Literature Survey Word Document - Copy 2

Localization performs a widespread function in purposes such as routing, goal monitoring etc.
which are all area dependent. The goal of Localization is to decide the bodily coordinates
(positional estimates) of a crew of sensor nodes in a terrestrial WSN surroundings Localization can
be considered as an unconstrained optimization problem. Initial strategy is to localize the sensor
nodes the use of a range-free localization approach referred to as Mobile Anchor Positioning
(MAP).In MAP method, the anchors cross via the community and broadcast their place as beacon
packets whilst on the move.The sensor nodes after amassing ample beacon packets
from cellular anchors are in a position to calculate their locations.The
estimated areas received are via regular mathematical method and to enhance the localization
accuracy, positive evolutionary and meta-heuristic algorithms have to be proposed.
Usage: .The localization is necessary in most applications,such as surroundings sensing,search and
rescue,and geographical routing and tracking; the function of every node need to be known.
GPS disadvantage:- GPS-equipped nodes with regarded coordinates in order to estimate their
positions. Most of these works think about the static beacon. While GPS gives fairly correct area
information, it may also no longer be viable for most randomly deployed WSNs. Firstly, GPS accessible
for WSNs are very costly, exceeding the fee of a sensor node.
Mobile Anchor Positioning
Beacon points
Possible locations of the sensor node
Beacon packets
1. Classification of localization techniques

Localization Techniques
Range Range Free

Based
Time DV Hop
RSSI Time Of fferen Angle of Centriod APIT
Arrival Arriv Arrival
Range based approach
Range based totally method makes use of the distance size to estimate the node’s
bodily location. Some of the techniques are: Re- ceived Signal Strength Indicator
(RSSI), Time of Arrival (ToA), Time Difference of Arrival (TDoA), Angle of
Arrival (AoA)[22]. Except RSSI, all the techniques want more hardware in
phrases of Antenna for transmitting and receiving the RF signals.
Received Signal Strength Indicator (RSSI)

In RSSI, acquired sign facts is used to calculate the location information. It is a
easy approach, however much less accurate. All these algorithms, makes use of
Anchor node or Reference node which is provisioned with GPS device. With
reference to Anchor node, final node’s place can be estimated.
Using Received sign strength, the distance between the transmitter and receiver
can be estimated as follows:
RSSI = −10n log10 d + A (1)
In Equation (1), A is RSSI value, n is the course loss exponent (value -2.7 to 4.3, varies for
indoor and out of doors environments) and d is the distance.
RSSI is one of the frequently used techniques for the indoor localization. Received Signal
Strength (RSS) is the approximate sign energy electricity obtained at the receiver power. It
is measured in phrases of decibel-milliwatts (dBm). The RSS is used to compute the
distance between Transmitter and Receiver. RSSI is an indicator of relative dimension of
acquired signal electricity which has arbitrary cost usually defined by means of the dealer
of the node.
A Radio module from Digi International, XBee has RSSI indicator in the shape of LED, whose
mild depth signifies the sign electricity .RSSI for XBee would be in the vary of -40dBm to -
100 dBm. Greater terrible price signifies a weaker signal [20]. Ath- eros WiFi chipset makes
use of RSSI values between zero and 60 Cisco makes use of a vary between zero and a
hundred Measured sign electricity in indoor environments would be various due
surroundings interference. Elnah- rawy et al. [21], have mentioned the simple obstacles of
the use of signal electricity in indoor surroundings was once discussed.
In [20], the authors have employed a hybrid goal localization scheme for cooperative 3D
wi-fi sensor networks. Their proposed work combines distance (RSSI information) and
attitude measurements (Angle of Arrival information). Based on attitude and distance
dimension they have derived a novel non-convex esti- mator.
Time of Arrival (TOA) and Time Difference of Arrival (TDOA) These techniques estimate the
distance primarily based on the arrival time of radio sign at the receiver side. Calculated
ToA is extended with a acknowledged propagation velocity offers the measured distance
between supply and receivers. For this approach, each sender and receiver want to be
synchronized. Time Difference of Arrival estimates the distance with the aid of measuring
the signal’s arrival time distinction between unknown and beacon node.
Angle of Arrival (AOA)

This method requires antenna array to estimate the attitude at which the sign transmitted
sign reaches the receiver and calculates the time distinction of arrival. AOA offers correct
estimation if the transmitter and receiver distance is small, it desires complex hardware
and cautious calibrations. If the distance between the transmitter and receiver increases,
small accuracy error would lead to large error in real function estimation.
Range free localization
Bulusu et al noted that embedding GPS carrier to every node in sensor community is no longer a
favored method for the following reasons [12]: (i) Cost-Sensor networks consists of large variety of
nodes deployed in a region. In such scenario, facilitating every node with GPS unit would be a dearer
solution. (ii) Limited battery power. (iii) For some applications, nodes may also be deployed in
indoors; reception of indicators from satellite tv for pc would be affected through climatic condition.
(iv) Inaccurate- GPS does now not provide very accu fee spatial coordinate, about 10-20 m of
inaccurate spatial information would be suggested [2], which is no longer proper for applications
which want specific geographic position.
Requirements of a localization scheme are:
1) Distributed –Nodes in the community compute their positions on their very own with the assist of
anchor nodes and instant neighbouring nodes.
2) Reduce the quantity of node to node communication.
3) Handle intermittent community conditions.
Classification of Range free localization algorithms are given below:
APIT: He et al [5] have made large contributions to the localization algorithms. They have proposed
vicinity based totally vary free localization algorithm referred as APIT (Anchor based totally Point in
Triangulation). This scheme has excessive powered transmitters and GPS. Node whose vicinity
records acquired via GPS is termed as Anchor node. This scheme divides the goal environment into
triangular areas between beacon nodes.
APIT algorithm works primarily based on the following procedure: a) Bea- con exchange- Exchanging
region facts from Anchors b) PIT Testing (Point in Triangulation) -Node chooses three anchors from
which a beacon used to be obtained and exams whether or not it is inner the triangle fashioned by
means of connecting these three anchor nodes. c) APIT Aggregation-This check is iteratively carried
out till all nodes are included for checking out or the required accuracy is attained.
a) Centre of Gravity Computation: APIT calculates the Centre of Gravity of the intersection of all of
the triangles in which a node resides to decide its estimated position.
APIT algorithm works nicely even all through irregular radio patterns and random placement of
nodes.
Adhoc Positioning System (APS)
APS is a distributed, makes use of hop by way of hop method. It is a mixture of GPS and Distance
Vector Method to supply the approximate spatial location. At least 3nodes (referred as landmark
nodes) are GPs enabled. Node exchanges their function records with its on the spot neighbors and
its on hand landmarks role estimate as well. Two strategies of hop by using hop distance
propagation are explored:
1) DV-Hop Propagation Method
2) DV-Distance Propagation Method
DV-Hop method: This technique is comparable to Distance Vector (DV) routing where every node
continues a desk and exchanges its information. Example of DV Hop is depicted in Fig 5. Assume
that node A, B and C are anchor nodes and X is unknown node.
1) Survey of Wireless Indoor Positioning Techniques and Systems
This paper affords an overview of wi-fi indoor positioning options and tries to classify
three exceptional methods and systems like triangulation, scene analysis, and proximity
are analyzed..Gives information indoor positioning strategies and systems. overall
performance, size standards, are mentioned and tradeoffs are observed
2) Received signal strength based room level accuracy indoor

localization method
This paper room-level accuracy is achieved by indoor localization of a

target, using already deployed WLAN infrastructure, but without actually
interfering with the network. Instead, for a limited number of points of a
map, a n-tuple of RSSI readings is recorded (n being the number of access
points (AP) in range).The target device only performs passive scans of the
existing access points(AP). A simple covariance based method is presented
which can achieve room level accuracy and simple to implement. The
quantity of AP sensed in a room is giant adequate (which is the case with contemporary
residential or workplace buildings) then lengthy time period variant in one or two AP’s
sign does no longer have an effect on the right decision
3)RSSI based Localization Scheme in Wireless Sensor Networks:

This paper focus on RSSI based localization in WSN which will show how to decrease
location error and enhance accuracy with the aid of use of a range of fashions and
techniques. After that we will focal point on how to make an algorithm scalable and
enhance the power effectivity by means of offering authentication and key management
for indoor and out of doors environment. RSSI based totally place extensively used
propagation model for distance estimation and one-of-a-kind methods like
trilateration, multilateration are used for position estimation.
4) Error Minimization in Localization of Wireless Sensor Networks
using Genetic Algorithm
In this paper Genetic algorithm is established to be advantageous in looking out a
answer house and subsequently can be modeled for the localization trouble in Wireless
Sensor Network (WSN). The method used in this paper for localization makes use of
two phases. The first section makes use of a common vary free localization algorithm
primarily based on Mobile anchor to estimate the place of a sensor node roughly. The 2d
segment is a put up optimization segment that makes use of Genetic algorithm which
will increase the accuracy of localization From the experimental effects it can be
considered that LMA-GA offers very excessive localization accuracy. From the
common error of MAP and LMA-GA strategies got from the experimentation, it is
located that the share of localization error has reduced by means of 84%. LMA-GA
additionally does now not require high priced hardware as in vary based totally
strategies and it does now not require flooding of messages as in standard vary free
algorithms. Thus it can be concluded that LMA-GA is a non-expensive and an
environment friendly method that offers very excessive localization accuracy.
5)Performance Analysis of Diverse Environment based on RSSI

Localization Algorithms in WSNs
In this paper,to minimum localization error proposed localization algorithm based on
RSSI for wireless sensor networks. It uses 3 Anchor nodes in triangle placement to
estimate Target node position using trilateration method and conducted the experiments
in outdoor and indoor environments to derive the path loss model.It shows that by
varying the anchor node density on localization error with minimum number of anchor
nodes thereby improve the accuracy of network The result displays that the outdoor
environment has better distance estimation than an indoor environment and achieved
minimum localization error by varying the density of anchor node
6) Localization Estimation Using Artificial Intelligence Technique in

Wireless Sensor Networks
This paper presents an intelligent localization scheme, which is an artificial neural

network (ANN) based localization scheme used to estimate the position of the unknown
nodes. The RSSI values of the beacons received from three anchor nodes are used as
the input to the proposed ANN for estimating the position of the mobile node. Only the
RSSI values of the beacons received from the anchor nodes are used for estimating the
position of the mobile node. It is recommended that the RSSI values of the beacon
request signal received by the anchor nodes also be utilized as it might result in better
localization accuracy. For Further work applications where online training is done, the
LM (Levenberg-Marquardt Algorithm) algorithm is recommended to be used for online
training the ANN as it is fast and has low memory requirements compared to other
learning algorithms. If training time is not important and memory requirement is not a
factor, the BR (Bayesian regularization) algorithm can be used.
7) An Effective Cuckoo Search Algorithm for Node

Localization in Wireless Sensor Network
In this paper, in order to reduce computational complexity and to increasing
optimizatation (modified Cuckoo Search) MCS algorithm is used. MCS algorithm uses
modified step size to enable the population to approach global optimal solution rapidly,
and the fitness of each solution is employed to build mutation probability to avoid local
convergence and along this to prevent the energy consumption caused by insignificant
search, the approach restricts the population in the certain range. Many experiments
have been performed to study the several factors like anchor density, node density and
communication range on the MCS algorithm with respect to average localization error
and localization success ratio for practical WSN applications .It is seen that when
compared with standard CS and PSO algorithm, the modified CS algorithm performs
better in terms of reducing average localization error and increasing the convergence
rate, which is favorable to reduce computational consumption and thus increases the
lifetime of sensor nodes. In future to apply modified CS algorithm, an experiment
system has to design to test for practical applications.
8) An Efficient Node Localization Approach with RSSI for Randomly

Deployed Wireless Sensor Networks
In this paper in order to enhance localization accuracy a route planning algorithm based
totally on grid scan which is the whole traverse in sensor subject is proposed. The
proposed method can grant the deployment uniformly of digital beacon nodes amongst
the sensor fields and the decrease computational complexity In order to enhance the
localization accuracy, the weighting feature is built based totally on the distance
between the nodes. Furthermore, an iterative multilateration algorithm is additionally
proposed to keep away from limit in the localization accuracy.
9) Bat Optimization Algorithm with MAP for Localization in Wireless

Sensor Networks
In this paper range-free localization methodology MAP-M&N doesn't involve the usage
of any hardware. In MAP-M&N methodology, the messages containing location data
square measure being shared among the nodes within the field and it doesn't need
flooding and complex computation for localization. the share of localized nodes is high
that indicates that MAP-M&N methodology is suitable for localization purpose. Since it
doesn't provide fine-grained accuracy, a meta-heuristic optimisation approach
particularly, bat optimisation algorithmic program has been applied over the results of
mobile anchor to scale back the share of localization error. The proposed meta-
heuristic method for localization is Bat Optimization algorithm with Mobile Anchor
Positioning (BOA-MAP) is utilized over the effects of MAP-M&N (Mobile Anchor
Positioning with Mobile Anchor & Neighbor) to reduce the share of localization error,
there by improving the localization accuracy. BOA-MAP algorithm significantly brings
down the RMSE based localization error by 98.53% when compared to MAP. meta-
heuristic BOA-MAP approach is better than using MAP-M&N Also, hybrid
optimization algorithms can be applied in order to minimize the localization error
further. Moreover bat optimization algorithm can also be combined with simulated
annealing and similar hybridization can be applied to further minimize the percentage of
localization error.
10) A Wireless Sensor Network with Soft Computing

