PoC based on Amarisoft equipment/software

The following pictures show an example of an Amarisoft scenario with 5G SA, MIMO 2x2, 50 MHz in the n78 band (TDD). The Quectel RM500Q-AE modem is able to achieve more than 350 Mbps (downlink). Private repository here.

PoC based on srsRAN platform

This testbed employs an Ettus B210 (NI USRP 2901) connected to an Intel NUC. UE is a Quectel RM500Q-GL modem connected to another Intel NUC. It supports 4G, 5G NSA (with srsEPC) and 5G SA (with Open5GS) with both real equipment and virtual hardware (ZMQ virtual radios). Private repository here.

Also supporting 4G with a Raspberry Pi 4 (low bandwidth). Private repository here.

Multi-connectivity demonstrator using real equipment

This testbed is part of our work for the 5G-CLARITY European project. To be published.

The following picture presents the topology of our multi-connectivity testbeds, including a 5G network, a Wi-Fi 6 network, and the required MPTCP entities.

The following picture shows our demonstrator.

Multi-connectivity virtual testbed

Multi-connectivity virtual testbed (used e.g. for the 5G-CLARITY European Project). Available here.

The following picture summarizes one of the scenarios, where the CPE (VM mptcpUe) is connected using 3 network interfaces (5G NR, Wi-Fi and Li-Fi) to the MPTCP proxy (VM mptcpProxy) through the Non-3GPP Interworking Function (N3IWF) of a 5G Core Network (implemented using free5gc).

The following picture summarizes another scenario, where the CPE is connected to several MPTCP proxies. The CPE acts as a switch (using OVS) for all the clients that may be connected, and uses different VLAN IDs to route the packets to the server through the different proxies (one VLAN per proxy).

You can check the following video with the execution of this testbed. Available here.

Basic LoRaWAN testbed

This basic LoRaWAN testbed is included in a private repository here.

Public repositories including sample firmware for different devices:

• Repository for TTGO LoRa32 here
• Repository for TTGO T-Beam Tracker here
• Repository for LilyGO T-HiGrow + T-HiGrow Lora Sheld here
• Repository for LilyGO T-HiGrow with Wi-Fi here

Demonstrator for LoRaWAN network slicing

To be added soon (once it is published).

PoC to improve LoRaWAN network capacity (for the paper "Collision Avoidance Resource Allocation for LoRaWAN")

Testbed using LoRaWAN gateways and nodes to implement the solution to reduce collisions (i.e. increase LoRaWAN network capacity) in our paper "Collision Avoidance Resource Allocation for LoRaWAN". Micropython code for Pycom's FiPy, to implement the device's part, available here. The following picture presents the testbed.

IoT network architecture demonstrator

Testbed using LoRaWAN gateways and nodes, connected to an SDN network that implements the solution to reduce MQTT traffic using multicast in our paper "A LoRaWAN Network Architecture with MQTT2MULTICAST". Implementation of the SDN network and the MQTT2MULTICAST protocol available here. The following picture summarizes the proposed architecture.

The following picture presents our proof-of-concept for the proposed network architecture.

PoC for multimedia transmission over LoRa

Solution to implement a contention mechanism in order to increase the data rate and thus being able to send images over LoRa. In this testbed, a sender takes a picture from a camera and from an SD card and sends it over LoRa to a receiver, which in turn sends it over Wi-Fi to a streaming server. This allows to send low-resolution images or low-quality videos over LoRa devices, which may be useful in many situations. Current tests show that an image of 10KB can be sent in around 4-6 seconds without contention and around 10-12 seconds with contention. The repository with the code will be made available here once this work is published. The following picture presents the testbed. You can check how the testbed works on the following videos, using contention here and without contention here.

Demonstrator for a LoRaWAN network supporting critical situations

Testbed that implements the solution in our paper "A LoRaWAN Testbed Design for Supporting Critical Situations: Prototype and Evaluation". This testbed is able to automatically recover if part of its network infrastructure is destroyed, e.g. after an earthquake. We implement the core network functionalities on top of a Kubernetes cluster (using containers) and also using OpenStack (using virtual machines) for comparison purposes.

The following pictures present our proof-of-concept for the proposed network architecture using real equipment (TTGO LoRa32 nodes and IMST Lite gateways) for the radio access network.

Weather station at the School of Informatics and Telecommunications Engineering

We have deployed a weather station from Bresser (BRESSER professional 7-in-1 Wi-Fi Weather Station) which will allow us to get environmental data (temperature, humidity, barometric pressure, wind speed and orientation, precipitation rate, UV level and light intensity) and to generate real traffic. It transmits data via 868 MHz to the base station, which stores historical local weather data shown in its display. The data sent via 868 MHz can be easily forwarded using LoRaWAN thus converting it in a LoRaWAN-enabled weather station, which can be used in our testbeds to generate LoRaWAN traffic using realistic data.

The following pictures present our weather station. Historical data is available at Weather Underground, Weather Cloud and AWEKAS.

We have converted this weather station onto a LoRaWAN-enabled device by using an ESP32 as a gateway between 868 MHz (FSK) and LoRaWAN. The first step was to include a Bresser 7-in-1 decoder in the BresserWeatherSensorReceiver Arduino library. The second step was to include program a LoRaWAN device (a Heltec Wireless Stick) to receive weather information via 868 MHz and then sending that information (using Cayenne LPP encoding) to a LoRaWAN network (TTN or our LoRaWAN testbed). This work is included in a private repository. The following picture shows the logs (from the serial monitor in Arduino) and the data received in The Things Networks.

We have also created a platform to show graphs with the weather station data. Since the weather station transmits these data to TTN using LoRaWAN, which in turn publishes the measurements using MQTT, our platform acts as a gateway between MQTT and an InfluxDB database, whose data is finally shown using a Grafana dashboard. This platform is available at a private repository. The following picture shows an example of this Grafana dashboard (including temperature, humidity, wind, rain, UV and light data, in addition to SNR, RSSI and battery information).

Demonstrator for TSN

The following picture shows our initial testing for our TSN network demonstrator. [UNDER CONSTRUCTION]

UAVs for demonstrators

This picture shows one of our DJI TELLO drones, which will be used for demonstrators (e.g. to test LoRaWAN to send pictures over the air, or real-time services over mobile networks).

This picture shows one of our AWS DeepRacer Evo with stereo cameras and a LiDAR sensor, which will be used for demonstrators (e.g. to test LoRaWAN to send its location, or to send environmental information via LoRaWAN).

Testbed for the 5G-CITY project

This picture shows our testbed with the partners from the 5G-CITY project. The different isles were connected using OpenVPN, which allowed us to have L2 connectivity. With this testbed, UGR contributed to two SDN-related papers ("An architecture for the 5G control plane based on SDN and data distribution service" and "The Roll of Data Distribution Service in Failure-aware SDN Controllers") and one LoRaWAN-related paper ("A LoRaWAN Testbed Design for Supporting Critical Situations: Prototype and Evaluation"), in addition to other experiments.