International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-2, Issue-4, April-2015

QUALITY IMPROVEMENT OF TRACKING AND POSITIONING OF

MOBILE SYSTEMS
1
PAVANKUMAR D, 2HR MAHADEVASWAMY
1
Research Scholar, 2Guide, Jain university Bangalore
E-mail: pavidn@[Link], hrmswamy65@[Link]

Abstract- In this initial research phase a novel model for mobile node localization and tracking has been developed for
Wireless Interfaced Networks for Next Generation communication. The precise location tracking or the localization of a
mobile user can play a vital role in Line of Sight (LS) as well as Non-Line of Sight (NLOS) communication paradigm which
is of course a key component of future oriented interfaced communication networks (Wireless Sensor networks and Wireless
Mobile Communication systems). To ensure optimal performance in terms of higher data communication rate, minimal data
drop, higher throughout and energy etc, the optimization in node localization plays a vital role. There are a number of factors
that might cause degradation of signals and communication linkage in the wireless mobile communication. In this initial
research phase, i have emphasized over localization or mobile node tracking in sensor based Wireless Mobile network, also
called Wireless Interface network

Keywords- Non-Line of Sight, Wireless Sensor Network, RSSI, Mobile Node, Betien.

I. INTRODUCTION network betien one node to another or betien one

mobile node to the other base station. In case of
Wireless Interface Network has great potential to be Interface communication, the mobile users are mobile
employed in next generation communication system while one terminal remains static as base station in
for broadband utilities, Indoor communication LOS or NLOS communication environment. In
systems, industrial applications etc. In major NLOS proposed research work, i have modeled our system
based communication scenarios, due to varied signal to function with LOS as ill as NLOS scenarios. More
degradation factors the radio signals gets varied precisely to evaluate the robust performance, i using
resulting into the inevitable requirements of optimal NLOS environment to estimate performance.
Received Signal Strength Indicator (RSSI)
calibration. Taking into consideration of the recent II. IMPLEMENTATION OF VARIOUS
researches for wireless mobile communication system LOCALIZATION SCHEMES.
node localization and tracking, especially in case of
NLOS communication paradigms, the RSSI factor There are a number of approaches for node
has been considered primarily. To strengthen the localization and as per research proposal and
performance and effectiveness, the Gray Prediction uniqueness of the proposed research initiatives, here
approaches, Dynamic triangular (DTN) location in this research a number of paradigms have been
approach etc have been advocated a lot. Thus, explored at first so as to ensure an optimal solution
considering these all motivations in this research for location estimation in mobile wireless
phase, i have employed a grey prediction algorithm communication scenario. In major applications, and
for RSSI estimation in NLOS wireless environment. especially in case of wireless mobile communication
Here, for specific case of Interfaced wireless with interface network, the localization of source
communication with mobile users in NLOS terminal becomes inevitable so as to accomplish QoS
environment, i have employed the Dynamic communication and efficient data transmission. In
Triangulation approach that gives precise results. this presented research work a number of localization
schemes have been considered for performing
Here, it must be noted, that in this research phase, i localization in varied circumstances and respective
emphasize in putting the foundation of localization efficiency has been evaluated. Ultimately a novel
and mobile node tracking. As per research proposal, i RSSI based weight predicted localization paradigm
supposed to enhance the overall localization and for mobile node localization has been developed for
tracking system using varied optimization and next generation interfaced wireless network. In this
localization approaches. These all research modelling research work, we have developed the two –
and development would be disused in ascending dimensional (2-D) node localization scheme. Taking
research phase. into consideration of non-linear behaviors in the
network conditions, here non-linear as well as linear
Theoretically, in general, the mobile node localization schemes have been developed. Initially the non-linear
or tracking is accomplished be estimating distance approaches have been employed such as maximum
parameters existing with wireless radio link or likelihood ad non-linear least square approach for

