A Social Potential Fields Approach for Self-Deployment and Self-Healing in Hierarchical Mobile Wireless Sensor Networks

Eva González-Parada, Jose Cano-García, Francisco Aguilera, Francisco Sandoval, Cristina Urdiales
2017 Sensors  
Autonomous mobile nodes in mobile wireless sensor networks (MWSN) allow self-deployment and self-healing. In both cases, the goals are: (i) to achieve adequate coverage; and (ii) to extend network life. In dynamic environments, nodes may use reactive algorithms so that each node locally decides when and where to move. This paper presents a behavior-based deployment and self-healing algorithm based on the social potential fields algorithm. In the proposed algorithm, nodes are attached to low
more » ... robots to autonomously navigate in the coverage area. The proposed algorithm has been tested in environments with and without obstacles. Our study also analyzes the differences between non-hierarchical and hierarchical routing configurations in terms of network life and coverage. allow one to adjust the transmission power levels at runtime, but only within a restricted interval, so power control techniques are usually combined with more complex approaches [6] . Some of them are based on strategically adding some redundant nodes so that there is more than one routing path between every pair of sensors in the network (k-vertex connectivity) [7] . The main challenge in these cases is the complexity of planning node deployment using a reduced number of redundant nodes. Alternatively, other methods propose spares for critical nodes. These spares could be passive extra nodes or active nodes nearby the critical ones. In this second case, it is necessary to select as the replacement the node that would cause the least degradation in the network [8] . Some methods work in a reactive way: rather than planning over the full topology, nodes are rearranged according to local criteria, like preserving connectivity or minimizing coverage loss in a given area. In these cases, algorithms can work with as little as one-hop information to detect the failure of critical nodes and decide the best strategy to recover from it [6] . In its simplest implementation, a failed node could be replaced by its nearest uncritical neighbor [9], whereas more complex approaches aim at minimizing movement overhead in order to extend network lifetime [10] . Other proposals deal with routing information to provide short paths between remaining nodes, as well [11] . These techniques are typically used in mobile WSN (MWSN), where nodes have some degree of mobility. MWSN can autonomously deploy themselves and also adjust their positions if part of the network fails (self-healing) [12] . MWSN are useful for long-term monitorization of large areas and also for emergency deployment of communication networks. In both cases, mobile nodes allow adaptation to changing conditions, including the specifics of the area and also the lifetime of the different nodes. Multiple robot systems (MRS) are adequate for MWSN [3, 13, 14] . In these systems, robots move to achieve the best possible coverage using all living nodes. Some approaches for deployment in WSN are based on deliberative algorithms to optimize efficiency [15] [16] [17] [18] and also for self-healing [3, 19] . Similarly, self-healing may rely on the deliberative relocation of nodes in the network [20]. Deliberative algorithms reason over a model of the environment. Hence, they tend to be computationally expensive and require information about the problem instance, including network configuration, environment layout, traffic, etc. Instead, reactive deployment assumes that local dispersion leads to global dispersion: hence, each node makes its own decisions based on local factors [21] [22] [23] [24] [25] . Reactive deployment is not optimized, so some features, like path redundancy, shortest path to the sink, etc., cannot be guaranteed. However, it is computationally less expensive and, hence, better suited for operation in unknown environments, adaptation to failure and also to dynamic structures where nodes can move. A popular approach to reactive deployment in MWSN is behavior-based deployment. In behavior-based algorithms, a node relies on several nuclear skills (behaviors). Each behavior associates an input instance to an output action. More complex, emergent behaviors are obtained as the combination of several simpler ones. A node stops moving when the combination of all its behaviors return a null vector. All virtual potential and forces approaches are implementations of behavior-based algorithms [23] [24] [25] . Alternatively, deployment could follow rules rather than behaviors (e.g., [26] ). Rule-based deployment is better fitted to deploy into a given topology, e.g., bus, backbone or ring networks, and also to impose hard constraints, e.g., fix a number of beacons for RSS-based localization [27] . However, behavior-based deployment adapts better to the environment, and it is more adequate for self-healing, especially for multi-node failure, since no fixed node structure needs to be preserved [28] . This work proposes a new behavior-based algorithm for deployment and self-healing of MWSN. The proposed algorithm is a variation of the social potential fields (SPF) [29] , originally proposed for navigation in a robot swarm. Force-based strategies have also been proposed for deployment, e.g., [23] , or for autonomous repair, e.g., [30] . In one case, forces tend to move nodes away from each other, whereas in the other, healthy nodes tend to move towards the center of the deployment area (assuming it is known). Our proposal is valid for both deployment and self-repair simultaneously,
doi:10.3390/s17010120 pmid:28075364 pmcid:PMC5298693 fatcat:f3ig3m7xcnf73fapmnkd2idxbe