This thesis provides a theoretical analysis of the effects of mobility, node density and a limited transmission range on the connectivity of a varying density of nodes in wireless sensor networks. Connectivity in cellular networks has the advantage of a fixed centralised infrastructure that can provide wide communication coverage. Wireless sensor networks, on the other hand, have a limited range. This limited range, coupled with nodes’ mobility, often results in network holes. As the architecture is de-centralised, there is no central node that monitors the nodes’ joining or leaving the network. The challenge of identifying these nodes, which is due to their dynamic nature of movement, is presented here.

Opportunistic connectivity addresses the challenge of providing connectivity to isolated mobile nodes. This is through the process of discovery of regions where good density of network nodes are available. The concept involves four key components. These are adaptive sampling, coverage, handoff and directional communication. These act on the minimisation of energy cost incurred with the discovery of related nodes and establishment of connectivity in the network. The window of time for communication is extended in an energy–efficient manner through coverage, handoff and direction for such delay–tolerant networks. The overall contribution of this thesis is a protocol design for opportunistic connectivity, its implementation and analysis, with reference to the conservation of energy and reduction of packet drops, in conjunction with protocol testing on an application scenario.

The thesis is structured into seven chapters. The first two chapters provide the background and the literature analysis. The third chapter deals with systems and tools which are used for the modelling and testing. It gives an insight into the different available tools and their ability to validate the parameter of our concept of an opportunistic connectivity protocol. Subsequently, the thesis discusses the design of the ‘adaptive Energy COnscious DElay Tolerant OpportUnistic Routing’ (ECO-DETOUR) protocol for such delay–tolerant networks in chapter four, as a four stage process involving adaptive sampling, coverage, direction and hand-off. Design of the protocol is followed by implementation in chapter five, which was performed using the OPNET and MATLAB environments. The chapter details the different conditions in which each of the four parameters are triggered and discusses the implementation of each of the four parameters as pseudo-code. Finally in chapter six the protocol is tested on a wildlife application scenario. The effectiveness of the protocol is measured in relation to the energy saved and the reduction in number of packet drops achieved under different mobility conditions. Results show that ECO-DETOUR achieves a 45% - 60% reduction in expended energy to set up communication and exchange data packets. The bulk of the saving in energy by the ECO-DETOUR protocol comes from adaptive sampling which is followed by coverage, handoff and direction.