Fault-tolerant Relay Node Placement in Heterogeneous Wireless Sensor Networks Xiaofeng Han, Xiang Cao, Errol L. Lloyd, and Chien-Chung Shen [Delaware University] INFOCOM 2007 전산학과 강유화 2008. 12. 11.

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Contents Introduction Network Model and Preliminaries One-way Partial Fault-tolerance Relay Node Placement Two-way Partial Fault-tolerance Relay Node Placement One-way and Two-way Full Fault-tolerance Relay Node Placement Heuristic Implementations Experimental Results Conclusion Discussion CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Introduction Fault tolerance in wireless sensor networks Energy depletion Harsh environment Malicious attacks One approach to achieve fault tolerance in WSN to deploy a small number of additional relay nodes to provide k (k ≥ 1) vertex-disjoint paths between every pair of functioning devices Relay node placement full fault-tolerance relay node placement (FFRP) partial fault-tolerance relay node placement (PFRP) relay node placement has also been studied in two-tiered homogeneous wireless sensor networks CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 3

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Problem definition Heterogeneous wireless sensor networks (H-WSNs) The different transmission radii of target nodes introduce asymmetric communication links One-way / Two-way partial fault-tolerance relay node placement (One-way / Two-way PFRP). One-way / Two-way full fault-tolerance relay node placement (One-way / Two-way FFRP). Contribution O(σk2) approximation algorithms for both One-way PFRP and Two-way PFRP; O(σk3)approximation algorithms for both One-way FFRP and Two way FFRP evaluate heuristic implementations of the proposed algorithms CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 4

Proposed two-tier (cluster-based) architecture “Fault-Tolerant Clustering of Wireless Sensor Networks” G.Gupta and Mohamed Yunis, 2003 IEEE proceedings of WCNC Relay node Sensor node

Network Model and Preliminaries Terms, Symbols and their Semantics CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 6

Network Model and Preliminaries One-way Steinerization. (Equation 1) Two-way Steinerization: (Equation 2) CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 7

Network Model and Preliminaries Minimum k-Vertex Connected Spanning Graph An important problem related to the relay node placement is to find a minimum k-vertex connected spanning graph (MKCSG). MKCSG problem is to compute a k-vertex connected spanning graph For undirected graphs, when k ≥ 2, this problem is NP-hard, For directed graphs, this problem is Np hard when k ≥ 1. CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 8

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 One-way PFRP One-way Partial k-Vertex Connected Graph O(σk2)-approximation algorithm. CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 9

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 One-way PFRP Definition 3.8: [Relay Component][5] We start at r (relay node), travel along each edge incident to r in G If we meet a relay node, we repeat the process; if we meet a target node, we stop. all of the nodes (target or relay) and edges visited in this recursive process form a relay component of r. Definition 3.3: [Super Path][4] G = (V ∪ R, E) be a one-way partial k-vertex connected graph for V a one-way path PG(u, v) in G is a super path, if u and v are target nodes and every interior node (if any) of PG(u, v) is a relay node Every relay node placed by Algorithm 1 is on exactly one super path → → CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 10

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Two-way PFRP Two-way partial k-vertex connected graph O(σk2)-approximation algorithm. CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 11

One-way and Two-way PFRP CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 12

One-way and Two-way PFRP Algorithm 3 for One-way (Two-way) FFRP In Step 6, Algorithm 3 connects the relay nodes on PG(u, v) with u and v as well as their in-neighbors and out-neighbors [full-connection] In Step 7, Algorithm 3 tests each cluster of relay nodes deployed in Step 6, and removes the cluster if the graph of the resulting network remains a directed k-vertex connected graph. O(σk3)-approximation algorithm. → CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 13

Heuristic Implementations Most existing approximation algorithms for MKCSG problem quite complicated and difficult to implement in computation- and communication constrained sensor networks Motivated by the heuristic algorithms, greedy heuristic algorithm as a practical alternative for the MKCSG problem repeatedly add the edges that can best help to improve the graph connectivity until the graph becomes k-vertex connected. Definition 7.1: [Contribution]: In a graph that is not k-vertex connected, the connectivity between node pairs is lower than k. unsaturated node pairs. The contribution of an edge uv (or uv) is defined as the number of unsaturated node pairs whose connectivity can be improved by the deployment of edge uv (or uv) in this graph. → → CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 14

Heuristic Implementations CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 15

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Experimental Results Performance of our algorithms using heuristic implementations Qualnet 3.8[9] as the simulation platform. Randomly place target nodes in a 1000m × 1000m 2D terrain To model a H-WSN, we set T (min) = 200m and T (max) = 500m, and let every target node use a random transmission radius between T (min) and T (max) in each simulation. CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 16

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Experimental Results The number of relay nodes first increases, then decreases One-way FFRP algorithm requires 5.9 times and 10.2 times more relay nodes than the One-way PFRP algorithm Two-way FFRP algorithm requires 4.6 times and 7.7 times more relay nodes than the Two-way PFRP algorithm CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Experimental Results The case of k = 4, N (target node) stands for the network size In a sparse network with 20 target nodes, when T(relay) increases from T (min) to T (max), and the number of relay nodes computed by the One-way and the Two-way PFRP algorithms drops 36% and 42% In a dense network with 60 target nodes, the performance of the One-way and the Two-way PFRP algorithms remain stable. CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 18

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Conclusions Addresses the problem of deploying a minimum number of relay nodes to achieve diverse levels of fault-tolerance in the context of heterogeneous wireless sensor networks Develop approximation algorithms O(σk2) approximation algorithms for one-way and two way partial fault-tolerance relay node placement O(σk3)- approximation algorithms for one-way and two-way full fault tolerance relay node placement. To support real applications, provide heuristic implementations of these algorithms performance of the proposed algorithms is much better than the performance ratios derived CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 Discussion Energy efficiency? One relay node failure? Survivability: ensure alternate path exists for sensor nodes when one of the relay nodes fail Convergence time in simulation result Locally optimal relay node placement in heterogeneous wireless sensor networks GLOBECOM 2005 Quanhong Wang   Kenan Xu   Takahara, G.   Hassanein, H.   Dept. of Electr. & Comput. Eng., Queen's Univ., Canada CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안

CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 References [4] E. L. Lloyd and G. Xue, “Relay node placement in wireless sensor networks,” IEEE Transactions on Computers, vol. 56, pp. 134–138, 2007. [5] J. L. Bredin, E. D. Demaine, M. Hajiaghayi, and D. Rus, “Deploying sensor networks with guaranteed capacity and fault tolerance,” in ACM MobiHoc, 2005 CS710 컴퓨터구조 특강 – 차세대 무선네트워크 및 보안 21

INFOCOM 2007 Session 47: Sensor Networks IV Fault-Tolerant Relay Node Placement in Wireless Sensor Networks: Problems and Algorithms Weiyi Zhang, Guoliang Xue, Satyajayant Misra (Arizona State University, US) Data Persistence in Large-scale Sensor Networks with Decentralized Fountain Codes Yunfeng Lin, Ben Liang, Baochun Li (University of Toronto, CA) Fault-tolerant Relay Node Placement in Heterogeneous Wireless Sensor Networks Xiaofeng Han, Xiang Cao, Errol Lloyd, Chien-Chung Shen (University of Delaware, US) Optimal Policies for Distributed Data Aggregation in Wireless Sensor Networks Zhenzhen Ye, Alhussein Abouzeid, Jing Ai (Rensselaer Polytechnic Institute, US)