E NERGY -A WARE G ENETIC A LGORITHM FOR M ANAGING W IRELESS S ENSOR N ETWORKS Abhishek Karpate.

Slides:



Advertisements
Similar presentations
Ali Husseinzadeh Kashan Spring 2010
Advertisements

Marković Miljan 3139/2011
Using Parallel Genetic Algorithm in a Predictive Job Scheduling
SELF-ORGANIZING MEDIA ACCESS MECHANISM OF A WIRELESS SENSOR NETWORK AHM QUAMRUZZAMAN.
O PTIMAL P ATH P LANNING FOR M OBILE R OBOT -T RAILER S YSTEMS Team 22: Siwei Wang Xin Yu Xi Li.
Sensor Network 教育部資通訊科技人才培育先導型計畫. 1.Introduction General Purpose  A wireless sensor network (WSN) is a wireless network using sensors to cooperatively.
Sec-TEEN: Secure Threshold sensitive Energy Efficient sensor Network protocol Ibrahim Alkhori, Tamer Abukhalil & Abdel-shakour A. Abuznied Department of.
KAIST Adaptive Triangular Deployment Algorithm for Unattended Mobile Sensor Networks Suho Yang (September 4, 2008) Ming Ma, Yuanyuan Yang IEEE Transactions.
Cooperative Multiple Input Multiple Output Communication in Wireless Sensor Network: An Error Correcting Code approach using LDPC Code Goutham Kumar Kandukuri.
1 Sensor Networks and Networked Societies of Artifacts Jose Rolim University of Geneva.
Applying Genetic Algorithms to Decision Making in Autonomic Computing Systems Authors: Andres J. Ramirez, David B. Knoester, Betty H.C. Cheng, Philip K.
Non-Linear Problems General approach. Non-linear Optimization Many objective functions, tend to be non-linear. Design problems for which the objective.
1 Lecture 8: Genetic Algorithms Contents : Miming nature The steps of the algorithm –Coosing parents –Reproduction –Mutation Deeper in GA –Stochastic Universal.
COMP305. Part II. Genetic Algorithms. Genetic Algorithms.
Adaptive Security for Wireless Sensor Networks Master Thesis – June 2006.
On the Construction of Energy- Efficient Broadcast Tree with Hitch-hiking in Wireless Networks Source: 2004 International Performance Computing and Communications.
Distributed Cluster Repair for OceanStore Irena Nadjakova and Arindam Chakrabarti Acknowledgements: Hakim Weatherspoon John Kubiatowicz.
Introduction to Evolutionary Computation  Genetic algorithms are inspired by the biological processes of reproduction and natural selection. Natural selection.
Intro to AI Genetic Algorithm Ruth Bergman Fall 2002.
Security in Wireless Sensor Networks Perrig, Stankovic, Wagner Jason Buckingham CSCI 7143: Secure Sensor Networks August 31, 2004.
Intro to AI Genetic Algorithm Ruth Bergman Fall 2004.
Evolutionary algorithms
Genetic Algorithm.
Vikramaditya. What is a Sensor Network?  Sensor networks mainly constitute of inexpensive sensors densely deployed for data collection from the field.
Low-Power Wireless Sensor Networks
College of Engineering Non-uniform Grid- based Coordinated Routing Priyanka Kadiyala Major Advisor: Dr. Robert Akl Department of Computer Science and Engineering.
SOFT COMPUTING (Optimization Techniques using GA) Dr. N.Uma Maheswari Professor/CSE PSNA CET.
Patch Based Mobile Sink Movement By Salman Saeed Khan Omar Oreifej.
An Intelligent and Adaptable Grid-Based Flood Monitoring and Warning System Phil Greenwood.
Efficient Deployment Algorithms for Prolonging Network Lifetime and Ensuring Coverage in Wireless Sensor Networks Yong-hwan Kim Korea.
Optimization Problems - Optimization: In the real world, there are many problems (e.g. Traveling Salesman Problem, Playing Chess ) that have numerous possible.
Maximum Network Lifetime in Wireless Sensor Networks with Adjustable Sensing Ranges Cardei, M.; Jie Wu; Mingming Lu; Pervaiz, M.O.; Wireless And Mobile.
한국기술교육대학교 컴퓨터 공학 김홍연 Habitat Monitoring with Sensor Networks DKE.
Benjamin AraiUniversity of California, Riverside Reliable Hierarchical Data Storage in Sensor Networks Song Lin – Benjamin.
G ENETIC A LGORITHMS Steve Foster. I NTRODUCTION Genetic Algorithms are based on the principals of evolutionary biology in order to find solutions to.
Computational Complexity Jang, HaYoung BioIntelligence Lab.
1 Sensor Deployment –Sequential Deployment 1. Sequential Deployment Deploy one or more sensors Get Feedback from them Deploy another set And, so on Disadvantage.
Derivative Free Optimization G.Anuradha. Contents Genetic Algorithm Simulated Annealing Random search method Downhill simplex method.
AD-HOC NETWORK SUBMITTED BY:- MIHIR GARG A B.TECH(E&T)/SEC-A.
G ENETIC A LGORITHMS Ranga Rodrigo March 5,
C OMPARING T HREE H EURISTIC S EARCH M ETHODS FOR F UNCTIONAL P ARTITIONING IN H ARDWARE -S OFTWARE C ODESIGN Theerayod Wiangtong, Peter Y. K. Cheung and.
Efficient Energy Management Protocol for Target Tracking Sensor Networks X. Du, F. Lin Department of Computer Science North Dakota State University Fargo,
A Dead-End Free Topology Maintenance Protocol for Geographic Forwarding in Wireless Sensor Networks IEEE Transactions on Computers, vol. 60, no. 11, November.
By Naeem Amjad 1.  Challenges  Introduction  Motivation  First Order Radio Model  Proposed Scheme  Simulations And Results  Conclusion 2.
MAE 552 Heuristic Optimization Instructor: John Eddy Lecture #12 2/20/02 Evolutionary Algorithms.
Evaluating Wireless Network Performance David P. Daugherty ITEC 650 Radford University March 23, 2006.
Cross-Layer Scheduling for Power Efficiency in Wireless Sensor Networks Mihail L. Sichitiu Department of Electrical and Computer Engineering North Carolina.
Link Layer Support for Unified Radio Power Management in Wireless Sensor Networks IPSN 2007 Kevin Klues, Guoliang Xing and Chenyang Lu Database Lab.
Wireless sensor and actor networks: research challenges
D Nagesh Kumar, IIScOptimization Methods: M8L5 1 Advanced Topics in Optimization Evolutionary Algorithms for Optimization and Search.
Rajani Muraleedharan and Lisa Ann Osadciw By: Mai Ali Sayed Ahmed.
Toward Reliable and Efficient Reporting in Wireless Sensor Networks Authors: Fatma Bouabdallah Nizar Bouabdallah Raouf Boutaba.
Genetic Algorithm Dr. Md. Al-amin Bhuiyan Professor, Dept. of CSE Jahangirnagar University.
Artificial Intelligence By Mr. Ejaz CIIT Sahiwal Evolutionary Computation.
1 Comparative Study of two Genetic Algorithms Based Task Allocation Models in Distributed Computing System Oğuzhan TAŞ 2005.
BORDER SECURITY USING WIRELESS INTEGRATED NETWORK SENSORS (WINS) By B.S.Indrani (07841A0406) Aurora’s Technological and Research Institute.
Scalable Coverage Maintenance for Dense Wireless Sensor Networks Jun Lu, Jinsu Wang, Tatsuya Suda University of California, Irvine Secon ‘ 06.
Wireless Sensor Network: A Promising Approach for Distributed Sensing Tasks.
EVOLUTIONARY SYSTEMS AND GENETIC ALGORITHMS NAME: AKSHITKUMAR PATEL STUDENT ID: GRAD POSITION PAPER.
A Secure Routing Protocol with Intrusion Detection for Clustering Wireless Sensor Networks International Forum on Information Technology and Applications.
Lecture 8: Wireless Sensor Networks By: Dr. Najla Al-Nabhan.
Genetic Algorithms And other approaches for similar applications Optimization Techniques.
Medium Access Control. MAC layer covers three functional areas: reliable data delivery access control security.
In the name of God.
Genetic Algorithms.
Introduction to Wireless Sensor Networks
Introduction to Genetic Algorithm (GA)
Net 435: Wireless sensor network (WSN)
Bluetooth Based Smart Sensor Network
Presentation transcript:

E NERGY -A WARE G ENETIC A LGORITHM FOR M ANAGING W IRELESS S ENSOR N ETWORKS Abhishek Karpate

O VERVIEW  Wireless Sensor Networks are quickly emerging as a technology for tracking and monitoring in many domains.  Military applications  Environmental applications  Health applications  Agricultural applications  Smoke detection  They consist of spatially distributed sensors which cooperate among themselves for any task.  The sensors are small integrated circuits which are embedded in wireless devices.

C HALLENGES Reliability and Robustness Sensor networks are not meant for frequent maintenance. They should be operated by sensors that are reliable and should be deployed in large numbers. Energy conservation Sensor networks have limited computational and communication capabilities. Algorithms need to be collaborative with energy aware communication. Real time data acquisition and processing There is a critical need for efficient data communication and data processing. Some techniques that are used are event ordering and synchronization. Data Management An embedded real time database is needed which stores the data of interest and provides results to different queries. Data privacy and Security The data collected is sensitive. It should be made sure that the data is properly transmitted and collected with no loss.

P REVIOUS W ORK Many Researchers approached the management of wireless sensors problem from one specific angle such as network life span and coverage. Very little has been done to optimize multiple objectives concurrently which is critical in this case, in particular the trade off between performance and consumed energy. Almost all current approaches provide static solutions that lack the flexibility to change priority according to the application domain. Few research studies looks at the impact of the density of the sensor nodes.

A PPROACH Area monitored by single sensor Area monitored by multiple sensors Set of Sensor Nodes Set of Sub-regions

T HE P ROPOSED M ODEL WSN and the set of areas it needs to cover can be represented by a bipartite graph G as follows: G = ( N ∪ A, E) ; where, N  Set of sensor nodes A  Set of area monitored E  Edge which exists between a node in N to a node in A if and only if that particular node monitors that area.

