# chapter{Literature GPS satellites, it can take more times to

chapter{Literature Review}label{chap:tr}section{Positioning Systems}Location-based services are location information services that collect and position geographic information on the map that can be accessed on the internet through appropriate electronic gadgets computers or smart-phones.When estimating the user’s location, there are many different techniques that can be used and with each technique having it’s own advantage and drawback. (Eg: When positioning a user at a particular location using only GPS satellites, it can take more times to receive the requested output depending upon the location of the user and the signal strength the user’s device is broadcasting. This feature can result in many drawbacks if you are monitoring an object in real-time. To ease this problem, they have introduced new techniques that combine two or more techniques.Example: The modern smart-phones, uses Wi-Fi hotspot with the cell tower data triangulation to find the GPS data for the user faster.)Smart-phones are not equipped with real GPS receivers because real GPS are expensive. Smart-phones are instead equipped with AGPSfootnote{AGPS – Assisted Global Position System} which are much faster than real GPS receivers because they ping the cell tower, then the cell tower which is equipped with a real GPS receiver ping the GPS satellites and make the results back to the cell tower and then the smart-phone much faster than the process of the smart-phone sending immediately the request to the GPS satellites. subsection{ extbf{Two methods used to Position an Object}}egin{itemize}item extbf{Triangulation:} is a technique used to estimate the user’s location by calculating the angle of arrival of the signal transmitted by the user’s mobile phone to the base tower.egin{figure}h!  centering  includegraphicswidth=60mm{figs/triangulation.png}  caption{Triangulation Technique.}  label{fig:fig 41}end{figure}This process is executed by converting the Antenna of the user’s mobile phone into a 0.1 seconds GPS receiver in order to be able to ping all the nearest cell towers with stronger signals to connect to that cell tower as setted up by the carrier of that mobile phone to call and receive phone calls and other related application such as internet surfing, etc. and to use this feature to estimate the current location of that cell phone.Cell towers have known position and unique cell ID which make this application easier and possible.The cell phone log it’s information from the connected cell tower using the phone’s clock and send the collected information to the cell tower to be processed by the BSSfootnote{BSS – Basic Service Set which is a component of IEEE 802.11 WLAN Architecture} computer and then the BSS computer send the exact location of the cell-phone.As shown in Figure
ef{fig:fig 41} to pinpoint a mobile signal, we need at least three or more cell towers to increase the accuracy of the output or the exact location of the user.Apart of the angle of arrival, there are other techniques that can be used in triangulation like extbf{Time of Arrival or Time difference of Arrival or Enhanced-observed Time difference} which are to estimate the distance of the user from the tower by using the time delay of the received signal at each tower between the time the tower ping a request to the mobile user and receive back the acknowledgement of that request. This can also be used to estimate the speed of the moving object.item extbf{Trilateration:} is a technique used by GPS satellite to estimate the location of the user by measuring the distance of the received signal. The Angle of arrival of the signal is unknown to this technique.This process is executed by calculating the intersection of three radiated spheres from the GPS satellites signals. The GPS satellites have known locations in space which are helpful to estimate the location of the user.As shown in  Figure
ef{fig:fig40}, today, mobile-phones do more than just calling and receiving calls. It also provides features that are helpful to track the user’s position in real-time and share their locations.  The camera’s also might provide the location of the user using the GPS features that record the location and time the picture or the video were taken and google photos also can estimate the place the picture or the video were taken using Google location history that is stored on the google account of the user.end{itemize}egin{table}caption{Pros and Cons of Location Based Services}label{tab:prco}egin{center}egin{tabular}{| c | c |}hline extbf{Pros} & extbf{Cons}\hlineKnow services the user need and to& With location shared publically\give recommendations to the companies.& User’s privacy and security is at risks.\Track User’s movement to might use this\improve predictive traffic routing. into kidnapping the user\Suggests nearby businesses&(eg. Children of a famous person)\and points of interest like Hotels, etc. other related criminals acts.\Faster response in case of emergencies can be found even when you don’t\Weather condition in your location.&want to be found (eg. famous people\Geo-tags posts and photos in social media&or their children or visiting friends and\to create a digital journal of&take pictures from that home,\places you visited in time.&they immediately give the address of\Facility to track your lost items.&that famous person in geo-tags).\Provide directions when navigating.&Burglars might know you are away, \Provide geographic relevant search&from the check-ins in your social media  \results and appropriate ads.&and go to steal at your home. \hlineend{tabular}end{center}end{table}section{Data Mining}Data mining is a process of analyzing and extracting useful information quickly from all your raw data and analytically discovering all the useful situations to give your enough insightful knowledge for making smart decisions and predicting the future outcome.  This is the reason why we applied data mining in our project to carefully analyze our data trajectories and make accurate predictions of future moves of our users driven to our analysis and preferences of the users as we will discuss in this thesis. cite{A}In Data Mining they are two categories of learning sets: egin{enumerate}    item extbf{Supervised learning:} the data include both input and output values. The supervised learning approach takes the input data which are labeled data vectors, process these input data to produce appropriate output and this process can repeat itself multiple times with taking the output of one process as the input of the next process.    The supervised learning algorithm has high performance rate due to it’s ability to analyze data very fast and produce fast and more precise output and this is the reason why supervised learning models have a wide range of applications especially in Neural Networks and Multi-layer perception, etc..        item extbf{Unsupervised learning}: the user does not provide labeled data to the machine during the training of the data set, this means that the machine has to extract the data by its own to produce the required output. Due to this, it’s not easy to assess the performance and accuracy of the result.    This method is very useful when we need to cluster the input information in classes on the basis of their statistical properties, for example in the case of density estimation, K-means, self-organizing maps, etc.     end{enumerate}subsection{Trajectory Data Processing}Initially, spatial trajectories are not authentic due to some noises during collecting the data trajectories and other factors such as receiving poor signals. This is the reason why we introduced the concept ofcite{A} $trajectory$ $processing$, for correcting these errors.In this section we will deal with extbf{Noise filtering, Segmentation} and extbf{Map matching}.\ The main goal of :egin{itemize}item extbf{Noise filtering:} is to filter the data trajectory, looking at unwanted data coordinates and removing them. This is performed by checking the trajectory the user took, and removing the points that fall away from other points in the trajectory, as shown in Figure
ef{fig:fig7}, the point $p_4$ is far away from the point $p_3$ and $p_5$ which means that the point $p_4$ will be removed from the trajectory.egin{figure}h!  centering  includegraphicswidth=80mm{figs/noise.jpg}  caption{Single Noise point in a trajectory.}  label{fig:fig7}end{figure}item extbf{Trajectory Segmentation:}, Is a process of setting a data trajectory into fragments by the time interval to help analyze more easily the data trajectory to give precise and more meaningful analysis.This property is much helpful during the clustering of the data trajectory.item extbf{Map Matching:} cite{G,I} is a process of assigning geographic coordinates (eg. latitude, longitude and altitude) to the digital map.Digital map has the properties to take these provided coordinates and convert them into a human readable format, providing some facilities to help the human navigate and showing some particular area of interest to the user on the map.Eg: If the user wants to know the traffic ahead on his trajectory, he/she can access it in real-time and the digital map can guide him/her from the point of origin to the destination using some feature like GPS to collect his/her current position and project it to the map and making it easier for him/her to predict the shortest path to his/her destination.end{itemize}section{Anomalies detection in The Trajectory}When collecting data coordinates of a moving object, the geographic coordinates are recorded with some time delay.The error varies depending upon the length of the time delay as to mention that the object will be at another location after that delay, so this may cause some misreading of our data collection and data analysis. In the chapter