Energy and coverage trade-offs in deploying a mix of mobile and stationary relays for disjoint Wireless Sensor Networks

Additional relay nodes (RNs) can be deployed within a partitioned Wireless Sensor Network (WSN) to restore connectivity among the partitions. In the case of lack of sufficient RNs to link all the partitions, some of the RNs can act as mobile data collectors (MDCs) to visit partitions for providing intermittent connection for the nodes. Determining the number of MDCs and stationary RNs is a challenge that deals with the trade-off between minimizing the maximum tour length and maximizing the coverage provided by both sensors and the stationary RNs. This paper proposes an RN placement algorithm to guarantee network connectivity while striving to balance the number of MDCs and stationary RNs so that the maximum tour length of MDCs is minimized and an imposed coverage constraint is satisfied. The proposed approach first determines Steiner points to connect partitions by using a Steiner Minimum Tree (SMT) heuristic. The number and location of these Steiner points to be occupied by stationary RNs is determined by using a cost function for each partition. The cost is computed based on the required RN count to connect a partition to the sink. The minimum cost partition is picked iteratively until the coverage constraint is met. The rest of the RNs are employed as MDCs which can tour the remaining partitions using existing algorithms. The proposed approach is evaluated with extensive simulations under a variety of conditions.