Localization Techniques for Track Cycling Applications
In this paper two soft computing techniques for WSN distance estimation were
presented for outdoor and indoor velodromes and used to to determine the distance
between the bicycle position moving on the cycle track and the coach. The first method
was based on ANFI (Neural Fuzzy Inference System) and second approach ANN
(Artificial Neural Network adopted and hybridized with heuristic algorithms such as
PSO, GSA and BSA hybridized i.e Particle Swarm Optimization(PSO),Gravitational
Search Algorithm (GSA), and Back tracking Search Algorithm (BSA) to select the
optimum number of neurons to select the optimum learning rate of ANN which
improves distance error between the bicycle and the coach. When comparing each
technique with ANN the hybrid GSA-ANN was more convenient than other in terms of
distance estimation accuracy. Therefore, GSA-ANN is suitable in both indoor and
outdoor environments and can be applied to any static or mobile WSN node. The
limitation of this study lies in implementing the hybrid GSA-ANN in real-time requires
a considerable amount of memory. The huge computation memory comes at the expense
of limited memory size and processor speed of the microcontrollers. In such a case, the
mobile node (bicycle) requires a microcontroller with a high speed and large memory
size such as an Arduino Due which leads to high power consumption, large size and
extra weight, which are considered critical issues in bicycle sensor nodes. The high
level of power consumption will reduces the battery life of the sensor nodes..
11) GWO-LPWSN: Grey Wolf Optimization Algorithm for Node

Localization Problem in Wireless Sensor Networks
This paper utilized GWO (Grey Wolf Optimization Algorithm ) for localization
problem and provided better results in finding location of unknown nodes. The
numerical computation results such as convergence rate (minimum computation time)
and success rate (maximum number of localized nodes)of proposed GWO algorithms
are noted and it is compared with other variants such as PSO and MBA algorithms
particle swarm optimization (PSO) and modified bat algorithm (MBA). The results of
GWO algorithm are compared with other meta heuristics approaches and thereby it
achieves better performance with respect to maximum number of localized nodes.
Further, this algorithm can be tested in movable node networks such as Mobile ad hoc
networks (MANET). In future, GWO algorithm can integrate with other variants of meta
heuristic algorithms to form a hybrid algorithm for efficient move in convergence and
diversity over the identification of maximum number of unknown node positions.
12) Localization in Wireless Sensor Networks: A Survey on

Algorithms, Measurement Techniques, Applications and Challenges
In this paper, localization techniques are classified into two broad categories:
centralized and distributed. Most of the applications in WSNs demand distributed
localization method as they are more convenient for online monitoring than centralized
system. Centralized and distributed localization system is further subdivided into range
based and range free method. Range based methods are more accurate than range free
methods. However,accuracy in range based methods are obtained with the cost of
additional hardware, which in turn consumes more energy and in many applications is
not suitable at all. Thus, range free methods are more desirable in many applications in
WSNs but obtaining higher accuracy in adverse channel conditions and environments
with different obstacles remains a future challenge for range free localization methods.
To improve the accuracy and robustness of the overall system, fusing the information
from different positioning systems with different physical principles lead to the
development of hybrid data fusion category in order to get optimal localization
accuracy.
13) A Survey of Localization using RSSI and TDoA Techniques in

Wireless Sensor Network: System Architecture
In this paper localization is carried by using two techniques RSSI and TDoA. In RSSI
the hardware is simple and low cost and advantage of TDoA over ToA is that the former
does require knowledge of the time at a transmission at the non line of sight (NLOS) and
requires synchronization between base stations in order to improve the accuracy of
localization. . In addition to RSSI andTDoA, three predominant algorithms are used
which are trilateration, multilateration and triangulation, and additionally the wi-fi is
applied. The main contribution of this paper is to design and development of WSN
localisation system, which are connected to wireless (long range )LoRa transceiver
module. The proposed design can be developed for small low-cost low-power LoRa
tracking device
14) Maximizing Lifetime of Wireless Sensor Networks Based on Whale

Optimization Algorithm
In this paper inorder to increase the lifetime of wireless sensor networks

(WSNs) Whale Optimization Algorithm (WOA) called WOTC is used.In
which discrete binary position of each Whale is calculated and represented in
binary format. It minimizes the number of active nodes with low energy
consumption within the selected nodes without losing the network coverage
and connectivity characteristics .Topology control of WSNs is a technique to
optimize the connections between nodes to reduce the interference between
them, save energy and extend network lifetime. In this paper proposed an
algorithm based on Whale Optimization Algorithm (WOA) called WOTC,
the paper provides a discrete version of the WOA, where the position of each
Whale is calculate and represented in a binary format. The proposed fitness
function is designed to consider two main target; a minimization in numbers
of active nodes, and low energy consumption within these nodes to
overcome challenges that face topology control to prolong the WSNs
lifetime. WOTC is compared with A3 topology depending on the number of
neighbors and their energies for active nodes, use a graph traversal function
to ensure that all nodes which selected in network are covered in the best
topology selection
15) Salp Swarm Algorithm for Node Localization in Wireless Sensor

Networks
In this paper, a node localization scheme is proposed based totally on a latest bio
stimulated algorithm referred to as Salp Swarm Algorithm (SSA).The proposed
algorithm has been evaluated and in contrast to 4 universal optimization algorithms,
specifically PSO, BOA, FA, and GWO,namely particle swarm
optimization(PSO),Butterfly optimization algorithm (BOA), firefly algorithm (FA), and
gray wolf optimizer (GWO) in phrases of localization accuracy, computing time, and
numerous localized nodes .The proposed algorithm has been carried out and validated in
different WSN deployments the usage of different numbers of goal nodes and anchor
nodes. In the future work, the proposed strategy can be hybridized with different
algorithm to decrease the localization error.
16)An Enhanced Hybrid 3D Localization Algorithm Based on APIT and DV-

Hop
In this paper, we propose an enhanced 3-D localization algorithm to solve the two
problems in the common localization algorithms, which is the low localization accuracy
of beacon nodes and the low coverage rate in sparse environment. Firstly, an enhanced
APIT algorithm and a 3D DV-Hop algorithm are designed to improve the robustness of
each algorithm in different situations. Then, a hybrid 3D localization algorithm is designed
with combining the advantages of APIT algorithm and DV-Hop algorithm, which can
further improve the localization accuracy of the beacon nodes in dense environments by
reducing the triangles in the triangle interior point test (PIT) and selecting good triangles.
In addition, the algorithm can combine the advantages of APIT algorithm and DV-Hop
algorithm localization algorithm to calculate the unknown node coordinates, and also
improve the location coverage of the beacon nodes in sparse environment. Simulation
results show that the proposed hybrid algorithm can effectively improve the localization
accuracy of beacon nodes in the dense environment and the location coverage of beacon
nodes in sparse environment.
17)Recent achievements in sensor localization algorithms
Wireless Sensors Network represents an attractive research area for the last years and it
is expected to continue for the next years because of its wide applications in our daily
life. Localization is one of the challenges that face researchers inextending the usage of
wireless sensors due to the high cost and indoor problems of Global Positioning System
GPS. Over the years a lot of algorithms have been applied to replace GPS either depending
on point to point distance Range based algorithms or depending on network information
as in Range Free algorithms. Both of them introduce reasonable cost but still suffer from
the problem of network coverage and look for the best management of power. The lack
of accuracy of all traditional algorithms and the massive development in optimization
techniques were the motivation for the researchers to find an accurate solution to the
localization problem using soft computing techniques. This research gives an overview of
Wireless Sensors Network applications and challenges and focuses on Localization
problem. It classifies the algorithms used to solve the problem and identified the issues of
them which represent a promising as well as challenging areas for future work. It
introduces the recent remarkable techniques of soft computing used in improving the
main issue of the localization which is accuracy. Localization is still a promising area to
research, develop and optimize.
18)Bat algorithm based on simulated annealing and Gaussian
perturbations
The BA is a new type of stochastic optimization techniques for global optimization. In this paper, we have
introduced simulated annealing into the standard bat algorithm and then use Gaussian perturbations to perturb the
solutions in the population, which can enhance the BA, while retaining a certain degree of ‘elitism.’ As the search
iterations con tinue, the temperature is gradually reduced, and consequently, the probability of accepting poor
solutions is gradually reduced. As a result, the overall convergence is enhanced, and the proposed SAGBA retains
the standard BA’s characteristics (e.g., simplicity and easy implement), but also speed up the global convergence
and improves its accuracy. The numerical results using 20 diverse test functions show that the proposed algorithm
(SAGBA) is better than the other two algorithms, which has been confirmed by statistical testing.
It is worth pointing out that we observed from our simulations that the performance will improve if random-
ness is reduced gradually in the right amount. Here, we have achieved this by using simulated annealing. It can be
expected that simulated annealing can also be used to hybridize with other algorithms. In addition, the diversity
of the solutions is controlled by using Gaussian perturbations, and thus, it may be useful to investigate how dif-
ferent perturbations and probability distributions may affect the convergence of an algorithm. It is highly needed
to compare various probability distributions and their role in randomizing stochastic algorithms.
Furthermore, though these preliminary results are very promising, there is still room for improvement. In the future
studies, the comparisons of SAGBA with other algorithms should be carried out. It will also be fruitful to apply
SAGBA to multi-objective optimization. In addition, it will be extremely useful to apply the proposed algorithm to
large-scale real-world design problems in engineering.
19)DV-max Hop: A Fast and Accurate Range-Free Localization

Algorithm for Anisotropic Wireless Networks
In this paper, we proposed an enhancement to the pioneer distance vector or DV-
Hop algorithm to compute localization estimation for anisotropic networks. The
recently proposed algorithms for anisotropic networks provide good estimation,
however, they are complex due to communication and computational overheads.
In addition, they might be unfeasible or un- desirable for low-cost low-power
location dependent protocols and applications. Our proposed scheme, DV-
maxHop, was able to reach good accuracy quickly by utilizing simpler, practical and
proven DV-Hop algorithm. We utilized our previously developed Pymote
simulation framework to generate several isotropic and anisotropic networks
based on desired shape, connectivity, neighbor density and communication range.
The comprehensive simulation results provided statistical and vi- sual analysis of
range-free localization algorithms. The simulation results of our scheme showed
better localization accuracy in different types of anisotropic networks, with much
faster convergence and low overheads, when compared to other existing
algorithms.
We introduced the formulation and simulation to show that our scheme can achieve multiple objectives of
accuracy and efficiency by selecting the optimal MaxHop. We are further improving the Multi-objective
Optimization solution. As a future work, we plan to study the effect of noise and signal fading during the
localization process. We are also planning to study and validate other new and up-to-date localization algorithms
including mobile anchor based schemes.
20)A Survey of Localization using RSSI and TDoA Techniques in

Wireless Sensor Network: System Architecture
This paper discussed review of localization techniques, localization algorithms and wireless transceiver
technologies. The paper also proposed the design and development of a wireless sensor node (WSN) for localisation
system. The WSN used the LoRa transceiver module for communication between dumb and beacon nodes. We
proposed the localization system using three beacon nodes and one dumb node. Future work is the node will be
developed for small size and low power consumption with self-power devices
21)Salp Swarm Algorithm for Node Localization in Wireless Sensor

Networks
Accurate node localization concerns many applications that adopt WSNs. In this paper, a node localization
algorithm has been proposed based on a novel bioinspired algorithm called Salp Swarm Algorithm (SSA) which
handled the node localization problem as an optimization problem. The proposed algorithm has been
implemented and validated in different WSN deployments using different numbers of target nodes and anchor
nodes. Moreover, the proposed algorithm has been evaluated and compared to four well-known optimization
algorithms, namely PSO, BOA, FA, and GWO, in terms of localization accuracy, computing time, and several
localized nodes. The obtained simulation results have proved the superiority of the proposed algorithm compared
to the other localization algorithms regarding the different performance metrics. In the future work, the
proposed approach can be hybridized with other algorithm to reduce the localization error.
 22)Resolution Limit of Positioning Error for Range-Free
Localization Schemes
The resolution limit of localization error is an important metric for range-free localization schemes. In this article, we have analyzed the
problem of the existing methods and proposed two algorithms to compute the resolution limit in ideal channel and in shadow-fading
channel, respectively. We have found that the derived resolution limit by the existing methods tend to be smaller than the real value. In
order to derive the real resolution limit in ideal channel, the important features of maximum movable distance and the concept of critical point
and corresponding propositions are given. Based on the propositions, we have designed the algorithm to calculate the real resolution limit in
ideal channel. In order to obtain the resolution limit in shadow-fading channel, we first turn random variables into definite values at each
time instant, then compute the instant resolution limit based on the proposed algorithm in ideal channel, and finally, calculate the
expectation of resolution limit. Simulation results have demonstrated the correctness of our derived resolution limit. The results also show that
the resolution limits or lower bounds derived by the existing methods are much smaller than the real values.
22)A Novel Range-Free Localization Scheme Based on Anchor Pairs Condition

Decision in Wireless Sensor Networks
In this paper, we proposed a novel range-free localization scheme, LAPCD, for anisotropic wireless sensor networks. In the
scheme, the geometric constraint based method is utilized. According to the relative location relationship among the anchor pairs and
unknown nodes, the anchor pairs are categorized into three categories, which are optimal, sub-optimal and unavailable, respectively.
For the first category, the expected distance from the unknown node to anchor is derived mathematically. For the second
category, the average hop progress of the routing paths between anchor pairs are utilized to estimate the distances. With the
proposed scheme employed, the impact on distance estimation caused by the detours existing in the routing path can be decreased,
hence improving the localization accuracy. Our simulation results indicate that the proposed scheme outperforms other algorithms for
the investigated node densities and the number of anchors, where the localization accuracy improves by an average of 31.4%
compared to the DV-maxHop. In future, we will manage to utilize the unavailable anchor pairs so as to further improve the
localization accuracy and the energy efficiency, while strengthening the theoretic analyses on the localization accuracy of nodes
Different Types of Localization Techniques
Measurement Scheme Accuracy Special Requirement
TOA Moderate Synchronization, dense beacons
TDOA High Synchronization, dense beacons

