L-MAC: a wake-up time self-learning MAC protocol for wireless sensor networks

Thanh Dinh, Younghan Kim, Tao Gu, Athanasios V. Vasilakos

Research output: Contribution to journalArticlepeer-review

40 Citations (Scopus)


This paper analyzes the trade-off issue between energy efficiency and packet delivery latency among existing duty-cycling MAC protocols in wireless sensor networks for low data-rate periodic-reporting applications. We then propose a novel and practical wake-up time self-learning MAC (L-MAC) protocol in which the key idea is to reuse beacon messages of receiver-initiated MAC protocols to enable nodes to coordinate their wakeup time with their parent nodes without incurring extra communication overhead. Based on the self-learning mechanism we propose, L-MAC builds an on-demand staggered scheduler to allow any node to forward packets continuously to the sink node. We present an analytical model, and conduct extensive simulations and experiments on Telosb sensors to show that L-MAC achieves significant higher energy efficiency compared to state-of-the-art asynchronous MAC protocols and a similar result of latency compared to synchronous MAC protocols. In particular, under QoS requirements with an upper bound value for one-hop packet delivery latency within 1 s and a lower bound value for packet delivery ratio within 95%, results show that the duty cycle of L-MAC is improved by more than 3.8 times and the end-to-end packet delivery latency of L-MAC is reduced by more than 7 times compared to those of AS-MAC and other state-of-the-art MAC protocols, respectively, in case of the packet generation interval of 1 min. L-MAC hence achieves high performance in both energy efficiency and packet delivery latency.

Original languageEnglish
Pages (from-to)33-46
Number of pages14
JournalComputer Networks
Publication statusPublished - 4 Aug 2016
Externally publishedYes


  • Wireless sensor networks
  • Duty cycle
  • MAC protocols


Dive into the research topics of 'L-MAC: a wake-up time self-learning MAC protocol for wireless sensor networks'. Together they form a unique fingerprint.

Cite this