Date(s) - 18 Apr 2018
1:10 PM - 2:00 PM
3043 ECpE Building Addition
Speaker: Ling Wang, ECpE Graduate Student
Adviser: Hongwei Zhang
Title: Distributed Scheduling and Power Control for Predictable IoT Communication Reliability
Abstract: Wireless networks are stepping into a new era. The vision of ubiquitous connection is boosting the emergence of new networks such as Internet of Things, vehicular networks and 5G. Compared to traditional human-oriented cellular networks with emphasis on capacity, these emerging networks and applications have placed stringent requirements on network latency and reliability. Mission-critical IoT applications such as wireless networked industrial control require reliable wireless communication over 99.999%. Due to co-channel interference and wireless channel dynamics (e.g. multi-path fading), however, wireless communication is inherently dynamic and subject to complex uncertainties. Joint scheduling and power control has been explored for reliable wireless communication, but existing solutions are mostly centralized or do not consider real-world challenges such as fast channel fading. Towards a foundation for mission-critical IoT application, we develop a distributed, field-deployable approach to joint scheduling and power control that adaptively regulates co-channel interference and ensures predictable IoT communication reliability in the presence of wireless communication dynamics and uncertainties. The proposed approach effectively leverages the Perron-Frobenius theory, physicalratio-K (PRK) interference model and feedback control for PRK model adaptation and transmission power update. We have shown that our approach improves concurrency by 70% than state of art fixed
scheduling while ensuring successful SINR tracking over time. As the first distributed scheduling and power control scheme that ensures predictable wireless communication reliability while considering realworld challenges such as fast channel fading, this work is expected to serve as a foundation for real-world deployment of mission-critical IoT systems.