Optimal channel selection for real-time uplink data transmissions in ambulances

  • Authors:
  • Ana Goulart;Wei Zhan;Robert Arnold

  • Affiliations:
  • Electronics and Telecommunications Engineering Technology Program, Texas A&M;Electronics and Telecommunications Engineering Technology Program, Texas A&M;The Academy for Advanced Telecommunications and Learning Technologies, Texas A&M

  • Venue:
  • Journal of Mobile Multimedia
  • Year:
  • 2009

Quantified Score

Hi-index 0.00

Visualization

Abstract

Through wireless internet access, ambulances take advantage of the widespread cellular coverage in rural and urban areas to transfer audio, video, and vital signs to the emergency room. In the current implementation adopted by the DREAMS™ ambulances, a designated channel is selected randomly. It is used to transmit high priority data, such as vital signs and audio. The remaining channels are used for video transmission. Whenever the communication system in the ambulance detects a certain threshold of packet losses in the designated channel, the communication system randomly switches the high priority data to another wireless channel. However, the designated channel selection process does not necessarily select the best available channel. The objective of this paper is to optimize this process through proportional-integral-derivative (PID) control and optimization with feedback. For each channel, an objective function is calculated. It includes a derivative term for fast response and an integration term for detection of small but consistent differences between channels, in addition to the proportional term. Using a causal real-time optimization algorithm, the maximum objective function is continuously selected. Thus, the proposed optimal channel selection algorithm enables the ambulance's communication system to intelligently shift the load to better quality channels without detailed information about the channels. The proposed algorithm combines key performance metrics (i.e., reliability and effective transmission rate), which can be calibrated with different weights. Using data from simulation and experiments in commercial cellular networks, we compare the performance of the ambulance with and without the algorithm to show dramatic improvements in the reliability and throughput of the ambulance's uplink transmissions.