Data Sourcing Random Access using Semantic Queries for Massive IoT Scenarios

Abstract

Efficiently retrieving relevant data from massive Internet of Things (IoT) networks is essential for downstream tasks such as machine learning. This paper addresses this challenge by proposing a novel data sourcing protocol that combines semantic queries and random access. The key idea is that the destination node broadcasts a semantic query describing the desired information, and the sensors that have data matching the query then respond by transmitting their observations over a shared random access channel, for example to perform joint inference at the destination. However, this approach introduces a tradeoff between maximizing the retrieval of relevant data and minimizing data loss due to collisions on the shared channel. We analyze this tradeoff under a tractable Gaussian mixture model and optimize the semantic matching threshold to maximize the number of relevant retrieved observations. The protocol and the analysis are then extended to handle a more realistic neural network-based model for complex sensing. Under both models, experimental results in classification scenarios demonstrate that the proposed protocol is superior to traditional random access, and achieves a near-optimal balance between inference accuracy and the probability of missed detection, highlighting its effectiveness for semantic query-based data sourcing in massive IoT networks.