Wavelength-selective fog-computing network for big-data analytics of Wireless Data

Smartphones generate a lot of data that can be analyzed. In a typical system, the data is input into the phones, converted to a wireless signal, received by a base station, and is converted to an electrical signal that can be sent to the cloud for processing. One step to speed up the task was to offload some of the processing to fog nodes. These fog nodes still communicate with a typical electrical trunk network. Our proposed architecture removes a large portion of the delay in the system by removing the hop to hop communication methods that are typically seen in implemented systems. The proposed architecture is shown to have a lower average and worst case delay in all cases but does not achieve as much throughput as a wavelength division multiplexed optical network is capable of. Throughput can be improved by running multiple lines, whereas the latency can not be significantly improved by adding extra hardware.