Domination Algorithms for Lifetime Problems in Self-organizing Ad hoc and Sensor Networks

Wireless sensor networks propound an algorithmic research problems for prolonging life of nodes and network. The domination algorithms can address some of fundamental issues related to lifetime problems in ad hoc and sensor networks. Most of the graph domination problems are NP-complete even with unit-disk-graphs. The investigation of the thesis addresses some of lifetime issues in sensor network with the approximate domination algorithm. In this work, we consider distributed algorithms of some important domination problems namely, maximum domatic partition problem (DPP), maximum connected domatic partition (CDP) problem, minimum connected dominating set (MCDS) problem, node-mobility transparent connected dominating set problem in context of unitdisk graphs and obtain solutions using state-of-the-art principles of well-known MIS (maximal independent sets). We incorporated self-organization feature to domatic partition for sensor networks. Domatic partition problems has variety of applications. In sensor networks our deterministic self-organizing domatic partition algorithm is used to provide maximum cluster lifetime in hierarchical topology control of sensor networks. Minimum connected dominating set is reported to provide a virtual backbone for ad hoc networks. The maximum lifetime of connected dominating set felt constrained to support virtual backbone in sensor networks. We modeled the maximum lifetime connected dominating set as connected domatic partition problem. We introduced a distributed algorithm for connected domatic partition problem. To our knowledge no such connected domatic partition is reported in literature. The minimum connected dominating set has drawn a considerable research interest and several approximation schemes are reported. We have introduced a collaborativecover heuristic and developed a distributed approximation algorithm for minimum connected dominating set problem using it with a single leader having an approximation factor of (4.8+ ln 5)opt+1.2, where opt is the size of any optimal CDS in G. This approximation provides an effective loss-less aggregation backbone for sensor networks. The results show the improvement in prolonging the life of sensor networks. The CDSbackbone gets disturbed by the mobility of nodes. We developed an integrated scheme adapting CDS to the node’s mobility transparently and efficiently. Adapting CDS to node-mobility is carried out by using four steps: i) reinforcing a self-organization to a multi-protocol relay(MPR) based connected dominating set, ii) reinforcing self-reconfiguration of CDS when a node becomes mobile or halts after mobile operation, iii) adapting CDS to mobile-node by tracking of mobile node for its location updates and iv) optimizing location updates using weighted CDS based on a Markov model.