მონაცემთა შეგროვება დინამიურ, ლიმიტირებული კავშირის მქონე ტიპის სენსორულ ქსელში

ავტორი


გივი ლობჟანიძე, ალექსანდრე ლობჟანიძე, თეიმურაზ ფესტვენიძე

კატეგორია

მოდელირება და პროგრამირება
სტატიის გადმოწერა

INTELLIGENCE GATHERING VIA DATA MULES IN SPARSE SENSOR NETWORK

1Givi Lobzhanidze, 2Aleksandre Lobzhanidze, 3Teimuraz Pestvenidze
2University of Missouri, USA, 1,3Akaki Tsereteli State University, Georgia;

lobzhanidze.givi@gmail.com, alex@lobzhanidze.com

Abstract – Our project presents, analyzes and proposes a reduced implementation for the architecture of a sparse network that is randomly deployed. These networks exhibit the unique characteristic of having their nodes deployed far apart and almost no communication between two adjacent nodes is possible making the data transfer really hard. The approach we propose exploits the presence of a mobile entity (or entities) that is called MULE. For future references we define Source as a randomly placed node on a geographical area and the MULE as the mobile entity that moves around the grid collecting data stored on the Sources. The MULE picks data from the Sources whenever it is in close range to the nodes. The data is buffered in the MULE’s inner memory and finally brought to an Access Point. The Sources transmit data in short ranges, therefore such an approach is beneficial in terms of the nodes’ energy savings making it more efficient over longer periods of time. For the purpose of implementation we have worked on the communication between Sources and the Sink, and some data retrieval and analytical tools are implemented as well. Our project presents the implementation of the proposed idea that experiences randomly deployed Sources in a sparse way that take data from environment and pass it to mobile entity whenever it is in close range.

Index Terms—Sparse Sensor Networks, Data Mules, Mobile Communication, Mica-2 Mote

I. INTRODUCTION

Advances in Wireless Embedded Systems provide for very powerful data gathering techniques. The small and inexpensive little motes can give us invaluable information that might help change the future of a whole nation. The small devices together with applications that include environment monitoring, data gathering, in-network data processing, data routing and finally data analysis are very efficient in terms of decision making. It doesn’t matter what kind of monitoring we do, be it just traffic control, warehouse monitoring or spying an enemy lines, Wireless Sensors Networks (WSN) are always useful. Our practical approach has to deal with a real-life application of Wireless Sensor Networks: helping the Central Intelligence Agency (CIA) to get information from uncontrolled areas where enemy positions are located. The intelligence agency has to report about possible danger factors that the enemy has managed to develop (i.e., higher temperatures than usual, high noises, high/low magnetic fields, etc.). This kind of sensor surveillance/espionage of a war field or just spying on some enemy territories would provide the CIA with useful information that would provide for better defense plans in the future.
The basic idea we propose is: firstly we learn in what (geographical location the potential danger is located; secondly the CIA agents deploy Sources randomly on the enemy lines from an air-plane. The number of Sources deployed cannot be too high because the enemy could find them easily then, so a method should be devised to deploy them efficiently. These Sources become data Sources and after they have been placed, the Sources start sampling valuable data and buffering it on their memory. Because the CIA agents do not have access into the enemy lines, a method has to be devised in order to gather information from the sources without anyone being physically present The solution found for this dilemma is the MULE system in which a mobile data-gathering system (MULE) copies data from sources whenever it finds itself in the close range of a Source. This MULE system would be easily flown around the area where the Sources are deployed once in a while in order to gather data from the Sources. After sources have been deployed, agent can carry mobile device and whenever he walks within a range of source, valuable data is copied to agent’s spying device. Later he can bring this device to headquarter and scientists will analyze data and reach some conclusions related to the changes that have been happening in the enemy lines.
The Wireless Sensor Network we are trying to deploy in this application should collect data from widely sensors and the design must be low in cost and work within limited energy budget because we want the Sources to be alive for as long as possible without needing a new battery, which is actually impossible to provide after deployment. The Sources will need to be spread over a large geographical area resulting in a sparse network. In our design no Sources can hear each other over the wireless connection, so a way of gathering the information stored on the Sources needs to be devised in the form of a MULE mobile entity which can approach the Source and pick the piece of information it needs. The Sources distribution might not be under our control and we don’t have full access to the Sources. This arise another concern: we should make sure that the Sources are deployed in a way so that we get enough information to analyze data and make decisions. The communication between the Source and the MULE must be quick and reliable, i.e. get the needed information and all redundant information is eliminated. For the purpose of this project we have decided to deploy 3 sources and one MULE. However we are not limited in the number of MULEs that can be deployed. The CIA might have several agents that are trusted by enemy and they all can bring information in an espionage configuration. In other words the more data the better. For the experiments we have used the Wireless Sensor Network lab and one of authors was playing the trusted agent, who is sent behind enemy lines and tries to take a tour of the mined Sources area in order to gather as much information as possible. In order to implement this application we have used 2 MICA2 motes as Sources, 1 MICA2 mote as a MULE who is also used as a sink. Fig. 1 (b) shows the device used in our experiment. The Tiny OS operating system was used to manage the motes, the nesC programming language and also some conversion modules written in Object Oriented Programming and Bash shell.

სტატიის გადმოწერა



——————————————————————–

Facebook Twitter Email Linkedin Google