Data Distribution Management in Mobile Networks

The patent introduces roadside units as intermediary devices that store data sets and enable local peer-to-peer transfers between mobile compute platforms. These intermediaries allow data to be exchanged directly between nearby devices without traversing the cellular network, thus maintaining data distribution reliability while dramatically improving transfer efficiency through local communication links.

The patent segments the data distribution system into multiple components: centralized data storage, roadside unit intermediaries, and mobile compute platforms. By dividing the system into these segments, data can be distributed through multiple pathways (cellular network to roadside units, then local transfers between mobile devices), improving overall efficiency while maintaining reliability through redundant distribution channels.

The patent implements preliminary action by having mobile compute platforms predict future proximity events based on current trajectory and velocity data. Before devices actually come into range, the system pre-establishes transfer plans and prepares data sets, so that when connections are made, transfers can begin immediately without connection instability affecting the process.

The patent applies dynamics by making the system adaptive to changing mobile device positions and connection states. The proximity prediction mechanism continuously updates based on real-time motion data, and transfer decisions are dynamically adjusted based on predicted proximity events, allowing the system to maintain reliability despite the inherently unstable nature of mobile ad-hoc connections.

The patent uses inexpensive, computationally lightweight trajectory and velocity data that are already available from standard mobile device sensors and GPS. Rather than implementing complex prediction algorithms, the system uses simple kinematic calculations based on readily available data, achieving effective proximity prediction without significant computational overhead.

The patent changes the parameters used for prediction from complex multi-factor models to simpler physical parameters like velocity, trajectory, and time-to-contact. By focusing on these fundamental motion parameters, the system achieves effective proximity prediction with minimal computational complexity, as these parameters are directly measurable and require simple mathematical relationships.