Range- basedAOA High Directional antenna
RSSI Moderate No
Neighborhood Low No
Range-free Area estimation Moderate Dense Beacons
Hop count Moderate Dense Beacons
FingerprintingRSSI High No
Localization strategies in WSNs
As mentioned in the previous chapter, for numerous applications of WSN, the localization of the nodes is a
fundamental information which must be associated to the sensor measurements. As a bridge between the physical
world and the digital world, WSNs are widely used to deal with sensitive information in many fields. Application
scenarios of WSNs include military, industrial, household, medical, marine and other fields, especially in natural
disasters monitoring, early warning, rescuing and other emergency situations. For example, by a smart dust
network, suspended nodes in the air space can detect pressure, temperature and other information of different
positions to monitor the quality of the atmosphere. Sensor nodes buried under the bed at different depths can
collect temperature, pressure and other data to observe the activity of the glacier [20]. Sensor nodes in birds’ nests
can help users to further research the living habits of birds [21]. In above mentioned applications, all collected
information is based on the accurate location of sensor nodes. Therefore, localization is one of the basic and core
technologies in WSNs [22].
In this chapter, the localization principle and process are discussed. A classification of localization strategies in
WSNs is provided, and some typical approaches are revisited and detailed.
19
2.3. EXTRA MODULES AIDED APPROACHES 20
2.1 Localization process

The objective of a localization process is to determine the position of an object of interest through a specific
system. According to the application specific requirements, appropriate algorithms can be chosen among existing
localization techniques.
For applications which require only coarse localization, the position is obtained directly, by determining the
proximity to an anchor or by using hop count methods or finger printing [23]. It should be noticed that coarse
localization methods are simple means to provide an initial estimate for a more accurate localization method.
For applications requiring a better accuracy, the localization methods include two steps: distance measurement
and position calculation. As illustrated in Figure 2.1, the first block estimates distance or angles of arrival (AOA)
from the received signal information: time of arrival (TOA), time difference of arrival (TDOA), received signal
strength indication (RSSI) and other available features. The second stage processes the distance and angular
information and estimates the position through several positioning methods associated with optimization
approaches.
All the relevant measurement techniques mentioned above: TOA, TDOA, DOA, RSSI, and localization
algorithms: Triangulation, Trilateration, Multilateration will be described further.
Signal Measured Estimated

Information Distances Angles of Position
Arrival
Measuremen Location
t Techniques Calculation
Figure 2.1: Two steps of localization process
2.2 Localization strategies overview in WSNs

Numerous methods have been proposed for localization in WSNs. From a hardware perspective, the
positioning strategies can be divided into two categories: extra modules
aided approaches and extra modules free approaches, as illustrated in Figure 2.2.
A detailed explanation of the two strategies will be addressed in the subsequent section.
GPS Method
Cellular Network Method

Extra Modules
Aided Approaches
Infrared Method
Ultrasonic Method
Localization Micro Inertial Navigation Method

Strategies
Range Free
Connectivity
Methods
Extra Modules Geometry

Free Approaches
Energy
AOA
Range Based
TOA/TDOA
Methods
RSSI
Figure 2.2: Application classification for WSNs
2.3 Extra modules aided approaches

For this class of methods, a specific hardware is dedicated to the localization. This approach includes GPS
method, cellular network method, infrared method, ultrasonic wave method, micro inertial navigation method, to
name just a few [24]. In the following, we will briefly address the above-mentioned localization techniques and
discuss the combination of these techniques and WSNs from a practical perspective.
2.3.1 GPS method

The Global Positioning System, abbreviated as GPS, is a satellite based positioning system developed by the
American military authorities in the 1970s. In the following four decades, GPS has been widely used in both
military and civilian fields for navigation, communications and monitoring. Similarly, the European Union and
Russia have build their position systems, named Galileo and Glonass Systems respectively. In recent years, China
has establishing the BeiDou Navigation Satellite System for pursuing a higher precision and extensive applications
[25].
In order to accomplish the task of tracking the displacement of a sensor, GPS modules can be integrated into
the nodes. However, there exist some drawbacks and limitations in it. In some environments, like underground
parking, underwater sites or indoor environments, GPS receiver can not communicate with the satellites [26].
Consequently, it is unfeasible to use GPS in these environments. Besides, high energy consumption is also a
drawback in employing GPS modules. Therefore, using GPS for localization in WSNs has some fundamental
limitations, which make us steering to other techniques.
2.3.2 Cellular network method

A cellular network is a communication network where the last link is wireless and used for mobile
communication. In that case it is an extra module free method because the cellular modem is also used for
communication.
The Global System for Mobile Communication is abbreviated as GSM, is a standard developed by the
European Telecommunications Standards Institute (ETSI) to describe the protocols for second-generation digital
cellular networks [27]. In the recent years, with the advent of third-generation and fourth-generation (4G)
technologies, mobile communications went up to a higher level [28]. In cellular wireless location systems, a
plurality of base stations receive signals from a mobile terminal simultaneously, and then the cellular network
accomplishes the localization process based on the measured parameters.
In WSNs, the sensor nodes can be equipped with the relevant modules and localized by cellular network .
However, there are many limitations and drawbacks in this method. Most importantly, it provides an unsatisfactory
localization precision, ranging from the order of tens of meters to hundreds of meters. In addition, the cellular
network module
is energy consuming, which contradicts the low cost objectives in WSNs. Therefore, there are still many multi-
aspects and challenging issues in this direction to be fixed.
2.3.3 Infrared method

Infrared ray is another kind of electromagnetic wave, whose wavelength ranges from the microwave to the visible
light wavelength. Due to its thermal effect and highly penetrating ability, infrared ray is extensively used in
medical treatment and industrial detection and control. With the advance of microelectronics and optical fiber
communication, infrared transducer became a low cost device which can be employed for object detection and
vehicle tracking [29].
When propagating in the space, the infrared ray would undergo deviations like reflection, refraction,
scattering, interference and absorption [30]. Due to these effects, it is nearly impossible to use infrared transducer
into WSNs for localization when the environment is complex. On the contrary, the sensor nodes with infrared
transducer in WSNs can be used with success for detection. To sum up, the use of infrared technology for localisation
in WSNs remains an open issue.
2.3.4 Ultrasonic wave method

Ultrasonic wave is a part of sound waves, whose frequency is beyond 20kHz [31]. Distinctly different from
the ordinary sound waves, the ultrasonic wave has the following characteristics: superior directionality, longer
transmission range, strong reflectivity and penetrability [32]. Due to the above-described features, the ultrasonic
wave is introduced successfully into engineering and health-care fields.
The principle of ultrasonic wave based localization can be addressed concisely as follows. The receiver
estimates the time of arrival or the difference of time of arrival and measures the propagation time t. The distance
between the sender and receiver can be estimated as (2.1):
s = vwt (2.1)
where vw is the velocity of the ultrasonic wave in the relevant transmission medium.
Since the ultrasonic wave suffers from interference and distortion in harsh environment with a variety of
obstructions, it is impractical to estimate displacement by means
of ultrasonic wave based localization techniques extensively [32]. On the contrary, the ultrasonic wave can
propagate steadily in the water, which enables the ultrasonic wave based localization techniques to some specific
application fields. In the marine environment monitoring, ultrasonic wave based localization technique is
considered as an advisable method. Many researchers are trying to integrate ultrasonic module into sensor nodes
and exploring the application in marine monitoring and navigation.
2.3.5 Micro inertial navigation method

Mainly based upon the acceleration sensor, direction sensor and gyroscope sensor, micro inertial navigation is
identified as an accurate navigation system estimating the movement parameters from the data sensed by the above
mentioned devices. A distinct feature is its independent estimation, namely, the localization process uses only
internal equipment without the help of outside systems [33]. This technique is originally employed for missile
guidance, and then aircraft and submarine navigation.
To overcome the limitations in GPS system and improve the localization accuracy, the researchers introduced
the micro inertial navigation method into WSNs. In the process of movement, the acceleration sensor obtains the
acceleration of the node movement and the orientation sensor obtains the node posture instantaneously. After the
acceleration data and the direction angles of the node are acquired, the displacement of the node’s movement can
be calculated through integral calculation. Micro inertial navigation system provides a satisfying accuracy in short
term but can exhibit deviation in long period. So, it can be used in addition to other localization systems (GPS or
cellular networks) when they become periodically unavailable. Despite that micro inertial navigation can gain a
higher precision in short range, a heavy energy consumption is inevitable. Therefore, it is imperative to make a
trade-off in practical applications to meet the diverse requirements [34].
2.4 Extra modules free approaches

Compared to the extra modules aided approaches, extra modules free approaches require no additional
components to assist the localization process. Namely, rather than supported by external localization systems and
internal mounted components, the ex-
2.4. EXTRA MODULES FREE APPROACHES 25
tra modules free approaches carry out the localization task merely by its own network parameters.
Extra modules free approaches are typically divided into two aspects: range free methods and range based
methods [35]. Compared to range-free localization, range- based localization provides higher precision. There are
many range-based localization techniques, such as those based on the measurement of TOA [36, 37], AOA [38,
39], TDOA [40, 41], RSSI [42] and so on.
RSSI-based algorithms have the following characteristics: low power consumption, simple hardware but high
sensitivity to environment. RSSI value heavily depends on the propagation channel. Signal reflection, multipath
propagation, noise and signal scattering have great influence on the received RSSI. Therefore, in practical
applications, establishing an accurate channel model to deduce the distance from the received RSSI value is crucial
to the performance of localization algorithms.
An in depth explanation for the two types of methods will be addressed in the following section.
2.4.1 Range free methods

Contrary to range-based algorithms, range-free algorithms accomplish localization through network and devices
features, such as network connectivity graph, device power consumption and conservation, geometric relationship
and many more, instead of ranging the distance between target and anchor nodes. In the sequel, several classical
localization algorithms: distance vector hop (DV hop) [43], approximate point-in-triangulation test (APIT) [44] and
centroid algorithm [45], will be revisited.
2.4.1.1 DV hop
Inspired by the classical distance vector routing scheme, DV-Hop algorithm is proposed in [46]. It involves
three steps in the localization process as follows:
1. In the initial step, an information table is built for each node according to the
landmark broadcast location and hop data. The data package is exchanged
between node and its neighbors. The table is denoted as (xi, yi, hi), where (xi, yi) is
the coordinates of the ith landmark and hi is the minimum hop count value from
the ith landmark to the target node who maintains this table.
Figure 2.3: Error analysis on DV hop count measurement
2. Secondly, the averaged size for one hop is estimated based on the distance cal-
culated between a landmark and other landmarks. The averaged size is
estimated by:
Sizei = Σ √ Σ , j
2i (2.2)
(xi − h
xji) + (yi − yj)
2
where (xj, yj) is the position of landmark j. In this step, the target node calculates the distance based on the
hop size value and the hop count number from at least three landmarks.
3. Finally, when the distance values are obtained, the relevant positioning method,
such as mutilateration, trilateration, linear least squares (LLS), non-linear least
squares (NLS) and so on [47], can be employed to find the position.
As shown in Figure 2.3, the dotted line denotes the actual distance between two nodes, and the solid line indicates
the hop direction and estimated distance by DV hop algorithm. Obviously, the DV-hop algorithm provides a low
accuracy due to the imprecise distance estimation.
A high density nodes deployment can provide a better accuracy. However, owing to its simplicity, this method
can be applied into some rough localization.
2.4.1.2 APIT
APIT is a range free localization algorithm, presented in [48]. The core idea of this method is to associate
Point-In-Triangulation Test (PIT) with area-based scheme to search the most likely target position. PIT is adopted
for narrowing the possible region where the target is located. We assume that many anchor nodes, whose location
is known by other means, are scattered in a wireless sensor networks. As illustrated in Figure 2.4, in every trial,
three anchors are selected to form a triangle and whether the position of the target is in this triangle or not is
decided. This process is repeated until all the triangles are considered. After finishing all the tests, the center point
of intersection area will be regarded as the estimated position. The localization accuracy of APIT depends on the
test number which is directly related with the anchor number. Unavoidably, the power and time consumptions
increase with number of anchors [49].
Figure 2.4: Principle of APIT algorithm

2.4.1.3 Centroid algorithm
Centroid localization algorithms estimate the position via geometric relationship between the landmarks and
the unknown nodes, instead of calculating the corresponding distance. These algorithms are suitable for the
wireless sensor networks with a certain number of landmarks whose positions are recognized by other
complementary schemes. Periodical packets containing position information are broadcasted among the networks.
When the number of received packets exceed a predefined threshold value, a stable transmission link is established
between a node and a landmark. An unknown node will be connected with many landmarks. Assuming that the
number of landmarks is more than 3, a polygon is formed by these landmarks. Then, the centroid position is
considered as the unknown node location.
As illustrated in Figure 2.5, scholars presented centroid localization algorithm based on tetrahedron. A
tetrahedron is limited by four anchor nodes L1, L2, L3, L4, which are connected with the target to be localized.
Then, the centroid of this tetrahedron is considered as the coordinates of the target. In [50], simulations were
performed to compare this method and the classical centroid algorithm. The results indicate that the tetrahedron
algorithm gives a higher accuracy than the traditional method in spite that larger calculation time is required due to
many estimation rounds. Other researchers in [51, 52, 53, 54] tried to improve the centroid algorithm by assigning
weighted factors to each anchors or associating correction schemes to reduce the localization error.
2.4.2 Range based methods

Contrary to the previous methods, range based methods perform position computation after distance
estimation. These methods involves two stages: distance measurement and position calculation.
Many techniques have been employed for distance measurement, for example, the techniques based on time of
arrival (TOA), time difference of arrival (TDOA), angle of arrival (AOA) and received signal strength indicator
(RSSI) [55]. After acquiring the distance measurement, geometric relationship is used to compute the node position
by triangulation, multilateration, trilateration, hyperbolic and so on.
In the next section, these techniques are discussed in detail.
L1
L2
O L3
D
L4
Figure 2.5: Diagram of tetrahedron algorithm
2.4.2.1 Distance measurement techniques

