Data aggregation techniques traditionally utilize cross-layer optimizations, which prove challenging to implement and are applicable only to a few related scenarios, while each new scenario requires a new approach. Proposed here is a framework that brings data aggregation to the application layer of the network stack while providing abstractions and defining interactions to make the framework flexible and modular. This framework, known as the data aggregation tunneling protocol (DATP), will make the benefits of aggregation attractive for many-to-one communication systems. The framework is modularized into the aggregation protocol, tree, function, scheduler, and end applications. DATP is considered in order to support the scalability of applications enabled by the smart grid. A complete DATP module was developed for the NS-3 network simulator and then tested in a smart grid distribution feeder scenario with optimal aggregation system parameters to determine the advantage gained by the system. Results show significant benefits for the aggregation system across a number of factors including the average message delay. Also developed was a data aggregation Internet application, which demonstrated the versatility of the application layer approach through an experiment in which thirty Internet-connected hosts joined the aggregation system and sent messages to a collector that gathered statistics about the system.