antennas or antenna arrays is being used. AOA system works along TODA system. Measuring unit measure the angle of the receiving signals by using antenna arrays or directional antennas. The intersecting of several directional pointers produces the positioning of the mobile unit. The main drawback of this system is accuracy. It is not problematic for outdoor environment but it is problematic for indoor environment because of the size of measuring unit. Shadowing is another problem for this system
Figure 4: Angle of arrival [10]
this occurred by multipath reflection from misleading directions an ultimate result error in measuring.
2.2.5 RSS:
It is based on the lost equation of the radio propagation. We know for free space; propagation loss LB is proportional to 1/r2. This equation is not suitable for real time application and mostly in urban areas and indoor environment where shadowing and multipath fading have major effect on radio propagation. For illuminating those problems we use advanced propagation models or fingerprinting techniques. Fingerprinting means collection of the actual field radio propagation distribution and some type of learning. For calculating the position probability estimation and neural network are being used. Most modern mobile wireless standards provide some reading called received signal strength indicator. Fingerprinting algorithms depends on RSSI. Here a receiver receives a positioning signal and it compares the signal pattern with the signal pattern which is already stored in fingerprinting database. By comparing, estimated position is calculated.
2.3 Bluetooth positioning:
For outdoor condition we use satellite based positioning system, on the other hand for indoor local wireless positioning system we use Bluetooth, GSM/UMTS and WLAN. Accuracy is too much important for positioning and the accuracy of infrared and ultrasonic system within sub decimetre range. Bluetooth, GSM, WLAN give precision of about some few meters when the condition is quite favourable. Time based positioning system is too much complicated and require high accuracy clock synchronisation, and not possible to deploy in current wireless tracking system because it needs extra expensive smart antennas. So normally signal based tracking system is being used and it is much popular for low cost and availability.
In GSM based tracking system, for obtaining high accuracy, need much denser base stations deploymentation. The accuracy for GSM and UMTS system is not good because normally the base stations are few hundreds meter or several kilometres far apart from each other.
WLAN technology becomes popular day by bay for indoor location by measuring signal strength. For determining the position of mobile WLAN device from different access point there are some algorithms like kNN (K nearest neighbour averaging), simple triangulation and Bayesian analysis [11]. In WLAN, signal strength can be sampled simultaneously from different access point and it is done at a per packet level.
WLAN or Wi-Fi is fixed but Bluetooth is ubiquitous and it is ‘always on’. On the other hand Wi-Fi is manually switched on when the connection is needed because it consumes more power than Bluetooth. Ubiquitous connection and communication are main advantages of Bluetooth.
Bluetooth tracking has some disadvantages [12], those are:
1. For measuring signal strength its need an active Bluetooth connection. So connection time is higher than usual.
2. Most of the Bluetooth mobile support only one connection at a time so it makes the triangulation too much difficult to calculate the actual location of the Bluetooth devices.
3. Frequency hopping is being used in Bluetooth which make the location assumption more difficult. Inquiring a device by using frequency hopping is twice than normal frequency hopping which is being used in typical ordinary connection [13]
2.4 Bluetooth Standards:
Bluetooth device has some standard. I will describe some of them which is being used in tracking.
2.4.1 Bluetooth 1.1 [6]:
It is ratified as IEEE standard 802.15.1 and this was released in 2002. Many errors which are found in the 1.0B was misplaced. This version gives support for non-encrypted channels. Radio Signal Strength Indicator (RSSI) was introduced in this version.
2.4.2 Bluetooth 1.2 [6]:
This version covers all the feature of version 1.1 and added some other new features.
1. It provides faster connection and device discovery.
2. It provides adaptive frequency hopping spread spectrum (AFH) which give better performance against interference because it avoid the use of busiest frequencies of the hopping spectrum.
3. Provide higher transmission speed which limits up to 721 Kbit/s.
4. Provides Extended Synchronous Connection (eSCO) for good voice quality of audio links. Here corrupted packets are retransmitted and also increase the audio latency for providing better support for concurrent data transfer.
