Client-Based Control Channel Analysis for Connectivity Estimation in LTE Networks

Advanced Cyber-Physical Systems aim for the balancing of restricted local resources of deeply embedded systems with cloud-based resources depending on the availability of network connectivity: in case of excellent connectivity, the off-loading of large amounts of data can be more efficient than the local processing on a resource- constraint platform, while this latter solution is preferred in case of limited connectivity. This paper proposes a Client-Based Control Channel Analysis for Connectivity Estimation (C3ACE), a new passive probing mechanism to enable the client- side to estimate the connection quality of 4G networks in range. The results show that by observing and analyzing the control traffic in real-time, the number of active user equipment in a cell can be determined with surprising accuracy (with errors well below 10^-6). The specific challenge addressed in this paper lies in a dedicated filtering and validation of the DCI (Downlink Control Information). In a subsequent step the data rates to be expected can be estimated in order to enable decision about the choice of network and the timing of the data offloading to the cloud. The proposed methods have been implemented and validated leveraging the SDR OpenAirInterface, a real- life LTE network and a distributed load generator producing a scalable network traffic by a number of LTE User Equipment.