Firstly, we will present the methods that can be used for the distance measurement step. These methods are
based on the measurement of the angle or distance, including TOA, TDOA, RSSI and AOA.
TOA
In TOA and TDOA methods, the distance is ranged from the transmission time between the transmitter and
the receiver [56]. The time of flight (TOF) recorded between two terminals can be is used to estimate the distance
by a simple multiplication by the transmitting velocity. The schemes for measuring elapsed time are classified into
two categories: one-way scheme and two-way scheme.
In one-way scheme, the transmitter sends signal to the receiver and the time delay
for this transmission is measured. The transmitter sends message at time t1 and the receiver gets the message at
time t2. Then a decoding delay or synchronization time tc is required to accomplish the time record.
Therefore, the time delay τ can be written:
τ = (t2 + tc) − t1 (2.3)
The one-way scheme is simpler, but it is essential to synchronize the two terminal clocks to reduce the
errors. High accurate clock synchronization is a challenging task and clock bias results in measurement errors. This
remark explains why two-way scheme is generally privileged.
In two-way scheme, the transmission time between two nodes is measured. The distance is one-half of the
measured time multiplied by the propagation velocity. The timing process of the two-way scheme is shown in
Figure 2.6.
Message
Sender
𝑡
𝑇p
𝑇p
𝑡s1 𝑡s2
Responder
𝑇r
∆
𝑡
𝑡
𝑡r1
𝑡r2
Figure 2.6: Timing process of the two-way scheme
The sender sends message to the responder at time ts1. The responder receives the message at time tr1.
After knowing the processing time Tr, the responder gives a feedback at time tr2 which is received by the sender at
time ts2. A clock bias ∆t exists between the sender and responder of two nodes. The propagation time is denoted as
Tp [57]. From the Figure 2.6 we can write:
tr1 = ts1 + Tp + ∆t (2.4)
tr2 = tr1 + Tr (2.5)

Finally,
ts2 = tr2 + Tp − ∆t (2.6)
Then by subtracting ts2 and ts1, we get the result
ts2 − ts1 = 2Tp + Tr. (2.7)
Then time Tp can be calculated by
ts2 − ts1 − Tr
T = . (2.8)
p
2
As shown in the previous equation, the clock bias ∆t is eliminated by this process.
In two-way scheme, the transmission time is obtained by recording packet receiving and sending times. In
spite that this two-way scheme does not need clock synchronization, inaccurate packet processing time in
terminals results in measurement errors.
An alternative method called TDOA is proposed to measure time difference between two propagation processes
[58]. A typical TDOA system will be discussed in detail in the following.
TDOA
The key concept of TDOA-based localization technique is to determine the location of the source by
evaluating the difference in arrival time of the signal at spatially separated base stations [59]. As shown in Figure
2.7, there are three signal receivers: RX1, RX2, and RX3, whose coordinates are known as (x1, y1), (x2, y2) and (x3,
y3). The objective is to determine the position of the transmitter with unknown coordinates (x, y). The reception
times in RX1, RX2 and RX3 are respectively t1, t2 and t3. This values can be combined to obtain the following
equation [60]:
√ √
√(x − x2)2 + (y − y2)2 − √(x − x1)2 + (y − y1)2 = c × (t2 − (2.9)
t1) (x − x3) + (y − y3) − (x − x1) + (y − y1) = c × (t3 −

2 2 2 2
t 1)
where c is the speed of light.

By solving the above nonlinear equations, the position of TX can be determined.
(𝑥1, 𝑦1 ) RX1 𝑡3 RX3
𝑡1
TX (𝑥3, 𝑦3 )
RX2 𝑡2
(𝑥2, 𝑦2 )
Figure 2.7: Terminals deployment of TDOA method
The main drawback of the TDOA technique is that the reception time difference can be fairly small,
especially in short distance measurement, and the distance estimation is not precise [61]. To overcome this
problem, the electromagnetic waves can be re- placed by acoustic waves. The propagation velocity is much smaller
and thus the time differences are largely increased.
As seen in Figure 2.8, ultrasound/acoustic and RF modules are simultaneously used in transmitter and
receiver. The principle is to measure the time difference between the propagation times of the acoustic and radio
signals [62].
In the initial localization step, the transmitter sends at time t0 the radio signal which is received by the receiver
at time tradio. After a fixed time delay tdelay, the transmitter sends the acoustic signal which is received at time tsound.
Figure 2.9 shows the time delay computation model for this type of TDOA [63]. The two received times can be
written :
d
tradio = t0 + (2.10)
and
vradio
d
tsound = t0 + tdelay + (2.11)
vsoun
d
where d is the distance between the transmitter and the receiver and vradio and vsound
Transmitter Receiver
Acoustic/
Acoustic/ Ultrasound
Ultrasoun module
d module
RF module
RF module
Figure 2.8: Hardware configuration of TDOA
are respectively the transmission velocity of the radio and acoustic signals. Using these two equations, the distance
between the transmitter and receiver can be calculated by:
vradiovsound
—t −t ) (2.12)
d=
(t
vradio − vsound sound radi
o
delay
It must be noticed that the transmitter time to is not present in this equation. More- over, the two times tradio and
tsound are measured in the receiver, and accurate synchronization between the transmitter and receiver is no more
needed.
Since the radio signal propagates far faster than the acoustic wave in free space, the value of vradio − vsound is
approaching to vradio, and tradio is also much smaller than tsound for short distances (i.e. indoor applications). For that
reason, we can write [64]:
t0 ' tradio (2.13)
and
d ' vsound(tsound − tradio − tdelay ) (2.14)

where tsound and tradio are measured at the receiver. There is no need to synchronize the transmitter and receiver.
t0 t𝑑𝑒𝑙𝑎𝑦
Transmitter
RF Acoustic
Receiver
t𝑟𝑎𝑑𝑖𝑜 t𝑠𝑜𝑢𝑛𝑑
Figure 2.9: Time delay computation model for TDOA
RSSI
The RSSI, which denotes the Received signal strength indicator, is a measurement of the receiver signal power. It
is available in most of receivers and can be used for distance measurement as it can be expected that its value
decreases with the distance [65]. Many RSSI based algorithms have been presented for unknown target localization
in wireless sensor networks. To characterize the relationship between the received signal strength and transmission
distance, several path loss models are built based on experimental data. In free space propagation, the relationship
between signal strength and transmission distance is expressed by Friis equation as [66]:
PtGtGrλ2
Pr(d) = , (2.15)
where 4π2dη L
– Pr and Pt are respectively the received and transmit powers,

– Gt and Gr are the antenna gains of the transmitter and receiver,
– L is the system loss,
– d is the radio transmission distance,
– η is the path loss exponent equal to 2 for free space,
– λ is the radio wavelength defined by
c
λ= (2.16)
f
where c is the light speed and f is the signal frequency.

For simplicity, the values of Gt, Gr and L are set to 1. Equation (2.15) is simplified
to:
Ptλ2
Pr(d) = (2.17)
4π2dη
From the relationship between the transmitted and received powers, we can define the path loss PL which
denotes the power attenuation during the propagation [67]:
Pt 2π
PL = =( )2 d η (2.18)
Pr λ
Substituting (2.16) into (2.18), we get:
Pt 2π
PL = =( )2 f 2 d η (2.19)
Pr c
This equation indicates that the pass loss is determined by two factors: radio frequency f and
transmission distance d. Path loss increases with frequency f and distance d.
The path loss exponent η is determined by the transmission environment. In usual environments, the free
space assumption is no longer verified. Multi-path and shadowing have great impact on factor η. A large number
of experiments indicate that the value of η is generally between 2 and 4 [68].
A simplified formula for RSSI computation is proposed in [69]:
Pt(d0)
P = (2.20)
r
dη
where d0 is a reference distance usually equals to one meter. On the study of the received power in the receiver, the
relation between RSSI and distance is interpreted as [70]:
d
) + Xσ (2.21)
Pr(d) = Pr(d0) + 10ηlog(
d
0
where the powers are expressed in dBm and Xσ is zero mean Gaussian distributed random variable whose mean
value is zero. This variable reflects the local variations of the received power due to fading and shadowing [71].
From the above equations, a popular RSSI channel model is presented as [72]:
Pr(dBm) = A(dBm) − 10ηlog(d) + Xσ (2.22)
where Pr is the received signal power, A is the signal power at a distance of one meter. Many RSSI based
localization algorithms are based on this channel model. These algorithms will be presented latter in this chapter.
AOA
Angle of arrival (AOA), which is also called as direction of arrival (DOA), can be used for location estimation
[73]. The AOA technique was firstly designed to estimate the location of objects in radar system, which is widely
applied in military and civilian fields. The receiver with multiple directional antennas measure the angle from the signal
reflected by the target 1. Generally, the angle information is extracted by two means [74].
(1) On receiving the signal, the antenna arrays in the reference station have
different phase informations. The angle of arrival can be calculated from the phase
difference.
(2) The angle of arrival can also be estimated by calculating the signal amplitude
at the main beam.
Algorithms developed by many authors make the direction of arrival estimation to become highly accurate and
able to provide very high resolution results. The first at- tempt to automatically localize signal sources using an
antenna array was proposed by Bartlett, which is referred to in the literature as the shift and sum beamforming
method or Bartlett method. It is based on calculating the power of the beamforming output for all the possible
directions [75]. The other conventional method is known as the Capon algorithm, which adds the constraint of
making the gain of the array unity in the looking direction of arrival and then minimizing the output power in the
other directions [76].
1. In classical radar systems the angle of arrival is determined by a rotary antenna.

2.4.2.2 Location calculation

In the previous section, the methods to estimate the angle or distance, have been discussed. However, these
informations need to be further processed by positioning algorithms to find the coordinates of the object.
The objective of the following sections is to present the algorithms that can be used to calculate the position,
such as Triangulation, Multilateration, Trilateration, linear least squares (LLS), non linear least squares (NLS),
projection onto convex sets (POCS) and semidefinite programming (SDP) [77].
Triangulation
Y T
(x, y)
θ𝑖
BS (x𝑖, y𝑖) X
Figure 2.10: Angle of arrival measurement.
When the angle of arrival is obtained, triangulation algorithm can be used for location estimation. As
illustrated in Figure 2.10, the transmitter T sends signal to the base station i. The angle of arrival θi is given by:
y − yi
tanθi = ( ) (2.23)
x− x
i
where (xi, yi) is the coordinates of the base station i; (x, y) is the coordinates of trans-
mitter T.
In triangulation, at least two base stations are needed for two-dimensional localization. The principle of
triangulation is shown in Figure 2.11. The location of transmitter T can be computed from the two angles θ1 and θ2
by:
T
Y
(x, y)
θ1 θ2
BS1 (x1, y1) X

BS2 (x2, y2)
Figure 2.11: Triangulation
L tan(θ2)
x=
(2.24)
tan(θ2) − tan(θ1)
L tan(θ1)tan(θ2)
y =tan(θ2) − tan(θ1)
where L is the distance between the two base stations, which can be calculated by:
√
L = (x1 − x2)2 + (y1 − y2)2 (2.25)
The accuracy of triangulation relies heavily on the measured angle of arrival. Im- proving the measurement
precision on arrived direction is a way to guarantee a higher accuracy. Meanwhile, employing more base stations
can also enhance the localization performance.
Compared to TOA method, AOA method has the following advantages. Time synchronization is not
required for measuring angle of arrival. The error caused by time measurement inaccuracy is avoidable. Less base
stations are needed to estimate the position in triangulation method. To find the position of one target, AOA needs
two base stations but TOA needs at least three base stations with known position. Moreover, base stations can
measure the arrived angle from the target without the cooperation of the target, which reduces the communication
overhead and makes the localization process less complex.
However, the drawback in AOA method may cause some limitations when it is applied in practical
localization process. When the distance between the target and base station is large, the measured angle value is not
accurate due to the varying transmission characteristics in long path. It is hard to overcome this problem and the
localization performance is reduced. Meanwhile, directional antennas or antenna array in base stations will bring
additional cost for localization system. In view of these features, AOA method is more applicable in radar
localization system.
Multilateration
Multilateration is a popular method for finding the position of a target. In this method, at least three anchor
nodes are needed for 2-D space localization. The equations for multilateration is expressed as:
2 2 2
 1

 (x − x )2 + (y − y 2) = 2
(x − x21) + (y − y2 )1 = d2
 (2.26)
d .
..
. .
.. ..


2 2 2
(x − xN ) + (y − yN ) = N
d
where (x, y) is the coordinates of the reference or unknown nodes, (x1, y1), (x2, y2),
· · · , (xN , yN ) are the coordinates of the N anchors. Then, this non-linear system of equations must be solved
by adequate methods to obtain the unknowns x and y.
In real environment, the distance measured from signal information is inaccurate due to multi-path, reflection,
shadowing and noise impact. Consequently, the position of the target can not be calculated exactly by
multilateration. To find an optimal position,
LLS, NLS and POCS methods have been employed and associated with multilateration. These methods will be elaborated
in the following sections.
Trilateration
(x1, y1) (x3, y3)

(x, y)
Unknown node
Anchor nodes
(x2, y2)
Figure 2.12: Trilateration
As shown in Figure 2.12, when the number of anchors is 3, the multilateration is also called trilateration.
Under minimum anchor configuration, the position can be found from three anchors, if they are not deployed in
straight line. The relationship between the unknown node and three anchor nodes is written by [78]:
(x − x1)2 + (y − y1)2 = d12

(x − x2)2 + (y − y2)2 = d22 (2.27)
(x − x3)2 + (y − y3)2 = d2 3
where (x, y) are the coordinates of the reference or unknown nodes, (x1, y1), (x2, y2), (x3, y3) are the coordinates of
the three anchors. By subtracting the first equation to the others, the system of equations (2.27) can be
transformed into the following matrix
form:
Qx = b (2.28)
where Q is a matrix of dimension 2 × 2, x is the coordinate vector, b is a vector of dimension 2.