5. Three-wire Universal Asynchronous Receiver/Transmitter (UART) is supported by Host controller Interface (HCI).
6. This version introduced retransmission and flow control for L2CAP
2.4.3 Bluetooth 2.0 [6]:
It provides all the feature of Bluetooth 1.2 version. It introduces Enhanced Data Rate (EDR) for higher and faster data rate. The normal data rate for this version is about 3 Mbp/s but in practical we get 2.1 Mbp/s. The high data rate is being achieved because it combines two modulations GFSK and PSK. EDR reduce multiple simultaneous connection complexity and reduce duty cycle and give low power consumption.
2.5 Physical parameter for positioning in Bluetooth:
There are some physical parameters for positioning in Bluetooth. The most relevant and common parameters are Received Signal Strength Indication (RSSI), the Link Quality (LQ) and the Bit Error Rate (BER). There are some other parameters like Transmit Power Level (TPL), Inquiry Result with RSSI and so on.
2.5.1 RSSI:
RSSI is a power level of a receiver which denotes either it is within or below/above of the Golden Receiver Power ranges (GRPR). It is 8 bit signed integer. GRPR is treated as an ideal receive power range. Positive RSSI means the RX power is above the GRPR value and negative RSSI means the RX power is below the GRPR value. If it is zero than it means that it is in ideal condition. The upper and lower threshold values of GRPR are droopily bound. Bluetooth has automatic power control. The transmitter and receiver negotiate to each other to adjust the transmitting power in such a way that receiver can receive it optimal power. Distance is a vital parameter for measuring the RSSI, because if the distance between transmitter and receiver is too close or too far, the previously mentioned mechanism does not work at optimum power level, ultimately the RSSI reading is higher or lower than ideal. This is known as X and then the reading is not usable for this condition. Class 1 device can maintain this feature but Class 2 and Class 3 device support it optionally.
Figure 5: Relation between GRPR and RSSI values [13]
There are some relation between the RSSI values and actual receiving power. The received power is measured by the Bluetooth dongle consist of Cambridge Silicon Radio chipset.
Figure 6: RSSI value vs. RX power level (dBm)[13]
In this figure positive values of RSSI indicate that receiving power is greater than the upper threshold, (here the value is -40 dBm). On the other hand the negative value of RSSI shows that receiving power is less than lower threshold value (here the value is -60 dBm). The limit for Golden Receive Power Range is -40 to -60 dBm. The minimum RSSI value is -10 dBm.
2.5.2 Link Quality or Bit Error Rate:
It is an 8-bit unsigned integer number and it quantifies the apparent link quality of the receiver. Vendors specify the exact coding of that integer. Various companies’ adaptor measures the link quality value at different time intervals. The Broadcom Bluetooth adaptor update the measurement every 5 second intervals and there is a little variation of the actual LQ value. BER is recorded by the chipset hardware and it gives a report about the average error which is encountered over the post period for the current connection. The range of the link quality from 0 to 255 and it is updated once per second by CSR chipset. Equations for converting link quality to BER are given below
β = 0, Ql = 255
(255 − Ql)* 0.0025, 255 < Ql ≤ 215
0.1 + (215 − Ql) * 0.08 215 < Ql ≤ 90
10.1+ (90 − Ql) * 0.64, 90 < Ql < 0
67.7, Ql = 0
From the equation it is clear that if the value of Link quality decreases BER also increase. When the link quality value is 255 means top value the bit error rate is zero.
Figure 7: Link Quality to BER relation for CSR Chipsets [12]
Link quality values depend on several factors. First it the mobile terminal is moving in slower speed, the link quality value is valid. But when the mobile terminal is moving in greater speed the link quality value is not up to the mark. Some other factors are adaptive power control and resulting GRPR, Channel Quality Driven Data Rate (CQDDR) and adaptive frequency hopping. All of those factors reduce the interference in the channels.
