Loading...

Efficient Data Collection in Wireless Sensor Networks

Samimi, Ghasem | 2014

604 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: English
  3. Document No: 45721 (52)
  4. University: Sharif University of Technology, International Campus, Kish Island
  5. Department: Science and Engineering
  6. Advisor(s): Hemmatyar, Ali Mohammad Afshin
  7. Abstract:
  8. Wireless Sensor Networks (WSNs) consist of few hundred to thousand sensor nodes that are usually deployedrandomly in a sensing field. The duty of sensor nodes is to sense the environment for data collection and sending these data to a sink node for future processes. Sensor nodes are equipped with a non-rechargeable battery. Hence, network lifetime is an important issue in WSNs. When a static sink node is used in a WSN, sensor nodes that are located in the neighborhood of the sink node willdrain out of energy faster than other sensor nodes which are farther away from the sink node, because traffic load is heavier on sensor nodes which are closer to the sink node. This problem is named “the sink neighborhood problem”, “crowded center effect” or “the energy hole problem”. To tackle with this problem, some methods are proposed by researchers. In this thesis, we have studied the network lifetime maximization problem by jointly considering sink mobility and routing.Moreover, we take the advantage of multiple mobile sink nodes to further improvement of network lifetime. A Mixed Integer Linear Programming (MILP) formulation called Optimal Multiple Mobile Sink Placement (OMMSP) is proposed to find the optimal routing toward a sink node and sink mobility pattern. In addition, to solve the MILP method in a polynomial time, we have proposed a heuristic approach named Heuristic Multiple Mobile Sink Placement (HMMSP). We have evaluated the performance of the proposed methods in several scenarios with typical WSNs topologies. In each scenario, we have compared the performance of our methods with static sink placement and conventional sink mobility methods. Our simulation results prove the benefits of using sink mobility and demonstrate that the proposed methods outperforms the others significantly in the terms of network lifetime, sensors residual energy and computation time. In particular, the network lifetime achieved by using OMMSP is about 3.66 times of the network lifetime achieved by using static sink placement, Beside, HMMSP yields a network lifetime which is about 3.51 times the network lifetime resulted by using static sink placement. To compare our proposed methods with conventional sink mobility methods indicates that OMMSP and HMMSP are more effective in prolonging network lifetime. In addition, we have studied the impact of multi sink on the network lifetime. Our simulation results shows that we can improve the network lifetime up to 47% by using two mobile sink nodes instead of a single mobile sink node. Besides, the network lifetime achieved by using three mobile sink nodes is about 1.95 of the network lifetime obtained by a single mobile sink node. We have also compared the computation time of each method in order to study the efficiency of OMMSP and HMMSP. Our simulation results show that HMMSP runs faster than OMMSP and is very useful in large scale WSNs. In particular, for an arbitrary WSN network (i.e. sensor nodes are distributed uniformly) with 60 sensor nodes and three mobile sink nodes, HMMSP results are less than 100 seconds while OMMSP results are about 900 seconds
  9. Keywords:
  10. Wireless Sensor Network ; Network Lifetime ; Mixed Integer Linear Programming ; Mobility Sink ; Multi-Sink

 Digital Object List

 Bookmark

No TOC