When are network coding based dynamic multi-homing techniques beneficial?

Mechanisms that can cope with unreliable wireless channels in an efficient manner are required due to the increasing number of resource constrained devices. Concurrent use of multiple communications technologies can be instrumental towards improving services to mobile devices in heterogeneous networks. In our previous work, we developed an optimization framework to generate channel-aware transmission policies for multi-homed devices under different cost criteria. Our formulation considers network coding as a key technique that simplifies load allocation across multiple channels and provides high resiliency under time-varying channel conditions. This paper seeks to explore the parameter space and identify the operating regions where dynamic coded policies offer most improvement over static ones in terms of energy consumption and channel utilization. We leverage meta-heuristics to find different local optima, while also tracking the intermediate solutions to map operating regions above 3 dB and 5 dB. Our results show a large set of relevant configurations where high resource efficiency can be obtained with the proposed transmission mechanisms.