Abstract
After the occurrence of a disaster one of the main needs for the rescue teams and volunteer helpers is a functional communication infrastructure even during the first hours. The disaster recovery system (DRS) originally presented by us in a previous work [1] is based on an IEEE 802.11s wireless mesh network which is set up by non damaged legacy, mesh capable and battery powered devices still available in the disaster region. For lifetime enhancement, which is a key challenge for the proposed DRS based on battery powered devices, a distributed algorithm was proposed which enhances the lifetime of such a system by an approach which allows to shut down non necessary nodes and to keep them for a later usage while still keeping network connectivity. A simulation based proof of concept of this algorithm has been given in [1], however, employing a rather high level of abstraction. In this paper we evaluate the DRS network performance using a full implementation of the distributed algorithm within the simulation and which is based on alternating sleep and awake states for each individual node. Additionally, the algorithm is extended by a novel approach whereby the decision when a node changes its state from sleep to awake or vice-versa is based on the number of neighbors of the corresponding node, allowing for a much simpler implementation on the individual mesh nodes. The achievable lifetime enhancement first is evaluated through a Monte-Carlo simulation. Finally, the network performance achievable with the proposed lifetime enhancement algorithm for disaster recovery systems (LEA-DRS) is validated by a network simulation in ns-3 evaluating a Voice over IP connection between two nodes (e.g. between two rescuers).