Efficient and guaranteed service coverage in partitionable mobile ad-hoc networks

In wireless ad-hoc networks, the network topology changes dynamically and unpredictably due to node mobility. Such topological dynamics are further exacerbated by the natural grouping behavior in the mobile user's movement, which leads to frequent network partitioning. Network partitioning poses significant challenges to the provisioning of centralized services in ad-hoc networks, since partitioning disconnects many mobile users from the central server. We propose a collection of novel run-time algorithms that adaptively ensure the centralized service is available to all mobile nodes during network partitioning, while minimizing the number of servers required. The network-wide service coverage is achieved by partition prediction and service replication on the servers, and assisted by distributed service selection on regular mobile nodes. Simulation results show that our algorithm efficiently achieves guaranteed service coverage to all nodes. To the best of our knowledge, there have been no similar approaches that use partition prediction to provision centralized services adaptively in partitionable mobile ad-hoc networks.