2(x1 − x2) 2(y1 − y2)
Q= " # (2.29)
2(x1 − x3) 2(y1 − y3)
" #
x
x= y (2.30)
" #
" # x2 − 2
2 − y 1 + d −2 d
1 x +y
2 2 2 2
2 1
= bb1 2 = x2 − x2 + y2 − y2 + d2 − d2 (2.31)
b 1 3 1 3 3 1
By an adequate choice of the anchor position, we can make sure that matrix Q is invert- ible. So, the calculated
position is:
" #
x
x = Q−1b where x= (2.32)
y
This solution for trilateration can also be written as:
" # " #
x x21− x22+ y2 1− y2 2+ d2 −
2
d2 1
= M , (2.33)
y x21 − x23 + y2 1− y23+ d2 3− d2 1

where M is a matrix of dimension 2 × 2, with the elements defined as follows.
1
M(1, y
1) = ( 1 — y3)/C (2.34)
2
1
M(1, 2) = (y 2
— y1)/C (2.35)
2
1 — x1)/C (2.36)
M(2, 1) = (x3
2
1 — x2)/C (2.37)
M(2, 2) = (x1
C = x1y2 − x2y1 − x1y3 + x3y1 + x2y3 − x3y2 (2.38)

SEARCHING TECHNIQUES
4
1. Introduction
The optimization is a significant tool in analysis of physical system and decision science. This is very much related to our real life
problems. For instance, airline companies does scheduling in order to minimize the cost. Manufacturers aim for maximizing the
efficiency in the design and various operations of their production sequence. Nature optimizes. Physical systems tend to reach a state of
minimum energy. The molecules in any isolated chemical system tend to react with each other till total potential energy of their electrons
is minimized. The rays of light follows path through which the travelling time gets minimized. Optimization came into existence in
1940s, when George Dantzig used some mathematical techniques and generated programs for scheduling timetables for military
application. Today, optimization consists of wide variety of techniques from Artificial Intelligence, fuzzy math, operations research and
computer science. In optimization problems, solutions are needed to be found which are optimal or near-optimal with respect to some
desired aim. Usually, the optimization problems are not solved in just one step, rather a sequence of steps are to be followed for problem
solving. Commonly used steps are to recognize and define problems, construct and to solve models, and evaluate and to implement
solutions. The steps include a) to check for the need of optimization b) Choosing design variables c) Formulating constraints d)
Formulation of objective function e) setting up variable bounds f) choosing an optimization algorithm g) obtaining solution or solutions.
The objective [2] depends on certain characteristics of the system, called variables or unknowns. The aim is to find values of
the variables that will optimize the objective. Generally the variables are restricted, or constrained, in these optimization problems. For
instance, quantities such as electron density in a molecule and the interest rate on a loan can never be negative [2]. The decision variables
may get values from discrete sets, bounded and additional constraints on basic resources, such as capital, labour, or supplies, restricting
the possible alternatives that are taken feasible [3].
Combinatorial optimization problems are concerned with the efficient allocation of limited resources to meet goals. Possible
objectives of a planning or optimization process are either to find an optimal solution of the problem or to find out a solution that is better
than some predefined threshold (the current solution). In this paper the optimization problems which are relevant for modern heuristics
are considered. Two important properties of optimization are considered 1) locality and 2) decomposability. The locality of a problem is
exercised by local search methods, whereas the decomposability is taken care of by recombination-based search methods. There are
mainly two types of classical optimization techniques namely: 1) Single-variable optimization 2) Multi-variable optimization. The multi-
variable optimization is further divided into three parts named as: i) with no constraints ii) with equality constraints iii) with inequality
constraints.
In single-objective optimization (SOO) problem deals with the maximization or minimization of the objective function based
upon a single variable given a constraint or an unconstrained problem. The SOO problems have a single variable in the given objective
function. The function may vary according to the different values of that variable. The function may have i) Relative or Local Minimum
ii) Relative or Local Maximum iii) Absolute or Global Minimum iv) Absolute or Global Maximum. The applications of SOO are related
to less complex real time problems. However, at small levels too optimization is needed.
© 2013, IJARCSSE All Rights Reserved Page | 1014

Punia et al., International Journal of Advanced Research in Computer Science and Software Engineering 3(7),
July - 2013, pp. 1014-1020

In Multi-objective (or multi-criteria or multi-attribute) optimization (MOO), two or more conflicting objectives are simultaneously
optimized with respect to a given set of constraints. Although, in real-world problems many times improvement in one objective leads to
the degradation of another. The applications of MOO can be easily seen in the field of network analysis, aircraft design, bioinformatics,
oil and gas industry, automobile design, product and process design and many more fields. Summing up, the optimization problems have
the following characteristics [3]:
 Availability of different decision alternatives.
 Number of available decision alternatives limited by additional constraints.
 Different effect by each decision alternative on the evaluation
criteria. This paper is categorized into the following three sections:
Section I)Problem Formulation. Section II) Description of various approaches on single objective and multi-objective optimizarion.
Section III) Conclusions.
2. Problem Formulation
The optimization problems may be categorized into Single Variable and Multi-Variable Optimization problems.
2.1 Single objective optimization problem
A single-objective optimization problem (SOOP) has the objective function (f (x’)), which must be minimized or maximized and a
number of constraints (g (x’)). Equation (1) shows the formula of the SOOP in its general form.
minimize f (x’)
s.t. g j(x’) ≥ 0 where ( j = 1,…..,m) (1)
x’ ∈ X ⊂Rn
where x’ is a vector of n decision variables, x’ = (x1, x2,........................,xn)T , and X represents a feasible region.
2.2 Multi-objective optimization problem

Similarly, multi-objective optimization problems (MOOP) with a number of objective functions are shown in equation (2).
(f (x’) = (f1 (x’); f2 (x’),……, fk (x’))T ) can be stated as follows:
minimize f (x’) = (f1 (x’), f2 (x’),……, fk (x’))T
s. t. gj (x’) ≥ 0 where ( j = 1,……,m)

(2)
x’ ∈ X ⊂Rn
The scalar concept of ―optimality‖ cannot be applied directly in the multi-objective model. Thus the notion of Pareto optimality
has to be entertained. Essentially, a vector x∗ ∈ S is said to be Pareto optimal for a multi-objective problem if all other vectors x ∈ S have
a higher value for at least one of the objective functions fi, with i = 1, . . . ,n, or have the same value for all the objective functions.
The formal definitions for multi-objective optimization problem are as following [4]:
 A point x∗ is said to be a weak Pareto optimum or a weak efficient solution for the multi-objective problem if and only
if there is no x ∈ S such that fi(x) < fi(x∗ ) for all i ∈ {1, . . . ,n} . Function f(x) is said to have local or relative minimum
at x as in figure 1.
 A point x∗ is said to be a strict Pareto optimum or a strict efficient solution for the multi-objective problem if and only
if there is no x ∈ S such that fi(x) ≤ fi(x∗ ) for all i ∈ {1, . . . ,n}, with at least one strict inequality. Function f(x) is said
to have relative or local maximum at point x as in figure 1.

Fig 1: Function f(x) having local, global maximum and local, global minimum at point
July x.
- 2013, pp. 1014-1020

3. Techniques for Optimization July - 2013, pp. 1014-1020

There are a few common techniques which are common to both single and multi- objective optimization problems. However there are
some advanced techniques which are applied to multi-objective optimization problems as these problems contain multi-dimensional
objectives to be satisfied. As found in literature (in figure 2) the different optimization techniques can be broadly classified into
following three categories [4]:
• Calculus-based techniques or Numerical methods.
• Enumerative techniques.
• Random techniques.
.
Fig. 2 The different search and optimization techniques [4]
Calculus methods, also known as numerical methods use a set of necessary and sufficient conditions which must be satisfied by the
solution of the optimization problem [4].Numerical methods further divided into direct and indirect methods. Direct search methods
deals with hill climbing in the function space by moving in local gradient direction. Whereas in indirect methods the gradient of the
objective function is set to zero and thus solution is get by solving these set of equations. All the calculus based methods assume strictly
the existence of derivatives and are local in scope too. These constrains limit their application in real-world problems; however in small
class of unimodal problems these can be efficiently used. Enumerative techniques tends to evaluate each and every point of the finite, or
discrete infinite, search space to sought optimal solution[4]. A well-known example of enumerative search technique is dynamic
programming. Thus in order to search each and every point enumerative needs to break down the problems even of moderate size and
complexity into smaller divisions.
Guided random search techniques are based on the concept of enumerative methods only but with the use of additional
information about the search space in order to seek the potential regions faster [4].Guided is further categorized into single-point and
multi-point search, means whether it is searching just with one point or with several points at a given time. For single-point search
technique, simulated annealing is widely used. It uses thermodynamic evolution in order to find states of minimum energy. For multi-
point search, where random choice is used as a tool to guide through a highly explorative search space, genetic algorithms are in trend.
They are basically used assuming that a near-optimal solution will be accepted; given the search space is huge, noisy, multimodal as well
as discontinuous.
3.1 Overview of Genetic Algorithms (GAs)

Gas are efficient, self-adaptable, self-repairable and robust, nature inspired search and optimization tool. GAs performs well in large,
complex and multimodal search space. GAs are modelled based upon the natural genetic principles where the potential solution is
encoded in structures known as chromosomes. These make use of problem or domain dependent knowledge to search potential and
promising areas; also called fitness function, in search space. Each individual or chromosome has a fitness value associated with it, which
describes its goodness compared to other individuals in the current population with respect to the solution. The genetic operators such as
selection, crossover and mutation are also inspired by the nature and are applied to chromosomes in order to yield better and potential
solutions. The sequence of steps taken in a GA to solve any optimization problem is shown in figure 3. GAs are adaptive
computational tools

modelled on the mechanics of nature. These efficiently exploit historical information to guess newly upcoming
July -offspring
2013, pp.with improved
1014-1020
performance. Genetic algorithms are heuristic search methods means it estimates the solution, which can be used for both solving
problems and modelling evolutionary systems.
GAs are preferred when the search space is huge, discontinuous, multi-dimensional, multi-modal and noisy. Whereas the classical
gradient search techniques are applied where there are tight constraints associated with the given problem. GAs have been found to
outperform both the gradient descent method and various forms of random search as literature shows [4, 9, 12, 13].
Fig 3: Basic steps of GA
3.1.1 Genetic Algorithms: Basic Principles and Features-Based on the theory of genetics, the GA encodes each
individual in the population with a chromosome [8]. This encoding represents the parameters for the objective function
being optimized. There are several different techniques for encoding parameters, performing the selection and the
alteration stages of the algorithm. The alteration stage is separated into Crossover and Mutation [5]. A GA uses a highly
abstract version of evolutionary processes to evolve solutions to some given problems. Each GA operates on a population
of given problems. Each GA operates on a population of artificial chromosomes. These are strings in a finite alphabet
(usually binary). Each chromosome represents a solution to a problem and has fitness, a real number, which is a measure
of the quality of the solution to the particular problem. GA starts with a randomly generated population of chromosomes
then carries out a process of fitness-based selection and recombination to produce a successor population, the next
generation [6]. During recombination, selecting parent chromosomes and their genetic material is recombined to produce
offspring chromosomes. Then these are then passed to the successor population. As this process is iterated, a sequence of
successive generations evolves and the average fitness of the chromosomes tends to increase until some stopping
criterion is reached. In this way, a GA ―evolves‖ a best solution to the given problem [6]. The information transfer
is done from one generation to next through the breeding of the trial solutions selected on the basis of their fitness, and
that is why the crossover i defines the defining feature of any GA [8]. The combined effect of crossover and fitness based
selection on any population of strings encoded trial solutions basically helps to enhance the occurrence frequency of
substrings to further convey their decoded trial solution above-average fitness value, at a rate proportional to difference
between the average fitness of all the trial solutions and the average fitness of whole population.
The features of GA which differentiate it from other search methods are as given below:
 GA works with coding of the parameter set, and not with the parameters itself.
 GA starts searching from a population consists of possible solutions, and does not iterate on single solution.
 GA does not use derivatives or other auxiliary information rather uses objective function information.
 GA does not use deterministic rules rather probabilistic transition rules.
A system of nonlinear equations are solved using [5] genetic algorithm techniques. To achieve this propose Gauss- Legendre integration
technique is used first to solve the system of nonlinear equations and then GA is used to find the results without converting the nonlinear
equations to linear equations. The standard coding scheme is used to accomplish the goal. Hence, the parameters of the search problem
are represented as bit strings. The obtained results are confirmed with the results obtained from numerical methods and hence it is shown
that GA is an efficient and effective approach to solve the systems of nonlinear equations that arise in the implementation of Gauss-
Legendre numerical integration.
The paper [9] presented a two-space genetic algorithm and also suggested that there is a general technique to solve minimax
and robust discrete optimization problems. Robust discrete optimization is a technique for structuring the uncertainty in decision-making
process. The goal is to find out a robust solution that has the best worst-case performance over a set of possible scenarios. The proposed
algorithm maintains two populations where the first population represents