Quality Improvement of Tracking And Positioning of Mobile Systems

49
International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-2, Issue-4, April-2015

localization which was further enhanced with losses, in this research model it has been assumed that
linearization and respective algorithms were the source node transmits signal at t=0 and the signal
evaluated for performance. is received at nth receiver at time . Thus the inter-
relationship between distance and time can be
In the initial phase of research work, here retrieved simply by
conventional Maximum likelihood scheme and non-
linear least square algorithm have been applied to
perform localization. Later considering the possible
enhancements and optimistic goal to schemes were
explored for localization. To make conventional
Linear Least Square scheme more effective the
generalization was done using weight factor
incorporation, where weight linear least square
outperformed convectional linear least square
approach. In this work, to examine the effectiveness
of the proposed systems, Cramer Rao Lower Bound
was also calculated and the respective performances
of all localization paradigms were evaluated. Here the
cramer rao bound was calculated for the variance for
all the unbiased localization paradigms and respective
performance evaluation was done. In this research
work received signal strength indicator approach and
time of arrival paradigm have been employed for
localization of nodes. In the initial phase time of
arrival has been considered for exploring the best
approach for localization, where in the second step
RSSI potential has been employed for mobile node
localization and positioning.

The novelty of this research work exist in the fact that

there are multiple transceivers in this research work
which certain circular equations are formed resulting
into more precise and accurate localization. This is
because of the fact that the doing so would facilitate
the conversion of noisy time of arrival into certain
circular equations which could be employed for
localization. Majority of existing systems have
employed the two time or arrival circles for
localization which given two possible position of
node that can’t be stated as a novel scheme. The set
of circular equations retrieved in the proposed system
can be employed for processing with optimization
schemes.

A brief of the developed research model can be

understood by observing the following expressions
and discussions. Consider = [, ] represents certain
location of a node and similarly = [ ,
] represents the available information or the In the expressed equations, the first term can be
coordinate of certain nth sensor or node n, = observed as the independent function of the location
1,2,3…, , where ≥ 3. Here in this implemented parameter, and therefore optimizing the above
research model minimal 3 nodes have been taken into equation could result into the equivalence of the
consideration. Thus, the respective distance between results of the optimization of the second term and in
certain passive source or nth sensor transceiver or such a way the maximum likelihood could be
receiver is given by, which can be estimated by obtained as
following expression.

Without taking into consideration of any generalized

Quality Improvement of Tracking And Positioning of Mobile Systems

50
International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-2, Issue-4, April-2015

i)Linear Schemes for Localization probability density function of the localization. In

In this research work, the fundamental concept of case of zero-mean Gaussian distribution, it can be
utilizing linear localization schemes was to transform stated that the optimization or maximization would be
the nonlinear expressions of time of arrival in correspond of a weighted form of non-linear least
localization scheme into equivalent linear equations square which is in fact maximum likelihood. To
having zero-mean disturbances with errors in range facilitate the maximization, we have taken into
measurement as negligible. Now, after transformation consideration of a logarithmic version, given by
the optimization cost function becomes a unimodel
that can assure to get global optimized solution for
precise localization. Thus, in this research phase,
three approaches called, linear least square, enhanced
weighted linear least square and subspace estimator In the expressed equations, the first term can be
have been employed for localization. Here the observed as the independent function of the location
novelty is implementation is the consideration of parameter, and therefore optimizing the above
weighted linear least square which is the enhanced equation could result into the equivalence of the
weighted form of linear least square scheme of results of the optimization of the second term and in
localization. Here it has been found the weighted such a way the maximum likelihood could be
linear least square can achieve better precise and obtained.
accurate localization if the mean and variance of the
linear equations are computed with weight factor. The III. IMPLEMENTATION OF CRAMER RAO
last scheme employed in this research phase was BOUND
subspace estimator that relates the location vector x
with the squared pair wise distances existing between In this research work, the prime objective of
the source node and the receiver in wireless network employing Cramer Rao Lower Bound is to generate
and here the localization has been accomplished the corresponding Fisher information matrix where
using eigenvalue decomposition scheme. the diagonal elements of the Fisher information
matrix inverse are supposed to be the minimal
ii)Implementation of Linear Least Square for achievable variance values. The generic process for
Localization: estimating Cramer Rao lower bound can be stated in
The linear least square approach converts the cost the following three steps:
function optimization equations into linear form of x 1. Estimate the second order derivatives of the
and then using conventional least square the logarithm of the probability distribution function
localization is accomplished.

