Reliable wireless data transfer for smart sensor shorts in football

Abstract

In football, most of the injuries occur in the lower extremities of the athletes. The leading cause of this is high muscle stress during explosive actions. To be able to prevent injuries, the Dutch Football Association (KNVB) and Dutch Hockey Association (KNHB) are working on smart sensor shorts in collaboration with the Delft University of Technology. The human movement scientists from the VU and RUG use these sensor shorts to develop a model which can predict potential incidents. The product’s goal is for the physical trainer to use it to monitor the team in real-time during a game or training.
The physical trainer will use a computer to analyze all the data. The sensor shorts consist of 3 Inertial Measurement Units (IMUs). Each IMU measures the angular velocity (dps), linear acceleration (g), and magnetic field (µT) with a ampling rate of 250 Hz. The high volume of data due to the high sampling rate induces challenges for wireless communication and design.
The sensor shorts can, at the moment, read out the IMUs and compress the sensor data. Besides, possible antennas for the sensor shorts have already been looked at. This thesis will continue with the communication link between the sensor shorts and the computer used by the physical trainer. The thesis will be focused on designing a reliable system for wireless communication. Multiple challenges must be faced to develop a reliable wireless communication system with minimal size. The challenges are the minimum amount of resources available, high data rate, large area to cover for communication, and obstruction due to the human body.
According to literature about monitoring devices currently described in the literature used for football, eight base stations are needed around the football field. The goal is to minimize the number of base stations so that the system can be easily set up. For the development of the system, first, the different Data Collection Protocols(DCP) are examined. Based on the network topology corresponding to the DCP, we want to choose a DCP that is the most efficient and reliable. The Direct Delivery protocol was chosen as most suited in combination with WiFi operating at 2.4 GHz. WiFi can deal with the high data rates and communicate over large distances.
The number of base stations needed is determined based on tests performed on the football field. The choice was made to place the antenna of the sensor shorts on the back of the athlete because there it has the most negligible chance of obstruction due to the athlete’s limbs. The athlete’s own body could cause attenuation of 10 dBm. Additionally, it was chosen to have the base station at 2.2 m height, to have fewer chances of obstruction due to the athletes on the football field. Multiple tests have been performed to determine how many base stations are needed. First, the possible signal strength values of the sensor shorts around the football field have been measured for a single base station. After that, a relation between the measured signal strength and possible loss and delay has been established. These results made it clear that more than one base station is needed. In tests with two base stations, it became clear that data might get lost when there is no base station in the radiation plane of the patch antenna of the sensor shorts. Therefore the choice was made to have four base stations, one on each corner of the football field, so that there is always a base station in the radiation plane of the patch antenna. Based on the test results, the choice was made to have four base stations, one on each corner of the football field. Assuming that the sensor shorts’ RSSI values are always higher than −80 dBm, the system with four base stations would have a loss of less than 5%. The upper boundary for the delay would be 5 s.