QoS-Heterogeneity and Buffer Management in Constrained Intermittent-Connected Mobile Sensor Networks

In intermittent-connected mobile sensor networks, generally nodes have a small number of neighbors and connections between the nodes rarely exist. This paper presents a new networking system for this type of network, aimed at maximizing resource utilization to achieve better system performance in terms of delay time and reliability. The system is motivated by the observation that a large number of sensor network applications, such as road monitoring sensor systems and animal tracking applications, have two kinds of co-existing data packets: One with soft realtime constraints while the other need to be delivered to the base station with high probability. Rather than focusing on energy consumption and network lifetime, other crucial issues are to design the communication system to satisfy the application requirements, and to manage the local buffer storage to avoid potential buffer overflow and data corruption or loss. Our proposed system achieves the above goals and makes significant improvements on system resource optimization. It is simulated in Matlab and evaluated under various practical scenarios. Simulation results demonstrate that our QoS-heterogeneous delay-tolerant routing approach performs better than ordinary epidemic routing and probabilistic forwarding strategies when packets with different types of QoS requirements exist in the network.