Page 113 - ITUJournal Future and evolving technologies Volume 2 (2021), Issue 1

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 1

is selected for every use case .

8.5 Route propagation

Route propagation occurs through two mechanisms: pig‐
gybacking and control packets. Piggybacking allows
routes to be shared without dedicated transmissions. Fig. 8 – Experimental setup.
Considering a RODENT packet carrying a data payload
of use case , the header contains RV ’s id number and WiFi
x

the best route BR . Wireless communications share a LoRa WiFi BS LoRa BS

common medium. Thus, overhears every packet from BLE

NBR(i), which allows to update RM opportunisti‐
i

cally. If stops overhearing route from its neigh‐

ij

bor NBR(i) e.g., because is down, will time out and
ij

j
will be removed from RM . To keep alive unused routes,
i
NBR(i) will send dedicated control packets. Control pack‐ N3 N1 N2 N4
j
ets are regular packets with an empty payload.
9. PROTOCOL IMPLEMENTATION
N5
Our implementation of RODENT is performed on Pycom
FiPy devices [21]. The speci icity of FiPy devices is that
they offer ive different RATs. These nodes take part in
the MTN and of load data to WiFi and LoRa base stations Fig. 9 – Farm monitoring scenario.
(BS). The hardware and irmware used are detailed in this
section. mented: RV monitoring and RV alarm .
The LoRa BS’s irmware is implemented in C. It listens
constantly for LoRa transmissions. Upon reception of a
9.1 Hardware
RODENT packet, it is unpacked and its characteristics are
Pycom FiPy nodes are composed of WSN hardware: wire‐ printed on the serial port. The WiFi BS is coded in Python.
less RAT, ESP32 CPU, little memory available which al‐ It listens for RODENT WiFi transmissions, unpacks them
lows ultra–low power usage. The available RATs are WiFi, and prints characteristics on stdout.
LoRa, Sigfox, LTE‐M, NB‐IoT and Bluetooth Low Energy
(BLE). Each RAT comes with different performances in 10. RODENT EXPERIMENTS
terms of energy consumption, economical cost, bit rate,
etc. RODENT performs route selection based on these To assess the performances of RODENT, we run experi‐
characteristics. FiPy nodes are coupled with Pytrack sen‐ ments on real hardware. We con igured the nodes to fol‐
sor shields which provide an accelerometer, a GPS and a low a speci ic scenario and measured the results. The ex‐
micro‐USB port. perimental setup and scenario are presented in this sec‐
The LoRa BS is a B‐L072Z‐LRWAN1 board [22]. The WiFi tion.
BS is an Edimax EW‐7811Un dongle [23] connected to the
main computer. A Trip Lite U223‐007 (7‐Port USB Hub) is 10.1 Setup
used to connect every device. The main computer is a Dell
Latitude 5590. It powers devices, collects and analyses The aforementioned devices in Section 9 are connected
results. to the main computer through the USB hub. Every node
and BS are powered at the same time and boot up imme‐
9.2 Firmware diately. As we can see in Fig. 8, every device is laying very
close to each other. The main computer reads the stdout
A port of MicroPython available as irmware for the FiPy of the WiFi BS and the serial ports of the nodes and LoRa
allowed us to implement RODENT in Python. Upon boot, a BS. Results are then computed of line, post‐experiment.
node computes its unique ID. Based on LM it boots up
i

the needed RAT and constructs routes. The node is then 10.2 Scenario
locked up in the main loop: i) select best route for each
RV , ii) add next payload to transmission buffer iii) send We simulate a farm monitoring use case. Smart agricul‐
x
every payload in buffer. Neighbor’s routes are added in ture can help farmers in their everyday life, but farms
RM upon reception. Neighbor’s payloads are appended are often an unfriendly environment for wireless sensors
i
in the transmission buffer. Nodes print on the serial port (large rural areas, tall crops...). MTN eases the techni‐
the characteristics of the packets sent. Upon the Pytrack’s cal dif iculties by offering nodes multiple possibilities of
button press, nodes switch between the two RV imple‐ communication (operator based networks, personal net‐

© International Telecommunication Union, 2021 97