INGENIEUR
humidity , based on the input provided by sensors . The system analyses the situation in real-time and commands the actuators to open specific water valves located in places where soil humidity is below the pre-set value .
Communication Technology IoT devices can be connected using various communication technologies . Communication technologies play a key role in successful deployment of IoT systems . They can be grouped based on standards , spectrum and application scenarios . Communication standards can be classified into long-range and short-range communication standards . Communication spectrum can be divided into licensed and unlicensed spectrums . Meanwhile , the IoT devices application scenarios can be based on device roles and network topology .
Long-range communication standards can cover up to tens of kilometers and are classified as low power wide area ( LPWA ) such as NB-IoT , Sigfox and LoRA . Short-range standards cover a distance within 100m . Examples include Zigbee , Z-Wave , radio frequency identification ( RFID ) systems , Bluetooth and near field communications-enabled devices . Longrange , large-scale deployments of IoT networks have the highest impact on the spectrum . The main technologies for LPWA IoT in licensed spectrums are LTE-M1 ( LTE Cat MTC ), LTE-M2 ( NB-IoT ) and ( future ) fifth generation ( 5G ) IoT . In unlicensed spectrum , LoRa ( based on spread spectrum ) and SIGFOX ( ultra narrowband ) are the most prominent technologies for LPWA IoT and are currently deployed widely . The main contentions for comparison between licensed and unlicensed technologies are device cost , time to market , network coverage , reliability and security . The choice of communication technology also depends on the application of IoT devices — whether they act as nodes or as backhaul networks — and the type of topology to be deployed . There are different types of network topology such as peer-to-peer ( P2P ) or line , star , mesh , ring , tree and bus topology .
Communication standards and protocols play central roles to enable end-to-end intelligent agricultural applications . In advanced countries that have long adopted precision agriculture practices , the development and design of farm-specific data networks have the greatest impact on their crop production . Farm level communication standards such as DIN 9684 , ISO 11782 and CAN have been long established to facilitate connectivity through electronics control units ( ECU ) that control a combination of several working machines or agricultural implements .
Internet The Internet forms the core network layer , where paths are provided to transmit and exchange data and network information between multiple subnetworks worldwide , enabling data availability and accessibility anywhere and anytime . The Internet has paved the way for cloud computing , where large amounts of data are gathered for storage and processing . IoT middleware and connectivity protocols are being developed to achieve the connectivity of heterogeneous systems and devices over the Internet . Service-oriented architecture ( SOA ), cloud-based IoT middleware and actor-based IoT middleware are examples of middleware which have been applied to support IoT .
Data Management and Processing Units The complexity of agricultural data can range from structured to non-structured data , which can take the form of texts , images , audio and video . To better understand the requirements of IoT data analytics , we need to explore and compare the properties of IoT data with those of the general big data . IoT data exhibit four main characteristics :
1 . Large-scale streaming data ( such as crop canopies streaming images from flying drones );
2 . Heterogeneity ( data collected from various agricultural IoT data acquisition devices resulting in data heterogeneity );
3 . Time and space correlation ( sensors devices are attached to specific locations , and thus the corresponding data have location and time-stamp ); and
4 . High noise data ( due to tiny pieces of data in IoT applications or data collected in harsh environment like moving tractors or flying UAVs . These types of data are subject to errors and noise during acquisition and transmission ).
12 VOL 84 OCTOBER-DECEMBER 2020