Energy Minimization Algorithm for Estimation of Clock Skew and Reception Window Selection in Wireless Networks

Area of Science:

• Computer Science
• Electrical Engineering
• Wireless Communication Systems

Background:

• Time synchronization is a critical challenge in wireless networks, especially for energy-constrained, battery-powered devices.
• Maintaining synchronized clocks is essential for reliable communication and efficient operation.
• Disconnections and clock skew can significantly degrade network performance.

Purpose of the Study:

• To develop methods for re-establishing connections with devices that have experienced long-term disconnections and unknown clock skew.
• To propose and evaluate novel approaches for follow-up clock synchronization using the confidence interval method.
• To optimize energy utilization while maximizing the probability of message reception in wireless networks.

Main Methods:

• Analysis of clock skew measurements.
• Development and simulation-based evaluation of algorithms for connection establishment and clock synchronization.
• Exploration of confidence interval methods for precise time alignment.
• Parameter selection analysis balancing energy consumption and reception probability.

Main Results:

• A method for re-establishing connections with long-disconnected devices with unknown clock skew was proposed.
• Two confidence interval-based approaches for follow-up clock synchronization were presented and evaluated.
• Extensive simulations demonstrated the effectiveness of the proposed algorithms.
• A growing time window size was shown to be more energy-efficient than repeated fixed-size windows for packet reception.

Conclusions:

• The proposed methods effectively address challenges in time synchronization for battery-powered wireless devices.
• Optimized parameter selection, including adaptive window sizing and shifting, can significantly reduce energy consumption.
• The study provides valuable insights for designing energy-efficient and reliable wireless communication systems.