The Bluetooth dongle also support the echo command. The echo command allows a random packet to be sent to distant devices and those devices a feed back packet holding the same bit sequence inward in the echo system. For measuring the bit errors echo command is used by comparing the transmitted and receiving data.
2.5.3 Transmit Power Level (TPL):
It is an 8 bit signed integer. TPL precise the Bluetooth modules transmit power level. During the connection TPL may vary due to the power control. For class 1 devices the value of the TPL is between +4 and +20 dBm. For class 2 devices it is normally +4 dBm and for class 3 devices it is 0 dBm. Sometimes there is no power control mechanism in class 2 and class 3 devices.
2.5.4 Inquiry Result with RSSI:
Inquiry result with RSSI was introduced in Bluetooth 1.2 standard. The procedure is same as conventional inquiry. The main feature of this technique is that clock offset and Bluetooth MAC address can be retrieve and RSSI value is provided during normal inquiry. No active connection is needed. The radio layer monitors the receiving power level of the current inquiry response which is sent by the response devices. The other name of Inquiry Result with RSSI is “corresponding RSSI”. The transmitted power is device specific so there is no side effect of power control in X. It gives good result than connection oriented RSSI. But it is still suffers from the GRPR values related zero RSSI.
2.5.5 Other Parameters:
By using Ranger’s BT 2100 Bluetooth USB adaptor it is possible to measure the absolute Rx power level through inquiry. This value has very good correlation to the distance. So it is a very good candidate for positioning. The absolute Rx values can be retrieving from certain specific devices. [13]
Figure 8: the relationship between Inquiries based Rx level Vs distance [13]
Chapter 3: DESIGN AND IMPLEMENTATION
3.1. Primary idea:
For tracking a mobile device there are several techniques and algorithms. Among them I choose RSSI to track a mobile device. RSSI is the most convenient and popular tracking system techniques but the accuracy level is not too good. In my project I will do my experiment in a room which is 15m long and 10m width. There are 3 access point, and 1 mobile device which are roaming in the room. Here I will calculate the actual location of the mobile device in the room from any of the access point. I will use triangulation algorithm to locate the mobile device. First of all I give the x and y coordinate of the mobile device and then I will measure the distance of the mobile device from 3 access points and then measure the relative RSSI values regarding distance. Then I will calculate the receiving power of the three access point and then again I calculate the distance with respect to receiving power. Then I will measure the accuracy of this tracking system.
3.2Design:
3.2.1 Basic recruitment of the application:
- One mobile device which we detect
- Class 2 Bluetooth dongle which transmit 4dBm power and the detecting range up to 10 m.
- Ethernet connection between the three access point means the Bluetooth dongle
- There will be a central data base system which will store all the information about the mobile device and those data can use for further analysis.
3.2.2. Design challenges:
For designing a system first of all we have to consider all the limitation or obstruction of the system. How to minimize the limitation it is a very vital issue for designing. If we do not consider those things earlier of our design we have to face a lot of problem when we try to implement the system.
Some limitation of this tracking system:
- Limited processing power
- Limited memory
- Slow network connection
- Battery life
- Interference from WLAN and Wi-Fi
- Shadowing,
- Fading
- GRPR value
- Not accurate data
- RSSI has low resolution
- Update rate is too slow
3.2.3Test Bed:
The experimental room is 15 m long and 10 m width and the area of the room is 150 m2. The unit grid size is 1×1m. We place three Bluetooth 1.2 standard adapters in three such grid positions to act as access points. All the access points are connected to the Ethernet through a USB port and all the connection are connected to the central data base for further analysis. A mobile host is in the room. The coordinate of the access point 1 is (0, 0), the coordinate of the access point 2 is (15, 0) and the coordinate of the access point 3 is (7.5, 10). Like there is a mobile unit in the room. We need to find out the distance of the mobile unit from the three access point and using some mathematical terms of tracking need to locate the exact location of the mobile unit.