K EY F EATURES OF THE M ODEL  The proposed model provides a flexible mechanism to incorporate various parameters of the problem, such as:  Heterogeneity of the sensor network  Optimization criteria  An any given instance, a solution is given by identifying a subset of nodes representing sensors (active sensors) that dominates the entire set of nodes representing the areas need to be covered.  Such a subset needs to be identified in a way that attempts to optimize various parameters related to performance and energy.

E VOLUTIONARY A LGORITHM  The evolution starts with n randomly generated strings.  The length of the string represents the number of sensors.  The state of each individual sensor is represented by a 1bit binary number called gene.  The gene defines the status of the nodes as follows: s i =

 Randomly generated strings are subjected to mutation.  Mutation is the random alteration of a gene of a chromosome.  Helps in reintroducing the random cells that have been lost. M UTATION Original String: Mutated String:

Crossover is an operator that combines two strings to produce a new string. The purpose behind crossover is that the resulting string takes better characteristics from both the parents The length of each substring to be swapped is taken as input from the user. The number of strings generated depends on the length of the substring Crossovers generated when the length of the substring is one-third the string length C ROSSOVER

As discussed there are the following sets and constants: S = {s 1, s 2, …, s n } the set of sensors A = {A 1, A 2, …, A m } the set of sub-regions We define fitness function, f ∀ S, ∀ A as follows: f = αP + βQ + γR where, P  Percentage of coverage Q  Quality of coverage R  Number of inactive nodes and α, β, γ are tuning parameters to customize the function according to the priority of each application F ITNESS F UNCTION

 The process of selecting better individuals to use as parents for the further generations.  The process of selection is stopped when for a few cycles no strings with better fitness are found. S ELECTION The healthiest strings are considered for future generations  Elitism can occur in one of the following forms: Pure Elitism: The strings that are considered are all obtained from the previous generation Partial Elitism: Some of the strings that are considered are obtained from the previous generation, others are randomly generated

A SSESSMENT A benchmark is needed to compare the performance of the algorithm. The performance of the algorithm is compared with the greedy Round-Robin algorithm. The algorithm is executed as follows: The desired number of sensors that need to be active are taken as input. Accordingly sensors are divided into groups. For any particular time slice a particular group of sensors are turned on.

E XPERIMENTS The simulations are carried out on three types of networks – densely, medium and sparsely populated The simulations are carried out in order to achieve the following: Have a set of options available for any desired situation. As needed, the required numbers of sensors are turned on and the requirement is fulfilled. Control the activation and deactivation of the sensors. A network that swings its main priority from coverage to energy saved.

R ESULTS Densely populated network Sensor networks having more than 50 sensors are classified as dense networks. Input ParametersOutput Parameters % Coverage Quality Coverage Energy Saved % Coverage Quality Coverage Energy Saved Case Case Case Case Case Table 1: Simulation test results using genetic algorithmTable 2: Simulation test results using round-robin algorithm Number of Active Sensors % Coverage Quality Coverage Energy Saved Case Case Case Case Case

Input ParametersOutput Parameters % Coverage Quality Coverage Energy Saved % Coverage Quality Coverage Energy Saved Case Case Case Case Case Table 3: Simulation test results using genetic algorithm Number of Active Sensors % Coverage Quality Coverage Energy Saved Case Case Case Case Case Table 4: Simulation test results using round-robin algorithm Intermediately populated network Sensor networks having less than 50 sensors and more than 25 sensors are classified as intermediately populated networks.

Input ParametersOutput Parameters % Coverage Quality Coverage Energy Saved % Coverage Quality Coverage Energy Saved Case Case Case Case Case Table 5: Simulation test results using genetic algorithm Number of Active Sensors % Coverage Quality Coverage Energy Saved Case Case Case Case Case Table 6: Simulation test results using round-robin algorithm Sparsely populated network Sensor networks having less than 25 sensors are classified as sparsely populated networks.

A NALYSIS OF R ESULTS  Genetic algorithms overall provide better coverage for the same amount of energy used as compared to the round-robin algorithm.  A better quality of coverage is attained by spending equal amount of energy in genetic algorithms which provide a much needed redundancy for critical applications.  The more densely populated a network is the better capable it is in producing a wider range of viable results.  As the network goes more sparse, lower amounts of energy would be saved in order to attain the some level of coverage or quality.  The more number of strings considered in the partial elitism the better are the chances of getting more viable results.  More efficient results are attained if the number of elements in the crossover operation are kept low because then the crossover is capable of producing more offspring.

C ONCLUSIONS  The project proposes a scheme by which the activation/deactivation process in a wireless sensor network can be adjusted to optimize multiple objectives.  According to the specific need at a given instance, the behavior of the wireless sensor network can be changed depicting networks of different priorities.  The proposed evolutionary algorithm presents different solutions to the optimization problem and outperforms standard greedy techniques in finding the most fit solution.  The graph theoretic model along with the algorithm can be further expanded by creating strings that represent various states of the sensors such as transmission, listening, active, sleep.

Thank You