View article

To address the “big data” problem introduced by machine-to-machine (M2M) applications, we argue in this paper that instead of focusing on serving individual machines with better quality, one should focus on solutions that can better serve the data itself. To substantiate, we consider a data gathering scenario, where machines perform source coding for collected data and communicate directly with the aggregator for sending back the data. Due to limited radio resources, however, it is not possible that all machines are allocated resources for uplink transmissions. To address the problem, we propose an algorithm for joint node selection, resource allocation, and source coding under resource constraints. The objective is to minimize information loss, instead of the conventional ones of maximizing the sum rates or maximizing the number of supported machines. Simulation results indicate significant benefits of …