Politecnico di Torino - Corso Duca degli Abruzzi, 24 - 10129 Torino, ITALY

+39 011 090 6100 info@tech-share.it

Device sequencer for a wireless network

Bluetoothneural networksportable electronicsWiFiWirelesswireless device sequencer

Introduction

Portable electronic devices such as smartphones, laptops, sensors, and smartwatches may be connected to form a wireless sensor network. In such networks, a central device is configured to process the signals received from remote measuring devices. In such occasions, these devices, which sample physical quantities or physiological parameters, may be wearable and be in contact with the user’s body. These devices, in general, record a person’s physical parameters and send them to a central device for data processing. Hence, the data collected from different sources need to be synchronized to provide meaningful information. However, the performance of the central-to-remote device connection may be affected over time by factors, such as, operating system workloads, temperature, and wireless protocol issues, thus causing misalignment in the received data as the synchronization between the devices is affected. These discrepancies are typically hard to identify, as they are strongly dependent on the wireless technology used and the internal electronics of both central device and remote nodes; therefore, difficult to prevent/compensate in general.

Technical features

The inventors have found that by centrally synchronizing the transmission of wireless commands from the central node to the peripheral devices of a wireless network, using a neural network trained ad-hoc on the wireless link, they are able to counteract problems that may arise from asynchronization of devices on the same network. The method consists of plurality of peripheral devices and a central device executing a sequence of cycles of operations. Each cycle of operations is associated with a reference time stamp (clock) of the central device. The peripheral devices each receive a set of instructions in order to determine the time delay (if any). Thanks to machine learning, the wireless link latency errors are predicted to enable data reception at unison in the system.

Possible Applications

  • Wireless Sensor Networks;
  • Wearable electronics;
  • Remote measuring devices.

 

Advantages

  • Real-time monitoring of information exchange;
  • Provides a status of the wireless network.