The Mobile Wireless Grid (MOWG) provides a platform for distributed computing and ad-hoc networking in the battlefield space with limited to non-existent external network connectivity through its framework to network multiple smartphones into a cohesive grid, partition complex data and compute problems across the available nodes, and combine results into a single cohesive solution. It thereby enables execution of complex clinical decision support, machine learning, and other algorithms that are too resource intensive to run on a single device in scenarios where it is not feasible to run such algorithms in the cloud. MOWG can work with an existing network of smartphones, or form an ad-hoc network on demand. It monitors the capabilities and constraints of the grid, including the number of devices and their hardware specifications, network speed and latency, other applications competing for device resources, and power constraints. Given these parameters, it dynamically assigns tasks to workers, optimizing both to maximize performance and minimize impact on the devices’ primary uses.
MOWG enables distributed computing in austere battlefield environments with limited to non-existent external network connectivity. There are a wide variety of distributed computing problems to which MOWG can be applied, and to ensure extensibility, MOWG is at its core a framework for deploying distributed computing applications on mobile devices. It handles the challenges of networking the devices into a cohesive grid and provides a MapReduce-like framework and an application programming interface (API) through which applications can be developed to utilize the capabilities of a mobile grid.