solution and the second population represents scenarios. Here the individual in one population is evaluatedJulywith respect
- 2013, pp.to1014-1020
individuals
in the other population. Both the populations evolve simultaneously and they converge to a robust solution and its worst-case scenario.
The minimax problems occur in many domains thus the given algorithm has a wide variety of applications. The GA is quite useful in
solving two-space minimax problem in a wide-case scenario. In this paper to illustrate the potential of two-space GA, a parallel machine
scheduling problem with uncertainty in processing time is solved. For this particular problem, good lower bounds are found and thus
algorithm’s performance is evaluated. The results confirmed that a two-space genetic algorithm is a very suitable technique for robust
and discrete optimization problems. Turkcan [10] used PSGA (Problem search genetic algorithm) for multi-objective optimization. In
multi- objective search, the key issues are guiding the search towards the global Pareto set and maintaining diversity. Here a new fitness
assignment method is proposed to find a uniformly distributed, well-diversified set of solutions that are very close to the global Pareto
set. A multi-objective optimization (MOP) problem formulation is stated as:
Min f(x) = (f1(x), f2(x),….,,fn(x))
s.t. x∈ X
where x is a vector of discrete decision variables and X is a set of feasible solutions. As the objectives conflict with each other, a number
of solutions known as Pareto-optimal or efficient solutions are found. A real world application of solving tool management and
scheduling problems simultaneously in flexible manufacturing systems (FMS) is taken as the problem definition. The proposed fitness
assignment method is taken as a combination of non-dominated sorting based method which is mostly used in multi-objective
optimization literature and aggregation of objectives method which is popular in the literature of operation research. With the use of
PSGA there is no need to do feasibility check hence reducing the significant amount of computation time. The PSGA is applied to single
objective optimization problems. The aim in single objective optimization problems is to find a single solution giving the minimum
objective function. PSGA was proposed by Storer et al.(1992). It is a local search method which provides a new neighbourhood structure
defined in the space of possible problem data perturbations. The proposed method, NSAPV, is a composite measure and gives higher
fitness values to the non-dominating solutions which are closer to the global Pareto-optimal set, have better aggregated objective function
value and less number of neighbors in objective space [9]. A system of linear equations is solved using GA [11] since it is difficult to
describe the solution set of a linear system with infinitely many solutions. A system of linear equations is a collection of two or more
equations with the same set of unknowns. To avoid the disadvantages of solving large system of linear equations such as inversion of
large matrixes, rounding errors, GA is effective and presents an efficient approach to solve the system of linear equations. The coding
scheme used is standard one and the parameters of the search space are represented as bit strings. The solution obtained is similar to
analytical one. The concept of Hydroinformatics is discussed in paper [12]. This field includes water supply management, design of
water distribution networks and systems, water resources, water supply management, watershed water quality management, waste water
management, irrigation scheduling. In paper the evolutionary algorithms have been described as a special case of a population based
approach. The efficiency of EAs in solving optimization problems has been outlined. The ability to handle mixed type of variables, non-
linear constraints, customizing for solving different classes of problems efficiently, and finding multiple trade-off optimal solutions in
the presence of multiple conflicting objectives are some of the commonplace in the field of hydroinformatics which have been discussed
in the paper.
In paper [13] the focus is on the study of evolutionary algorithms for solving multi-objective optimization problems with a
large number of objectives. The proposed algorithm dynamical multi-objective evolutionary algorithm (DMOEA) is compared with the
already existing algorithms for solving multi-objective optimization problems. A new definition of optimality (named as L-optimality) is
also proposed which not only considers the number of improved objective values but also considers the values of improved objective
functions but also takes into account the value of improved objective functions if all the objective functions have the same importance.
The Simulations and comparative experiments indicated that the newly developed algorithm MDMOEA can converge to the true L-
optimal front and it maintains a widely distributed set of solutions. However, even if it is proved that L-optimal solutions are subsets of
Pareto-optimal solutions; even then L-optimal solutions cannot be obtained only by choosing from Pareto-optimal solutions, which
utilize MOEAs based on the Pareto-dominance concept.
The paper [14] presents two new approaches for transforming single-objective problem into a multi-objective problems. The
multi-objectivization approach is used to translate SOOP into MOOP and then applies EMO. The advantages of multi-objectivization
such as reduction of the effect of local optima, increasing the search path to global optimum, or making the problem easier. The two new
multi-objectivization approaches based on addition of new objectives are as: 1) Relaxation of the constraints of the problem. 2) Addition
of noise to the objective value or the decision variables. These new approaches give more freedom to explore and a reduced likelihood of
getting trapped into local optima. The characteristics and effectiveness of the proposed approaches are investigated by comparing the
performance on single-objective problems and multi-objective versions of those same problems. Using numerical examples, it is showed
that the multi-objective versions produced by relaxing constraints are providing good results and the addition of noise can obtain better
solutions when the function considered is multimodal and separable.
The authors [15] proposed a new algorithm for multi-objective optimization called ―Neighborhood Cultivation GA
(NCGA)‖. The recent studies such as SPEA2 or NSGA-II, demonstrated that some of the mechanisms are important such as the
mechanisms of placement in an archive of excellent solutions, assign of fitness, sharing without parameters, selection and reflection the
archived solutions to the search population. Not only NCGA includes these mechanisms but also the neighborhood crossover. NCGA is
compared with SPEA2 and NSGA-II with some test functions and it shows that NCGA is a robust algorithm to find Pareto-optimum
solutions. The effect of neighbourhood crossover is made clear

International Journal of Future Generation Communication and Networking
through the comparison between the case of using neighborhood crossover andVol. the case of using
9, No. normal
8 (2016), pp.crossover
327-344in NCGA. The
authors [16] have suggested a nondominated sorting-based multi-objective EA (MOEA), called nondominated sorting genetic
algorithm II (NSGA-II). Multi-objective evolutionary algorithms (EAs) which use nondominated sorting and sharing have been
criticized mainly for the mentioned reasons such as: 1) computational complexity O ( MN3 ) (where is the number of objectives
and is the population size); 2) the need for specifying a sharing parameter, and 3) nonelitism approach. The proposed algorithm
alleviates all the above three difficulties. Particularly, a fast nondominated sorting approach with O (MN2) computational
complexity is presented. A selection operator is also presented that creates a mating pool by combining the parent and offspring
populations and selecting the best solutions. Simulation results from difficult test problems show that the proposed NSGA-II, in
most of the problems, is able to find much better spread of solutions and better convergence near the true Pareto-optimal front as
compared to Pareto-archived evolution strategy and strength-Pareto EA—the two other elitist MOEAs that pay special attention
to creating a diverse Pareto-optimal front. Furthermore, the definition of dominance is modified in order to solve constrained
multi-objective problems efficiently and effectively. The simulation results obtained from the constrained NSGA-II on a number
of test problems, which includes a five-objective seven-constraint nonlinear problem, are matched up with another constrained
multi-objective optimizer and NSGA-II offered much better performance.
The authors [17] have presented a new distributed genetic algorithm for multi-objective optimization problems. The
proposed approach uses island model with a distributed genetic algorithm and an operation is performed for sharing Pareto-
optimum solutions with the total population. The Pareto-optimum solutions are needed to be derived for designers in multi-
objective optimization problems. Not only the accuracy but also the diversity of the solutions is needed to be high as the Pareto-
optimum solutions are the set of optimum solutions that are in relationship of trade-off. Indexes are introduced that can evaluate
the performance of the algorithm. The indexes taken are population size, error, coefficient of variation and cover rate. These can
be applied the problems that have more than three objectives to be achieved. High accuracy is achieved by the effect of the
distributed population and the high diversity of solutions is achieved by the sharing effect. The numerical examples which have
more than three functions are taken as test problems to examine the effects. In paper [18] proposed a novel parallel hybrid
algorithm which combines multi-objective and single objective genetic algorithm. The results confirmed that this approach
outperforms traditional parallel versions of multi-objective genetic algorithm. This algorithm is proposed as the literature shows
that the majority of the multi-objective genetic algorithms are computationally expensive, thus they are often parallelized. In this
paper the single objective (SOGA) evolutionary algorithm is combined with multi objective evolutionary algorithm (MOGA) in
heterogeneous island model and has it has outperformed the traditional island model. The experiments showed that adding SOGA
with island model can be more effective than adding MOGA island, thus it leads to better utilization of computational resources.
Also in some of the cases it has reduced the need for function evaluation during evolution, and it leads to the reduction in run-
time of the optimizer.
Range free localization techniques are best suited for these requirements of network.
DV-HOP and APIT range free techniques have gained attention in most of research
efforts. So,in this paper
,performance of these techniques is evaluated on the basis of accuracy obtained while

implementeing them in WSN . Detail analysis of localization error is done by considering
different network parameters effecting it directly. Further structured deployment of
anchors is proposed that gave more reliable results.
Recent advancements in semiconductors, communication [10] and networking technologies are driving
pervasive deployment of large scale wireless sensor networks (WSNs). Due to availability of tiny, cheap and
smart sensors and appropriate RF circuitry for data transmission, WSN has become one of the most promising
technologies [3]. WSN is formed using very small but capable sensors that sense, gather and transmit the
information to a large network of such sensors. There are pronounced applications of WSN, like providing
faster warnings from disasters, fine-spun observations of the surrounding environment, performing more
efficient agriculture, all with important economic significance. Similarly, applications in battle field
surveillance, healthcare or habitat and structural monitoring help to increase our daily welfare. Localization is
one of the most important key techniques in WSN because the location knowledge of sensors is helpful in
most of the applications i.e. for coverage, placement, routing, location service, target tracking, and rescue.
Hence, location evaluation is a demanding technical challenge for researchers and academic community [9].
So, it is highly desirable to design cost effective, scalable, efficient and reliable localization mechanisms for
WSNs. These sensor node localization schemes [8] have different features used for different applications.
Localization methods are broadly categorized into two categories: Range based and Range free localization
techniques [11]. Range based techniques [4] use range measurements like received signal strength [12], angle
of arrival and time of arrival for location estimation whereas Range free methods omit the use of range
measurement techniques, instead network topology and connectivity information is used to find the
ISSN: 2233-7857 IJFGCN
Copyright ⓒ 2016 SERSC

Vol. 9, No. 8 (2016), pp. 327-344
location of sensors placed in the network. Therefore,range free techniques overpower the range based
techniques in terms of lesser power consumption, low cost, no extra hardware, less installation and
computational complexity [5]. In return,these features of Range free techniques have attracted a lot of research
efforts in recent years. In this paper we have implemented DV-HOP[1],[2] and APIT[7][6] using MATLABTM
.The results are being analysed for various network parameters under random and uniform placement strategy
of sensor nodes in the network for DV-HOP and APIT individually in sections 3and 4 respectively. After
achieving best suitable network parameters for both the techniques, comparison is further extended by
proposing a structured deployment strategy of anchors .Using second and third order error analysis tools ,the
detail analysis is reported in section 5. Analysis based conclusion is drawn in section 6.
1. Simulation Methodology and Environment

Among the main range free localization techniques-Centroid,DV-HOP,APIT and MDS-MAP. DV-HOP is
one of the techniques that brought a dramatic change in the fashion of localization by considering multi hop
sensor nodes in the network after a simple centroid technique. APIT,then considered area based approach in
localization of sensors in WSN with much improvement in reliability of technique. So, these two techniques
are yet to be explored by researchers.we have implemented these techniques in MATLABTM.Simulations are
done assuming certain parameters and setup environment as shown below in table 1:
Table 1. Simulation Setup
S.n Parameters Specifications

o.
1. Area 100x100 square area
Sensor nodes static

2.
Total nodes 100

3.
Anchor nodes
4. 8
Unknown nodes 92
5.
Placement strategies of Random, Uniform ,Square

Anchor nodes
6. Regular
7. Radio Ranges 25-50m
8. Simulations done 300
As shown in table 1, area of 100x100 sq.units is taken and the total of 100
sensor nodes of which 8 are anchor nods and 92 are unknown nodes are
deployed in a way that they are connected to each other to form a network.
1.1. Network Scenarios

Deployment strategy of these sensor nodes are varied to form 3 types of scenarios- Random deployment,
Uniform deployment, Square Regular deployment.
1.1.1. Random Deployment: Unknown sensor nodes are distributed in random fashion
and also the anchors are placed randomly as shown in figure 1 below.
Copyright ⓒ 2016 SERSC

Vol. 9, No.8, (2016)
100
Red-Anchors,
90 Black-Unknown
80
70
60
y axis
50
40
30
20
10
0 20 40 60 80 100
Figure 1. Random Placement Model
1.1.2. Uniform Deployment: Unknown nodes and anchor nodes are deployed randomly
but the whole area is divided into grids to place the sensors at equal distances. This is
shown in the Figure 2 below:
100
90
80
70
60
Red-Anchors,
y axis
Black -Unknown
50
40
30
20
10
Copyright ⓒ 2016 SERSC 329

Vol. 9, No.8, (2016)

0
0 10 20 30 40 50 60 70 80 90
x axis
Figure 2. Uniform Placement Model
1.1.3. Square Regular Deployment: Unknown nodes are placed randomly at the corners
of grids and the anchors are placed in concentric squares to form a regular network as
shown in Figure 3.

Vol. 9, No.8, (2016)
100
90
80
70
60
Red-Anchors,
y axis
Black-Unknown
50
40
30
20
10
0 20 40 60 80 100
x axis
Figure 3. Square Regular Placement Model
For the presented deployment strategies, DV-HOP and APIT localization methods are implemented in the
software.

Vol. 9, No.8, (2016)
Figure 4. Flow Chart of DV-HOP Localization
Figure 4 shows the working of DV-HOP that after the network is initialised, the packets of information
about the nodes are broadcasted to all nodes and then DVHOP algorithm is performed by calculating hop
counts.