The scenario of the room
● AP3 (7.5, 10)
► m u
| ||||||||||||||
●
AP1 AP2●
AP1 AP2●
Figure 9: A room 15 m long and 10 m width with 3 access points and a mobile unit
The position of the mobile unit is (px , py)
The distance between the AP1 to mobile unit is = sqrt {(x1-px) 2 + (y1-py) 2}
The distance between the AP2 to mobile unit is = sqrt {(x2-px) 2 + (y2-py) 2}
The distance between the AP3 to mobile unit is = sqrt {(x3-px) 2 + (y3-py) 2}
Putting the (x,y) coordinate of the mobile device and equations for the three access points become
For AP1 = sqrt {(px) 2 + (py) 2}
For AP2 = sqrt {(15-px) 2 + (py) 2}
For AP3 = sqrt {(7.5-px) 2 + (10-py) 2}
Flow diagram of tracking process:
Figure 10: flow chat of tracking process
3.2.4 Measurement of the RSSI value with respect to distance:
In [14] this experiment they measure the RSSI value associated with distance between receiver and sender. I take those values for my experiment and find a relation between the distance and RSSI
. Table 2: RSSI values with respect to distance
Distance X (m)
|
Value of RSSI dB
|
0.75
|
7
|
1.75
|
4
|
2.25
|
4
|
2.50
|
3.5
|
3.00
|
3.0
|
3.50
|
3.0
|
3.80
|
2.5
|
4.40
|
3.0
|
4.55
|
2.0
|
4.75
|
1.0
|
5.15
|
3.0
|
5.50
|
0.5
|
5.75
|
2.0
|
6.00
|
2.5
|
6.50
|
1.0
|
6.65
|
2.5
|
7.30
|
2.0
|
7.80
|
0.0
|
After plotting the value this graph is generated and it is a third order polynomial function
Graph 1: 3rd order polynomial function of RSSI w.r.t. distance
Here we can observe that most of the points not touch the feting curve, so it has less accuracy
I also use forth order polynomial equation for this experiment and this the graph for also and this the graph for
Graph 2: 4th order polynomial function of RSSI w.r.t. distance
3.2.5 Measuring the Prx value:
After evaluating the value of RSSI it is possible to calculate the value of Prx. The conversion between the RSSI and Prx is mentioned below:
Normally the RSSI reading considers has lower resolution for the GRPR values of the RSSI.
3.2.6 The calculating the distance with respect to Prx :
From the radio propagation model we got the equation for receiving power [15][16]
PRX = PTX + GTX + GRX + 20 log(c/4Πf) – 10 n log (d) - Xa
Here
PRX= Receiving power level in dB
PTX = Transmitted power in dB
GTX = Transmitter antenna gain
GRX =Receiver antenna gain
C = Velocity of light, 3×108 m/s
f = Operating frequency 2.44 GHz
n = Attenuation factor, for free space it is 2
d = The separating distance between the transmitter and receiver
Xa = Normal random variable, whose standard deviation equal to α
Putting the value of c and f the equation become
PRX = PTX + GTX + GRX - 40.2 – 10 n log (d) - Xa
PRX = PTX + G- 40.2 – 10 n log (d)
Where G = GTX + GRX = -4.82 dBi
Xa value is ignored because it is a random error which is unknown.
After calculating the value of PRX we are able to determine the value of d again
The equation for d stands for
d= 10^ [(PTX-40.2-PRX +G)/10n]
Though we use Class 2 device so the transmitting power is 4dBm
d= 10^ [(4-40.2-PRX +G)/10n]
Converting dB to dBm:
For converting dB to dBm we have to add 30.
1 dB= 1dBm+30;
Converting dBi to dBm:
It is same as dB.
1 dBi= 1dBm+30;
CHAPTER 4: Results and Discussion
4.1 Result
I take several values from the three access points and plot them with respect to RSSI and got those graph.
