Publication
Chaos: Versatile and Efficient All-to-All Data Sharing and In-Network Processing at Scale
An important building block for low-power wireless systems is to efficiently share and process data among all devices in a network. However, current approaches typically split such all-to-all interactions into sequential collection, processing, and dissemination phases, thus handling them inefficiently. We introduce Chaos, the first primitive that natively supports all-to-all data sharing in low-power wireless networks. Different from current approaches, Chaos embeds programmable in-network processing into a communication support based on synchronous transmissions. We show that this design enables a variety of common all-to-all interactions, including network-wide agreement and data aggregation. Results from three testbeds and simulations demonstrate that Chaos scales efficiently to networks consisting of hundreds of nodes, achieving severalfold improvements over LWB and CTP/Drip in radio duty cycle and latency with almost 100 % reliability across all scenarios we tested. For example, Chaos computes simple aggregates, such as the maximum, in a 100-node multi-hop network within less than 90 milliseconds.