Connectivity restoration in disjoint wireless sensor networks using limited number of mobile relays

Disjoint Wireless Sensor Networks (WSNs) can be reconnected by placing additional relay nodes in the damaged areas. However, in some cases there may not be enough relays to reconnect all the partitions with the sink node. In such a case, some of the relays can exploit their motion capabilities and temporarily act as a mobile data collector (MDC) between partitions providing intermittent connectivity for the nodes sitting in those partitions. Nonetheless, due to increased data latency intermittent connectivity creates, the number of such MDCs need to be minimized. On the other hand, given that the energy resources for an MDC is limited, an upper bound on the travel distance overhead for an MDC needs to be imposed. This paper proposes a relay placement algorithm which guarantees connectivity by maximizing the number of stable connections while meeting the maximum tour constraint on the MDCs. The approach first determines the number and location of relays to restore connectivity by establishing stable links using a Steiner Minimum Tree (SMT) heuristic. Assuming that the number of available relays is less than the needed count, the algorithm determines how many of the available relays need to be stationary and how many of them should act as MDCs. By initially assuming all relays as MDCs, an iterative procedure is followed to reduce the MDC count while meeting the maximum tour length constraint. Specifically, groups of partitions are created and assigned to MDCs for touring. The proposed approach is validated with extensive simulations under a variety of conditions.