Browsing by Author "Majham, Mike"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Gs-mac: A scalable and energy-efficient mac protocol for greenhouse monitoring and control using wireless sensor networks
    (NM-AIST, 2023-08) Majham, Mike
    In recent years, wireless sensor networks have been widely applied in agricultural greenhouses to monitor and control farming-related parameters. These networks are composed of multiple sensor nodes, usually deployed in an ad hoc fashion. This form of farming attains the best quality and crop yield. However, the sensor nodes are energy constrained. So it is essential to reduce node power consumption to extend the network lifetime. An extensive analysis of weather conducted at Tanzanian locations show that weather parameters do not vary frequently. These characteristics motivate a duty cycling strategy, to minimize node power usage and increase the network lifetime. Therefore, this work proposes Greenhouse Sensor MAC (GS- MAC), a medium-access-control (MAC) protocol designed for greenhouse monitoring and control: energy conservation and scalability are the primary objectives, with latency being less crucial. To minimize idle listening, nodes implement a Time Division Multiple Access (TDMA) scheme and periodically sleep. Nodes in close proximity organize into clusters. Unlike traditional duty cycling techniques, GS-MAC avoids periodic node synchronizations. Instead, nodes communicate by maintaining strict schedules provided by cluster heads, to avoid collisions, over-emitting and minimize the duty cycle. GS-MAC also uses short node addresses to reduce packet overheads. Finally, GS-MAC adopts a contention approach on reserved time slots allocated between communication rounds to maintain scalability. GS-MAC has been implemented on MATLAB software, with simulation parameters (i.e., sensor nodes' current consumption characteristics) obtained from actual hardware. The experiment results show that GS-MAC extends the network lifetime by at least 2.7 times more than previous research, with traffic loads sent every 1-60 seconds