Graph3: the position of the mobile device from the AP1
Table 3: tabulated values of D1 and d1 with respective RSSI
Mean error from the access point 1:
Here, the actual mean position of the mobile device from AP1= 8.981m
The estimation mean position of the mobile device from AP1= 10.3m
The mean error = (10.3-8.981) = 1.319m
From the graph it is clear that few calculation points are very close to its actual point.
Graph 4:The position of the mobile device of the AP2
Table 4: tabulated values of D2 and d2 with respective RSSI
Mean error from the access point 2:
Here, the actual mean position of the mobile device from AP2= 8.843m
The estimation mean position of the mobile device from AP2= 9.274m
The mean error = (9.274-8.843) = 0.431m
Graph 5: Position of the mobile device from the Access Point 3:
Table 5: tabulated values of D3 and d3 with respective RSSI
Mean error from the access point 3:
Here, the actual mean position of the mobile device from AP3= 8.981m
The estimation mean position of the mobile device from AP3= 9.64m
The mean error = (9.64-8.981) = 0.659m
In this graph we see some calculated point is very close to its actual position.
4.1.2 Analysis the result:
►Here we saw that we can calculate the distance of any mobile unit from any particular access point at a time. It is not possible to determine three differences from three access points at the same time because in Bluetooth only one connection is supported at a time.
► GRPR values of RSSI are another problem for getting accurate result. When the RSSI values is 0 then the Prx vales lies between the -40 dBm to -60 dBm but we don’t know the exact value of the Prx. If the values of RSSI are less than -10 dB we get 0 values for Prx. For that reason most of the time we don’t get exact values.
► We don’t know the exact value of RSSI with respect to distance because I took the values from other reports and then make the equation for the RSSI by best fitting in MATLAB. There is two equation for the RSSI with respect to distance, one is forth order polynomial and another is 3rd order polynomial. We follow the forth order polynomial function. I also measure the distances by using the third order polynomial function but don’t get accurate values. If we use forth order polynomial function most of the measuring values of RSSI in the report [14] touch the graph.
4.2 DISCUSSION:
Here I will discuss about the interference between the WLAN and Bluetooth mainly Wi-Fi and Bluetooth and some advanced technique of Bluetooth tracking system.
In discussion about Wi-Fi and Bluetooth interference we discuss what the reason for interference and what the effect of interference and how to mitigate the interference.
In advanced Bluetooth technique I discuss some new technology for tracking which has good precision and accuracy than Bluetooth.
4.2.1 WLAN and Bluetooth interference:
Bluetooth and WLAN are operating in the same frequency. Both of them operate at 2.4 GHz. So there is interference between those two networks. Both WLAN and Bluetooth fail benevolently for the cause of interference. For minimizing the effect communicating protocol will be very robust and there must be error checking and correcting mechanism and resending of corrupted packets. Though interference decreases the data rate so need to send data again and again. In some condition interference is too much severe like when a Bluetooth enabled mobile phone is very near to an in service microwave oven and communication disrupt completely.
4.2.2 How Bluetooth and Wi-Fi interfere:
Wi-Fi and Bluetooth reside in a section of the 2.4GHz ISM band that is 83 MHz wide. Bluetooth use frequency hopping spread spectrum (FHSS) and there will be 79 different 1 MHz-wide channel in this band.
On the other hand Wi-Fi use direct sequence spread spectrum (DSSS). There is no hop and do not able to change the frequency and leftovers centered that is 22 MHZ- wide. There is only three overlying channel instead of 11. So there is no more than three different Wi-Fi networks operate in close proximity to one another.