iii)Implementation of Nonlinear Approaches for

Localization
In non-linear approach of localization it has been
tried to achieve the location of source node using
conventional non-linear least square algorithm and
maximum likelihood approach of localization
estimation. Here it has not been assured to
accomplished global convergence because of its
multimodal cost optimization function. Furthermore,
here it was explored with Non-Linear least Square
because of its efficiency with unknown or unavailable
noise information. To explore further enhancement
the weighted version Non-linear least square
algorithm which is nothing else but the Maximum
likelihood was developed and employed by using
noise variance. The developed system exhibited more
optimal results to achieve CRLB.
iv)Implementation of Maximum Likelihood based
Localization
Consider a scenario with known error distribution;
the maximum likelihood scheme optimizes the

Quality Improvement of Tracking And Positioning of Mobile Systems

51
International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-2, Issue-4, April-2015

Phase-3
Fig [Link]-Linear Least Square, Maximum
Likelihood, Gauss Newton-Non-Linear Space,
IV. RESULT Steepest Descent Non-linear least square scheme,
Steepest Descent –Maximum Likelihood based
In this research phase the entire models developed Localization.
and accomplished analysis was carried out in 6 steps.
These phases of implementation and evaluation are as
follows:
Phase-1
Fig 1- Performance Analysis for Newton-Ramphson,
Gauss-Netwon and Steepest Descent local search
optimization based node Localization

Phase-4
Fig 4-Performance Analysis with LLS, Non-Linear L,
Weighted LLS, Two-step WLS, ML, SubSpace and
Cramer Rao Lower Bound Analysis

Phase-2
Fig 2- Performance Analysis for Newton-Ramphson,
Gauss-Netwon and Steepest Descent local search
optimization based node Localization.
Phase-5
Fig 5-Localization Performance Analysis with Linear
Least Square, Weighted Linear Least Square, Two-
step Weighted Least Square, Subspace estimation
based localization analysis.