Vol. 9, No.8, (2016)
Working of APIT is shown in the Figure 5 that neighbour information is calculated using RSS values and then test of
triangulation is done to locate the nodes shown below:
Figure 5. Flow Chart of APIT Algorithm

Vol. 9, No.8, (2016)
References
 [ 1] Hui Liu, Student Member, IEEE, Houshang Darabi, Member, IEEE, Pat Banerjee, and
Jing Liu IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART C: APPLICATIONS
AND REVIEWS, VOL. 37, NO. 6, NOVEMBER 2007
 [ 2] Attila Buchman, Claudiu Lung CogInfoCom 2013 • 4th IEEE International Conference
on Cognitive Infocommunications • December 2–5, 2013 , Budapest, Hungary
 [ 3] Hetal P. Mistry,Nital H. Mistry 2015 Fifth International Conference on Advanced
Computing
Copyright & Communication
ⓒ 2016 SERSC Technologie 2327-0659/15 $31.00 © 2015 IEEE DOI334
10.1109/ACCT.2015.105
Vol. 9, No.8, (2016)
 [ 4] Sivakumar.S, R. Venkatesan, R. Venkatesan International Journal of Computer

Applications (0975 – 8887) Volume 43– No.12, April 2012
 [ 5] Z.Mary Livinsa 1 , Dr.S.Jayashri Proceedings of 2013 IEEE Conference on Information
and Communication Technologies (ICT 2013) 978-1-4673-5758-6/13/$31.00 © 2013
IEEE
 [ 6] Shiu Kumar, Seong Min Jeon, Seong Ro Lee The Journal of Korea Information and
Communications Society '14-09 Vol.39C No.09
http://dx.doi.org/10.7840/kics.2014.39C.9.820
 [ 7] Jing Cheng and Linyuan XiaSensors 2016, 16, 1390; doi:10.3390/s16091390
www.mdpi.com/journal/sensors
 [ 8] XihaiZhang,JunlongFang,andFanfengMeng Hindawi Publishing Corporation Journal of Electrical
and Computer Engineering Volume 2016, Article ID 2080854, 11 pages
http://dx.doi.org/10.1155/2016/2080854
 [ 9] S.Sivak umar,Y.P.Geetha International Journal of Applied Engineering Research, ISSN 0973-4562 Vol. 10
No.81 (2015) © Research India Publications; http://www.ripublication.com/ijaer.htm
 [ 10] Sadik Kamel Gharghan , Rosdiadee Nordin , and Mahamod Ismail , Sensors 2016, 16,
1043; doi:10.3390/s16081043 www.mdpi.com/journal/sensors
 [ 11] R.Rajakumar, J.Amudhavel, P.Dhavachelvan, and T.VengattaramanHindawi Journal
of Computer Networks and Communications Volume 2017, Article ID 7348141, 10 pages
https://doi.org/10.1155/2017/7348141
 [ 12] Anup Kumar Paul 1,2,* and Takuro Sato 2 J. Sens. Actuator Netw. 2017, 6, 24;
doi:10.3390/jsan6040024 www.mdpi.com/journal/jsan
 [ 13] Nur Aisyah Azmi, Shafiqa Samsul, Mohd Ismifaizul Mohd Ismail, Rudzidatul Akmam
Dziyauddin Yoshihide Yamada, Mohd Fitri Mohd Yakub 2018 2nd International Conference on
Telematics and Future Generation Networks (TAFGEN)
 [ 14] Mohammed M. Ahmed1,, Essam H. Houssein, Aboul Ella Hassanien, Ayman Taha, and
Ehab HassanienSpringer International Publishing AG 2018 A.E. Hassanien et al. (eds.),
Proceedings of the International Conference on Advanced Intelligent Systems and
Informatics 2017, Advances in Intelligent Systems and Computing 639, DOI 10.1007/978-
3-319-64861-3 68
 [ 15] Huthaifa M. Kanoosh,1 Essam Halim Houssein ,2 and Mazen M. Selim1 Hindawi Journal
of Computer Networks and Communications Volume 2019, Article ID 1028723, 12 pages
https://doi.org/10.1155/2019/1028723
 [16]Lianjun Yi, Ganzhou, Jiangxi,Miaochao Chen, HefeiAn Enhanced Hybrid 3D
Localization Algorithm Based on APIT and DV-Hop iJOE ‒ Vol. 13, No. 9, 2017
https://doi.org/10.3991/ijoe.v13i09.7255
 [17]Eman Saad , Mostafa Elhosseini ,Amira Yassin Haikal “Recent achievements in sensor localization
algorithm” , Alexandria Eng. J. (2018), https://doi.org/10.1016/j.aej.2018.11.008
 [18] Xing-shi He ,Wen-Jing Ding ,Xin-She Yang “Bat algorithm based on simulated
annealing and Gaussian perturbations” DOI :10.1007/s00521-013-1518-4
 [19] Xing-shi He , Wen-Jing Ding , Xin-SheYang “DV-max Hop: A Fast and Accurate
Range-Free Localization Algorithm for Anisotropic Wireless Networks” : DOI
10.1109/TMC.2016.2632715, IEEE Transactions on Mobile Computing
 [20] Huthaifa ,M. Kanoosh Essam Halim,Houssein and Mazen M. Selim “Salp Swarm
Algorithm for Node Localization in Wireless Sensor Networks” Hindawi Journal of
Computer Networks and Communications Volume 2019, Article ID 1028723, 12 pages
https://doi.org/10.1155/2019/1028723
Vol. 9, No.8, (2016)
 [21] L. Gui, B. He, F. Xiao and F. Shu “Resolution Limit of Positioning Error for
Range-Free Localization Schemes” 1937-9234 © 2019 IEEE.
http://www.ieee.org/publications_standards/publications/rights/index.html
DOI:10.1109/JSYST.2019.2927276
[21] Xingcheng Liu, Senior Member, IEEE, Feng Han, Wenjie Ji, Yitong Liu and
Yi Xie “A Novel Range-Free Localization Scheme Based on Anchor Pairs
Condition Decision in Wireless Sensor
Networks”http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
DOI 10.1109/TCOMM.2020.3020553, IEEE Transactions on Communications
[1] Loukas Lazos and Radha Poovendran, “Serloc: Secure range-independent local-
ization for wireless sensor networks,” in Proceedings of the 3rd ACM workshop
on Wireless security. ACM, 2004, pp. 21–30.
[2] Isaac Amundson and Xenofon Koutsoukos, “A survey on localization for
mobile wireless sensor networks,” Mobile entity localization and tracking in
GPS-less environnments, pp. 235–254, 2009.
[3] Nabil M Drawil, Haitham M Amar, and Otman A Basir, “Gps localization
accuracy classification: A context-based approach,” IEEE Transactions on
Intelligent Transportation Systems, vol. 14, no. 1, pp. 262–273, 2013.
[4] Elliott Kaplan and Christopher Hegarty, Understanding GPS: principles and
applications, Artech house, 2005.
[5] Trevor M Ellis, R Barry Bousfield, Lucy A Bissett, Kitman C Dyrting, Geral-
dine SM Luk, ST Tsim, Katharine Sturm-Ramirez, Robert G Webster, Yi Guan,
and JS Malik Peiris, “Investigation of outbreaks of highly pathogenic h5n1
avian influenza in waterfowl and wild birds in hong kong in late 2002,” Avian
Pathol- ogy, vol. 33, no. 5, pp. 492–505, 2004.
[6] Alex Varshavsky, Mike Y Chen, Eyal de Lara, Jon Froehlich, Dirk Haehnel,
Jeffrey Hightower, Anthony LaMarca, Fred Potter, Timothy Sohn, Karen Tang,
et al., “Are gsm phones the solution for localization?,” in Proceedings of the
7th

Vol. 9, No.8, (2016)
IEEE Workshop on Mobile Computing Systems and Applications. IEEE, 2005, pp.
34–42.
[7] Lucia J Zamorano, Lutz Nolte, A Majeed Kadi, and Zhaowei Jiang, “Interac-
tive intraoperative localization using an infrared-based system.,” Neurological
research, vol. 15, no. 5, pp. 290–298, 1993.
[8] David J Schlegel, Douglas P Finkbeiner, and Marc Davis, “Maps of dust
infrared emission for use in estimation of reddening and cosmic microwave
background radiation foregrounds,” The Astrophysical Journal, vol. 500, no. 2,
pp. 525, 1998.
[9] Joseph L Rose and Peter B Nagy, “Ultrasonic waves in solid media,” The
Journal of the Acoustical Society of America, vol. 107, no. 4, pp. 1807–1808,
2000.
[10] MG Silk and KF Bainton, “The propagation in metal tubing of ultrasonic wave
modes equivalent to lamb waves,” Ultrasonics, vol. 17, no. 1, pp. 11–19, 1979.
[11] Oliver J Woodman, “An introduction to inertial navigation,” University of
Cam- bridge, Computer Laboratory, Tech. Rep. UCAMCL-TR-696, vol. 14, pp.
15, 2007.
[12] John E Bortz, “A new mathematical formulation for strapdown inertial naviga-
tion,” IEEE transactions on aerospace and electronic systems, , no. 1, pp. 61–
66, 1971.
[13] Sassan Alizadeh, Michael W Brandt, and Francis X Diebold, “Range-based es-
timation of stochastic volatility models,” The Journal of Finance, vol. 57, no. 3,
pp. 1047–1091, 2002.
[14] Mohsen Jamalabdollahi and Seyed Zekavat, “Toa ranging and layer thickness
computation in nonhomogeneous media,” IEEE Transactions on Geoscience
and Remote Sensing, vol. 55, no. 2, pp. 742–752, 2017.
[15] Amer Catovic and Zafer Sahinoglu, “The cramer-rao bounds of hybrid toa/rss
and tdoa/rss location estimation schemes,” IEEE Communications Letters, vol.
8, no. 10, pp. 626–628, 2004.
[16] Slavisa Tomic, Marko Beko, and Rui Dinis, “3-d target localization in wireless
sensor networks using rss and aoa measurements,” IEEE Transactions on
Vehic- ular Technology, vol. 66, no. 4, pp. 3197–3210, 2017.

Vol. 9, No.8, (2016)
[17] Dragos Niculescu and Badri Nath, “Ad hoc positioning system (aps) using
aoa,” in Proceedings of the Twenty-Second Annual Joint Conference of the
IEEE Com- puter and Communications. Ieee, 2003, vol. 3, pp. 1734–1743.
[18] Jose Velasco, Daniel Pizarro, Javier Macias-Guarasa, and Afsaneh Asaei,
“Tdoa matrices: Algebraic properties and their application to robust denoising
with missing data,” IEEE Transactions on Signal Processing, vol. 64, no. 20,
pp. 5242–5254, 2016.
[19] Li Cong and Weihua Zhuang, “Non-line-of-sight error mitigation in tdoa
mobile location,” in Proceedings of the IEEE Conference on Global
Telecommunica- tions. IEEE, 2001, vol. 1, pp. 680–684.
[20] Pooyan Abouzar, David G Michelson, and Maziyar Hamdi, “Rssi-based dis-
tributed self-localization for wireless sensor networks used in precision agricul-
ture,” IEEE Transactions on Wireless Communications, vol. 15, no. 10, pp. 6638–
6650, 2016.
[21] Yayoi Sakiyama, Nguyen KN Thao, Nguyen M Giang, Shinji Miyadoh, Duong V
Hop, and Katsuhiko Ando, “Kineosporia babensis sp. nov., isolated from plant
litter in vietnam,” International journal of systematic and evolutionary microbi-
ology, vol. 59, no. 3, pp. 550–554, 2009.
[22] Daniel Butzke, Robert Hurwitz, Bernd Thiede, Sigrid Goedert, and Thomas
Rudel, “Cloning and biochemical characterization of apit, a new l-amino acid
oxidase from aplysia punctata,” Toxicon, vol. 46, no. 5, pp. 479–489, 2005.
[23] S Chen, AW Palmer, KTV Grattan, and BT Meggitt, “Fringe order identifica-
tion in optical fibre white-light interferometry using centroid algorithm
method,” Electronics Letters, vol. 28, no. 6, pp. 553–555, 1992.
[24] Dragos Niculescu and Badri Nath, “Ad hoc positioning system (aps),” in
Proceedings of the IEEE Conference Global Telecommunications. IEEE, 2001,
vol. 5, pp. 2926–2931.
[25] Luca Rigazio, Brice Tsakam, and J-C Junqua, “An optimal bhattacharyya cen-
troid algorithm for gaussian clustering with applications in automatic speech
recognition,” in Proceedings of the IEEE International Conference on Acous-
tics, Speech, and Signal Processing. IEEE, 2000, vol. 3, pp. 1599–1602.

Vol. 9, No.8, (2016)
[26] Tian He, Chengdu Huang, Brian M Blum, John A Stankovic, and Tarek F Ab-
delzaher, “Range-free localization and its impact on large scale sensor
networks,” ACM Transactions on Embedded Computing Systems (TECS), vol.
4, no. 4, pp. 877–906, 2005.
[27] Hongbin Tan and Feng Liu, “Research and implementation of apit positioning
algorithm in wsn,” in Proceedings of the 9th IEEE International Conference on
Fuzzy Systems and Knowledge Discovery (FSKD), 2012, pp. 2212–2215.
[28] Nirupama Bulusu, John Heidemann, and Deborah Estrin, “Gps-less low-cost
outdoor localization for very small devices,” IEEE personal communications,
vol. 7, no. 5, pp. 28–34, 2000.
[29] Hongyang Chen, Pei Huang, Marcelo Martins, Hing Cheung So, and Kaoru
Sezaki, “Novel centroid localization algorithm for three-dimensional wireless
sensor networks,” in Proceedings of the 4th International Conference on Wire-
less Communications, Networking and Mobile Computing. IEEE, 2008, pp. 1–
4.
[30] Jan Blumenthal, Ralf Grossmann, Frank Golatowski, and Dirk Timmermann,
“Weighted centroid localization in zigbee-based sensor networks,” in Proceed-
ings of the IEEE International Symposium on Intelligent Signal Processing.
IEEE, 2007, pp. 1–6.
[31] Paolo Pivato, Luigi Palopoli, and Dario Petri, “Accuracy of rss-based centroid
localization algorithms in an indoor environment,” IEEE Transactions on
Instru- mentation and Measurement, vol. 60, no. 10, pp. 3451–3460, 2011.
[32] Bo Zhu and Shu Chen, “An improved centroid localization algorithm for wireless
sensor network,” Chinese Journal of Sensors and Actuators, vol. 6, pp. 024, 2010.
[33] Guangjie Han, Huihui Xu, Trung Q Duong, Jinfang Jiang, and Takahiro Hara,
“Localization algorithms of wireless sensor networks: a survey,” Telecommuni-
cation Systems, pp. 1–18, 2013.
[34] James J Caffery, “A new approach to the geometry of toa location,” in
Proceed- ings of the 52nd IEEE Conference on Vehicular Technology, 2000,
vol. 4, pp. 1943–1949.
[35] Ismail Guvenc and Chia-Chin Chong, “A survey on toa based wireless localiza-
tion and nlos mitigation techniques,” IEEE Communications Surveys & Tutorials,
vol. 11, no. 3, 2009.