When Bluetooth and Wi-Fi operating in the same area, the single 22MHz wide Wi-Fi channels cover the same frequency space as 22 of the 79 BT channels which are 1 MHz wide. So Bluetooth frequency frequently lies on Wi-Fi frequency band and interference occurred according to the signal strength.[17]
4.2.3 Frequency hopping of Bluetooth device:
When there is interference in a channel, Bluetooth device try to omit this problem by hopping to the next channel and until it do not find a non-interfering channel. When it do this process on Asynchronous Connection Less (ACL) link, the data throughput become worse and its become lesser and lesser. And when it do this process on Synchronous Connection Oriented (SCO) links, sometimes packets may lost like voice packets. The reasons for that in SCO BT device do not use Automatic Repeat Request (ARQ).
On the other hand when Wi-Fi device meet interference from a BT transmission, it automatically slow its transmission rate. It will spend more time to transmit packet than before. In Wi-Fi data or packet do not lost but throughput rate slow down to an intolerable level.
4.2.4 Calculating the BER when Bluetooth and Wi-Fi interference each other:
Normally a Bluetooth or WPAN network operates in Wi-Fi or WLAN network and interference was occupied and ultimate result degradation of quality of service like voice and data. Here we see how they hamper their throughput. Here Bluetooth power is 0dBm, and Wi-Fi power is varying and it is fluctuate between -20 to 20 dBm.[18]
Tab 6: BER to power ratio reliance for 10,000 bits
Figure 11: Wi-Fi and Bluetooth to power ratio reliance
Here the power carrier frequency of Wi-Fi is 200 Hz. Most of the time Bluetooth do not interference with Wi-Fi at low power which indicate either Class 2 or Class 3 device. But when the power of Wi-Fi is increasing, then there are some bit error happens in Bluetooth transmission. On the other hand, Wi-Fi data transmission has some bit error when the power lies between -14 to -4 dBm. Normally Bluetooth devices use less power than Wi-Fi. In real world scenario when there is no bit error in Wi-Fi but Bluetooth has some. So Bluetooth is less defiant to the interference in real world conditions area.[18]
4.2.5 Different technique to mitigate the interference:
Most of the time Bluetooth and Wi-Fi work together in a same place, so there is a big possibility to be interference. So minimizing the interference level some mechanism was proposed by Bluetooth SIG and the IEEE 802.15 working group. There is two mechanisms. 1. Collaborating mechanisms
2. Non-collaborating mechanisms
Collaborating mechanisms: Collaborating coexistence mechanism is defined as a mechanism where WPAN and WLAN communicate and collaborate to lessen the communal interference.
4.2.5.1Some collaborating technique is mentioned below: [17]
4.2.5.1.1 TDMA (Time Division Multiple Accesses): This technique allows Bluetooth and Wi-Fi to alternate transmission. Here BT can not support SCO links but it support piconet.
4.2.5.1.2 MEHTA (the Hebrew world for “conductor”): In this technique there is a packet transmission request. It grants acquiescence to broadcast a packet based on parameters including signal strength and the difference between Bluetooth center frequencies and 802.11. SCO link is supported by this system.
4.2.5.1.3 Deterministic frequency nulling: This technique used in congestion with MEHTA. In this process 1 MHz wide null is inserted in the 22 MHz- wide 802.11 carrier accompanies with the current BT center frequency.
4.2.5.2Non-collaborative mechanisms:
In non-collaborative mechanism, there is no communication between WPAN and WLAN. Some non-collaborative mechanisms are mentioned below:[17]
4.2.5.2.1Adaptive packet selection and scheduling: In this process a Bluetooth Media Access Control (MAC) level enhancement use a frequency usage table to amass statistics on channels that face interference. By using packet scheduling algorithms, this table can be access. When a channel is good for transmission then a schedule transmission took place.
4.2.5.2.2 Adaptive frequency hopping: In adaptive frequency hopping channels are classified and alter the hopping sequence regularly to avoid channels which commit the most interference.
