![](https://csdnimg.cn/release/download_crawler_static/10887287/bg13.jpg)
10 Q. Le-Dang and T. Le-Ngoc
Serving as a backbone infrastructure, wired communication technologies such as
fiber optic networks are normally used at this layer due to their stability, large band-
width, and low delay characteristics. However, if the IoT components are mobile, for
instance, mounting on a service truck, wireless communication technologies need
to be used due to their flexibility and mobility support. In these cases, depending on
the traffic volume, different scenarios can be envisioned. First, for low bandwidth
IoT components (GPS, temperature sensors, etc.), LTE and LoRaWAN can be used.
LTE networks have wide coverage and higher bandwidth; however, LTE radios and
subscriptions are quite expensive so it is only suitable for a limited number of mobile
IoT devices. On the other hand, LoRaWAN networks offer wide coverage at low cost
but at narrow bandwidth. For large bandwidth IoT components (such as cameras
on road inspection vehicles), wireless communications are normally not adequate
or too expensive for real-time streaming and offline storage is required for latter
upload.
The data management layer is the central point that oversees the operation of
Smart City devices and data. This is where the collected data from the entire city is
stored and processed. It also provides interface to remotely monitor, configure, and
control IoT devices in the data acquisition layer. Usually, the logical components
of the management layer are separated into two subsystems: device management
(control plane) and data management (data plane). Device management system
provide facilities for collecting, storing status reports from deployed devices as
well as a uniform platform for controlling devices with different capabilities from
different vendors. Monitoring device availability, firmware update and reconfigu-
ration are some of the typical functionalities on this device management system.
Data management system facilitates data collection, storage, and visualization from
all the IoT devices in the network. Since there are multiple devices from different
vendors, with different capabilities, sensing different environmental factors, data
should be translated and stored in a unified format for ease of access and usage
afterwards. As this layer is critical to the Smart City infrastructure, it must
be protected against threats from the Internet, such as via the use of firewalls.
Additionally, the management layer maintains access control over devices and
data to prevent misuse of information and services provided by Smart City IoT
infrastructure.
To facilitate these functionalities, two main approaches can be identified based
on the placement of the data management platform: On-premise and cloud-based
approach. In the On-premise approach, all the above functionalities are hosted
on local computing facilities such as a data center. In this case, the operational,
administration and maintenance of these facilities will be on the shoulder of the
data center’s operator. This approach offers the benefits of performance and data
privacy. Since the data sink is placed close to its sources, the performance of
data archiving in terms of bandwidth and delay can be guaranteed easily. This is
especially advantageous for real-time data. In addition, since data is stored locally,
the operator has the complete control over the data, how to use it and its privacy
control. However, this approach can be costly to set up, maintain and scale up this