Vol. 9, No.8, (2016)
[36] Saipradeep Venkatraman, James Caffery, and Heung-Ryeol You, “A novel toa lo-
cation algorithm using los range estimation for nlos environments,” IEEE
Trans- actions on Vehicular Technology, vol. 53, no. 5, pp. 1515–1524, 2004.
[37] Jan-Jaap Van de Beek, Magnus Sandell, and Per Ola Borjesson, “Ml estimation
of time and frequency offset in ofdm systems,” IEEE transactions on signal
processing, vol. 45, no. 7, pp. 1800–1805, 1997.
[38] Paul C Chestnut, “Emitter location accuracy using tdoa and differential
doppler,” IEEE Transactions on Aerospace and Electronic Systems, , no. 2, pp.
214–218, 1982.
[39] Ryota Yamasaki, Atsushi Ogino, Tsuyoshi Tamaki, Takaki Uta, Naoto Mat-
suzawa, and Takeshi Kato, “Tdoa location system for ieee 802.11 b wlan,” in
Proceedings of the IEEE Conference on Wireless Communications and
Network- ing, 2005, vol. 4, pp. 2338–2343.
[40] KC Ho, “Bias reduction for an explicit solution of source localization using tdoa,”
IEEE Transactions on Signal Processing, vol. 60, no. 5, pp. 2101–2114, 2012.
[41] Kutluyil Dogancay and DA Gray, “Bias compensation for least-squares multi-
pulse tdoa localization algorithms,” in Proceedings of the IEEE International
Conference on Intelligent Sensors, Sensor Networks and Information
Processing, 2005, pp. 51–56.
[42] Wing-Kin Ma, Ba-Ngu Vo, Sumeetpal S Singh, and Adrian Baddeley,
“Tracking an unknown time-varying number of speakers using tdoa
measurements: A random finite set approach,” IEEE Transactions on Signal
Processing, vol. 54, no. 9, pp. 3291–3304, 2006.
[43] Cesare Alippi and Giovanni Vanini, “A rssi-based and calibrated centralized lo-
calization technique for wireless sensor networks,” in Proceedings of Fourth
Annual IEEE International Conference on Pervasive Computing and
Communi- cations Workshops. IEEE, 2006, pp. 5–pp.
[44] Ossi Kaltiokallio, Maurizio Bocca, and Lasse M Eriksson, “Distributed rssi pro-
cessing for intrusion detection in indoor environments,” in Proceedings of the
9th ACM/IEEE International Conference on Information Processing in Sensor
Networks. ACM, 2010, pp. 404–405.

Vol. 9, No.8, (2016)
[45] Ahmed Faheem, Reino Virrankoski, and Mohammed Elmusrati, “Improving

rssi based distance estimation for 802.15. 4 wireless sensor networks,” in
Proceedings of the IEEE International Conference on Wireless Information
Technology and Systems (ICWITS). IEEE, 2010, pp. 1–4.
[46] Mohit Saxena, Puneet Gupta, and Bijendra Nath Jain, “Experimental analysis
of rssi-based location estimation in wireless sensor networks,” in Proceedings
of the 3rd International Conference on Communication Systems Software and
Middleware and Workshops. IEEE, 2008, pp. 503–510.
[47] David C Jenn, “Transmission equation for multiple cooperative transmitters and
collective beamforming,” IEEE Antennas and Wireless Propagation Letters,
vol. 7, pp. 606–608, 2008.
[48] Chun-Pang Wu and Hen-Wai Tsao, “A 110-mhz 84-db cmos programmable gain
amplifier with integrated rssi function,” IEEE Journal of Solid-State Circuits,
vol. 40, no. 6, pp. 1249–1258, 2005.
[49] Hyo-Sung Ahn and Wonpil Yu, “Environmental-adaptive rssi-based indoor lo-
calization,” IEEE Transactions on Automation Science and Engineering, vol. 6,
no. 4, pp. 626–633, 2009.
[50] Yapeng Wang, Xu Yang, Yutian Zhao, Yue Liu, and Laurie Cuthbert, “Blue-
tooth positioning using rssi and triangulation methods,” in Proceedings of the
IEEE Conference on Consumer Communications and Networking. IEEE, 2013,
pp. 837–842.
[51] LC San Martin De Viale, AoA Viale, S Nacht, and M_ Grinstein, “Experimen-
tal porphyria induced in rats by hexachloro-benzene. a study of the porphyrins
excreted by urine,” Clinica Chimica Acta, vol. 28, no. 1, pp. 13–23, 1970.
[52] Abdalla OA Ahmed, Moawia M Mukhtar, Marly Kools-Sijmons, Ahmed H Fa-
hal, Sybren de Hoog, Bert Gerrits van den Ende, Eduard E Zijlstra, Henri Ver-
brugh, AM El Sir, Ahmed M Elhassan, et al., “Development of a species-
specific pcr-restriction fragment length polymorphism analysis procedure for
identification ofmadurella mycetomatis,” Journal of clinical microbiology,
vol. 37, no. 10, pp. 3175–3178, 1999.
[53] Maurice S Bartlett, “Periodogram analysis and continuous spectra,” Biometrika,
vol. 37, pp. 1–16, 1950.

Vol. 9, No.8, (2016)
[54] Jack Capon, “High-resolution frequency-wavenumber spectrum

analysis,” Pro- ceedings of the IEEE, vol. 57, no. 8, pp. 1408–
1418, 1969.
[55] Kenneth Levenberg, “A method for the solution of certain non-
linear problems in least squares,” Quarterly of applied
mathematics, vol. 2, no. 2, pp. 164–168, 1944.
[56] Eduardo Cassano, Francesco Florio, Floriano De Rango, and
Salvatore Marano, “A performance comparison between roc-rssi
and trilateration localization techniques for wpan sensor networks
in a real outdoor testbed.,” in Proceedings of the IEEE on
Symposium Wireless Telecommunications, 2009, pp. 1–8.
References
[1] J.H. Holland, Adaption in Natural and Artificial Systems, University of Michigan Press, Ann Arbor,
Michigan, 1975.
[2] J. Nocedal and S. J. Wright, Numerical Optimization.1993 [Online]. Available:
www.lania.mx/~
c coello/EMOO/knowles01.ps.gz .
[3] F. Rothlauf, Design of Modern Heuristics, Natural Computing Series, Springer-Verlag Berlin Heidelberg,
2011.
[4] S. Bandyopadhyay, S. Saha,, Unsupervised Classification Similarity Measures, Classical and
Metaheuristic Approaches, and Applications, Springer-Verlag Berlin Heidelberg 2013.
[5] Ibrahiem M.M. El-Emary and Mona M. Abd El-Kareem, ―Towards Using Genetic Algorithm
for Solving Nonlinear Equation Systems,‖ World Applied Sciences Journal, Vol. 5, No. 3, pp. 282-289,
2008.
[6] J. McCall, ―Genetic Algorithms for Modeling and Optimization‖, Computational and Appl.
Mathematics, pp. 205-222, 2005.
[7] M. Caramia, P. Dell’Olmo, ―Multi-objective management in Freigt Logistics Increasing Capacity,
service level and safety with Optimization Algorithms‖, 2008.
[8] T. Bäck, Evolutionary Algorithms in Theory and Practice: Evolution Strategies, Evolutionary
Programming, Genetic Algorithms. Oxford University Press, N.Y., 1996.
[9] W. Jeffrey, Herrmann , ―A Genetic Algorithm for Minimax Optimization Problems‖, Evolutionary
Computation, 1999. CEC 99. Proceedings of the 1999 Congress on Vol.2, 1999.
[10] A. Turkcan, M. S. Akturk,‖ A problem space genetic algorithm in multiobjective optimization‖, Journal
of Intelligent Manufacturing, Vol. 14, No. 3-4, pp. 363-378, Dec. 2003.
[11] A. Dahoud, Ibrahiem M.M. El-Emary, Mona M. Abd El-Kareem, ―Application of Genetic Algorithm
in Solving Linear Equation Systems‖, MASAUM Journal of Basic and Applied Science, Vol.1, No.2, pp.
179-185, Sept. 2009.
[12] K. Deb, Single and multi-objective optimization using evolutionary computation, Wiley, 2008.
[13] X. Zou, Y. Chen, M. Liu, L. Kang, ―A new evolutionary approach for solving many-objective
optimization problems‖, IEEE Transactions on systems, MAN and cybernetic- part B: cybernetics, vol.
38, No. 5, Oct 2008.
[14] S. Watanabe, K. Sakakibara,―Multi-objective approaches in a single-objective optimization
environment‖, IEEE Congress on Evolutionary Computation,(CEC 2005), pp. 1714-1721, Vol. 2, Sept.
2005.
[15] S. Watanabe, T. Hiroyasu, M. Miki, ―NCGA: Neighborhood cultivation genetic algorithm for
multi-objective optimization problems‖, Proceedings of the Genetic and Evolutionary Compution
Conference (GECCO’2002), New York, USA, pp. 458–465, 2002.
Vol. 9, No.8, (2016)
[16] K. Deb, A. Pratap, S. Agarwal, T. Meyarivan, ―A Fast and Elitist Multiobjective Genetic Algorithm:
NSGA–II.
IEEE Transactions on Evolutionary Computation”, Vol. 6, pp. 182–197, No. 2, April 2002.
[17] T. Hiroyasu, M. Miki, S. Watanable, ―Distributed Genetic Algorithms with a New Sharing
Approach in Multiobjective Optimization Problems‖, 1999 Congress on Evolutionary Computation,
Washington, D.C., Vol. 1., pp. 69-76, July 1999.
[18] M. Pilat, R. Neruda, ―Combining multiobjective and single-objective genetic algorithms in
heterogeneous island model‖, IEEE Congress on Evolutionary Computation, (CEC 2010), pp. 1-8, July
2010.

Literature Survey Word Document - Copy 2

Uploaded by

Copyright:

Available Formats

You might also like

Literature Survey Word Document - Copy 2

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Literature Survey Word Document - Copy 2

Uploaded by

Copyright:

Available Formats

Localization performs a widespread function in purposes such as routing, goal monitoring etc.

Mobile Anchor Positioning

1. Classification of localization techniques

Range Range Free

Range based approach

Received Signal Strength Indicator (RSSI)

RSSI = −10n log10 d + A (1)

Angle of Arrival (AOA)

Range free localization

2) Received signal strength based room level accuracy indoor

This paper room-level accuracy is achieved by indoor localization of a

3)RSSI based Localization Scheme in Wireless Sensor Networks:

5)Performance Analysis of Diverse Environment based on RSSI

6) Localization Estimation Using Artificial Intelligence Technique in

This paper presents an intelligent localization scheme, which is an artificial neural

7) An Effective Cuckoo Search Algorithm for Node

8) An Efficient Node Localization Approach with RSSI for Randomly

9) Bat Optimization Algorithm with MAP for Localization in Wireless

10) A Wireless Sensor Network with Soft Computing

11) GWO-LPWSN: Grey Wolf Optimization Algorithm for Node

12) Localization in Wireless Sensor Networks: A Survey on

13) A Survey of Localization using RSSI and TDoA Techniques in

14) Maximizing Lifetime of Wireless Sensor Networks Based on Whale

In this paper inorder to increase the lifetime of wireless sensor networks

15) Salp Swarm Algorithm for Node Localization in Wireless Sensor

16)An Enhanced Hybrid 3D Localization Algorithm Based on APIT and DV-

19)DV-max Hop: A Fast and Accurate Range-Free Localization

20)A Survey of Localization using RSSI and TDoA Techniques in

21)Salp Swarm Algorithm for Node Localization in Wireless Sensor

22)A Novel Range-Free Localization Scheme Based on Anchor Pairs Condition

Different Types of Localization Techniques

Measurement Scheme Accuracy Special Requirement

TOA Moderate Synchronization, dense beacons

TDOA High Synchronization, dense beacons

Range-free Area estimation Moderate Dense Beacons

Hop count Moderate Dense Beacons

Localization strategies in WSNs

2.1 Localization process

Signal Measured Estimated

Figure 2.1: Two steps of localization process

2.2 Localization strategies overview in WSNs

Cellular Network Method

Localization Micro Inertial Navigation Method

Extra Modules Geometry

Figure 2.2: Application classification for WSNs

2.3 Extra modules aided approaches

2.3.1 GPS method

2.3.2 Cellular network method

2.3.3 Infrared method

2.3.4 Ultrasonic wave method

2.3.5 Micro inertial navigation method

2.4 Extra modules free approaches

2.4.1 Range free methods

Figure 2.3: Error analysis on DV hop count measurement

Figure 2.4: Principle of APIT algorithm

2.4.1.3 Centroid algorithm

2.4.2 Range based methods