4.3 Advanced positioning techniques:
Though there are two methods for tracking like direct and indirect tracking. Direct method is too much accurate than indirect method and its need programming in Bluetooth devices. For advanced technique we need high accuracy and need to consider all the factors which affect the accuracy like humidity, presence of movement of people, temperature, multipath, diffraction, scattering and reflection. Recently most common methods are RSS-based location fingerprinting. In fingerprinting there are some predefined location dependant characteristics. By adjusting those characteristics it take one of the predefined models. Fingerprinting has two stages: online and offline stages. In offline stage survey is done inside the whole area. In this survey it measures the signal strength of neighbouring BSs and gathers the information of positioning coordinates. In online stage several positioning algorithm is being used. Those algorithms are created on the basis of previously measure signal strength at different points and recently measure signal strength.
Few methods are describing below in brief:
► Probabilistic methods: Sometime probability density function method work well than RSSI method for estimating the position. RSSI method is really questionable for the golden receive power range. By combining the transmit power level with RSSI, precision level become well. By using probability density function method, it is possible to get the accuracy level up to 2m and standard deviation of 1.2m. [19]
►K-nearest neighbour averaging: In this algorithm the smallest root mean square error is being used. Here K location entries which was previously reserved in location database and run time unknown location is compared with smallest root mean square error. We can get the final value by averaging the coordinate of the k-th location.
►Smallest M-vertex Polygon (SMP): In location database there is adequate information about the distance from any access point to any candidate points. M-candidate locations can measure the distance from every access point by using the database. At least one candidate location has M-vertex polygon to each access point. Smallest polygon has the shortest perimeter. Averaging the coordinates of the vertices of least polygons, give the finishing estimated location.
Chapter 5: Conclusion and Future work:
5.1Future work:
In near future in a single mobile unit, there will be combination of different wireless access technology. The future devices must support basic cellular connection and also some other extra feature like GPS connection, Wi-Fi and Bluetooth. If it is possible to use all the technology in a single mobile unit then it is sure it will provide great accuracy and precision. So in near future there will be a lot of research will take pace in this sector.
Already some works have been done on combination method. In [20], demonstration systems which use three dissimilar location sensing methods like GPS, 802.11 WLAN and Bluetooth. Here, simple cell identification method is being used for Bluetooth positioning. In future there will be some system which will combine all the individual result and store the information is a database and using different classical algorithms provide very accurate result. For higher accuracy, TN, KNN and SMP, the SELFLOC algorithms can use in combined.
5.2 Conclusion:
Bluetooth tracking is a booming positioning tracking system. Everyday the Bluetooth enabled devices is increasing and this technology will be prevailing all over the world and it has great market value in near future. So most of the renounced companies are trying to improve this tracking system for human convenience. Some tracking algorithms do not work properly and had some limitation. So those technique need to improve. RSSI is commonly used tracking system but it is incompatible for location system. Like RSSI, transfer power level also incompatible for position estimation. Link quality totally depends on Bluetooth class of the mobile device. If we measure the LQ in AP side, most of the time Bluetooth class is unknown. If we measure the LQ from the mobile host side, fingerprinting is too much dependent on the algorithms of BER-to-LQ which is device specific. Overall LQ has inferior location accuracy. So we need to use any other algorithms. Inquiry based tracking system give some good accuracy which is acceptable.
Here I depict an indoor tracking system based on the receive signal strength measurement. Though I do not measure the RSSI values directly, I am really confused about the accuracy of this system. Triangulation method was used to determine the exact location of the mobile device but the Prx values directly depend on RSSI values, so sometimes we got improper result. It is possible to improve the accuracy if we use the combination methods.
Reference:
[1]. Y. Fukuju, M. Minami, H. Morikawa, and T. Aoyama, "DOLPHIN: An
autonomous indoor positioning system in ubiquitous computing environment," in Proc. IEEE Workshop on Software TEchnologies for Future Embedded Systems, 2003.
[2] P. Bahl and V. N. Padmanabhan, "RADAR: An in-building RF-based user location and tracking system," in Proc. IEEE INFOCOM 2000, Mar. 2000, pp. 775-784
[3] B. Fuglede and F. Topsoe, "Jensen-Shannon divergence and Hilbert space embedding," in International Symposium on Information Theory (ISIT 2004), 2004.