Quality Improvement of Tracking And Positioning of Mobile Systems

52
International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-2, Issue-4, April-2015

Phase-6 it is because their evaluation becomes biased for

Fig 6 -Maximum Likelihood, Gauss Newton and higher disturbances and noise situations.
Steepest Descent based local search optimization
based Maximum Likelihood, Subspace Estimation, REFERENCES
Linear Least Square, Non-linear Least Square, Gauss
Newton and Steepest Descent based Non-linear least [1] Farmani, M.; Moradi, H.; Asadpour, M., "A hybrid
square based localization, enhanced weighted Linear localization approach in wireless sensor networks using a
mobile beacon and inter-node communication," Cyber
Least Square and Two-step weighted Linear Least Technology in Automation, Control, and Intelligent Systems
Square based localization schemes and its (CYBER), 2012 IEEE International Conference on , vol.,
performance analysis with Cramer Rao Lower no., pp.269,274, 27-31 May 2012
Bound. [2] Kumar, Sudhir; Tiwari, Shriman Narayan; Hegde, Rajesh
M., "3-D mobile node localization using constrained volume
optimizationover ad-hoc sensor networks," Communications
(NCC), 2014 Twentieth National Conference on , vol., no.,
pp.1,6, Feb. 28 2014-March 2 2014
[3] Mihaylova, L.; Angelova, D.; Bull, D.R.; Canagarajah, N.,
"Localization of Mobile Nodes in Wireless Networks with
Correlated in Time Measurement Noise," Mobile
Computing, IEEE Transactions on , vol.10, no.1, pp.44,53,
Jan. 2011
[4] Zhong Zhou; Zheng Peng; Jun-Hong Cui; Zhijie Shi;
Bagtzoglou, AC., "Scalable Localization with Mobility
Prediction for Underwater Sensor Networks," Mobile
Computing, IEEE Transactions on , vol.10, no.3,
pp.335,348, March 2011
[5] Amini, Arghavan; Vaghefi, Reza Monir; de la Garza, Jesus
M.; Buehrer, [Link], "GPS-free cooperative mobile
tracking with the application in vehicular networks,"
Positioning, Navigation and Communication (WPNC), 2014
11th Workshop on , vol., no., pp.1,6, 12-13 March 2014
[6] Priyadharshini, C.; Rubini, K.T., "Integration of route
lifetime prediction algorithm and particle swarm
optimization algorithm for selecting reliable route in
MANET," Computing, Communication and Applications
(ICCCA), 2012 International Conference on , vol., no.,
pp.1,6, 22-24 Feb. 2012
[7] Miles, J.; Muknahallipatna, S.; Kubichek, R.F.; McInroy, J.;
Muralidhara, H., "Use of radio propagation maps in a single
moving beacon assisted localization in MANETs,"
Computing, Networking and Communications (ICNC),
2014 International Conference on , vol., no., pp.871,877, 3-
6 Feb. 2014
[8] Xiuming Zhu; Pei-Chi Huang; Song Han; Mok, AK.; Deji
Chen; Nixon, M., "RoamingHART: A Collaborative
Localization System on WirelessHART," Real-Time and
CONCLUSION Embedded Technology and Applications Symposium
(RTAS), 2012 IEEE 18th , vol., no., pp.241,250, 16-19
April 2012
In this research work, we have found that the mean
[9] Yuehu Liu; Hao Yu; Bin Chen; Yubin Xu; Zhihui Li; Yu
square error for localization in maximum likelihood
Fang, "Improving Monte Carlo Localization algorithm using
estimator scheme has exhibited better performance genetic algorithm in mobile WSNs," Geoinformatics
and can be considered as the optimal solution to (GEOINFORMATICS), 2012 20th International Conference
achieve cramer rao lower bound in minimal iterations on , vol., no., pp.1,5, 15-17 June 2012
with higher signal to noise ratio outcomes. In this [10] Wenqiang Guo; Yubo Deng; Yi Yang; Heshun Ouyang;
research model in the last phase, the implementation Lian Li, "A novel outdoor localization with cellular
of two-step weighted linear least square scheme has topology in opportunistic networks," Automatic Control and
Artificial Intelligence (ACAI 2012), International
accomplished better response at higher signal to noise Conference on , vol., no., pp.239,242, 3-5 March 2012
ratio. And other paradigms of localization have
exhibited sub-optimal accuracy in localization. Here [11] Chia-Ho Ou, "A Localization Scheme for Wireless Sensor
Networks Using Mobile Anchors With Directional
it has also been observed that the two optimal Antennas," Sensors Journal, IEEE , vol.11, no.7,
localization schemes maximum likelihood and two pp.1607,1616, July 2011
step weighted linear least square has exhibited less [12] Xiuming Zhu; Pei-Chi Huang; Song Han; Mok, AK.; Deji
mean square error in localization which was found to Chen; Nixon, M., "RoamingHART: A Collaborative
be less than Cramer Rao bound with SNR ≤ 0dB and Localization System on WirelessHART," Real-Time and
Embedded Technology and Applications Symposium

Quality Improvement of Tracking And Positioning of Mobile Systems

53
International Journal of Advances in Electronics and Computer Science, ISSN: 2393-2835 Volume-2, Issue-4, April-2015
(RTAS), 2012 IEEE 18th , vol., no., pp.241,250, 16-19 NETWORKVishal Garg1, Mukul Jhamb2Department of
April 2012. Computer Science & Engineering, Kurukshetra University.
[13] International Journal of Advanced Research in Computer [14] Guibin Zhu, Qiuhua Li, Peng Quan and Jiuzhi Ye, “A GPS-
and Communication EngineeringVol. 2, Issue 6, June2013 free localization scheme for wireless sensor networks”, in
Copyright to IJARCCE [Link] TRACKING the Proc. Communication Technology (ICCT), 12th IEEE
LOCATION of MOBILENODE inWIRELESS SENSOR International Conference on vol., no., pp.401-404, 11-14
Nov. 2010.



Quality Improvement of Tracking And Positioning of Mobile Systems

54