[4] Marc Haase and Matthias Handy, “BlueTrack – Imperceptible Tracking of Bluetooth Devices”, Ubicomp Poster Proceeding, 2004.
[5] Gil Held, “Data Over Wireless Networks Bluetooth, WAP,& Wireless LANs”
McGraw-Hill
[6] Wikipedia, “Bluetooth” last update 25 th of August, retrieved on 22nd of August
[7] Image, “Triangulation method” http://emhain.wit.ie/~p02ac03/analysis&design_files/image007.jpg
[8] Cisco, Americans Headquarter, “Wi-Fi Location-Based Services 4.1 Design Guide”, May 20, 2008
[9] Image, “Time of Arrival”, http://www.ietr.org/IMG/jpg_toa_gsm.jpg
[10] Image, from magazine, http://www.gisdevelopment.net/magazine/middleeast/2006/july-aug/22_2.htm
[11] Y. Zhao, "Standardization of mobile phone positioning for 3G systems," IEEE
Commun. Mag., vol. 40, pp. 108-116, Jul. 2002.
[12] A. Madhavapeddy and A. Tse, "A study of Bluetooth propagation using
accurate indoor location mapping," Ubicom 2005, LNCS 3660, Springer-
Verlag Berlin Heidelberg, pp. 105-122, 2005
[13] A. K. M. M. Hossain and W.-S. Soh, "A comprehensive study of Bluetooth
signal parameters for localization," in The 18th Annual IEEE International
Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC'
07), 2007.
[14] Silke Feldmann, Kyandoghere Kyamakya, Ana Zapater, Zighuo Lue,” An indoor Bluetooth-based positioning system: concept, Implementation and experimental evaluation”, International conference on Wireless networks, 2003
[15] Sheng Zhou and John K. Pollard,”Position Measurement using Bluetooth”,
IEEE Transaction on Consumer Electronics, 2006
[16] Antti Kotanen, Marko Hännikäinen, Helena Leppäkoski, Timo D. Hämäläinen, ” Experiments on Local Positioning with Bluetooth”, IEEE journal 2003
[17] Wi-Fi™ and Bluetooth™ - Interference Issues by hp invent, January 2002.
[18] Jan Mikulka, Stanislav Hanus, “BLUETOOTH AND WI-FI COEXISTENCE MODELING” 2005
[19] K. Wendlandt, M. Berbig, and P. Robertson, "Indoor localization with
probability density functions based on Bluetooth," in Personal, Indoor and
Mobile Radio Communications, 2005. PIMRC 2005. IEEE 16th International
Symposium on, Berlin, Germany, 2005, pp. 2040-2044.
[20] D. Graumann, W. Lara, J. Hightower, and G. Borriello, "Real-world
implementation of the location stack: teh universal location framework," in the
Fifth IEEE Workshop n Mobile Computing Systems & Applications (WMCSA
2003), 2003, pp. 122-128.
+ comments + 3 comments
HND Assignment Help is the #1 Assignment help company offering hnd business management assignments at most affordable prices and 100% plagiarism free content. Contact for hnd business assignments at hndassignmenthelp@gmail.com. We are pioneer in btec hnd assignments since more than 7 years and have provided btec hnd assignment help for students around the world. Login http://assignmenthelpblog.blogspot.co.uk or http://btechndcourse.blogspot.co.uk for more details.
HND Assignment Help
HND Assignment Help
HND Assignment Help UK
HND Assignment Help in UK
HND Assignment Help in London
HND Assignment Help in UK
HND Assignment Help in London
HND Assignment Help in UK
Students who get enrolled in contract law in the United States often find it hard to complete their assignments as well as concentrate on their studies. This is when they look for Contract Law Assignment Help online.
Buy online the best price rechargeable international mobile calls to India from the USA, Canada, Australia, and the UK with the help of amantel.com
Post a Comment