June 2021

Testing Virtual Cloud Environment

In June 2021 we have progressed towards testing and finalizing our experiment workflow and the virtual cloud environment. The core and radio access network of the Ericsson 4G/5G system have been deployed at Lake Wheeler, and internal testing has started in coordination with Ericsson engineers. AERPAW platform was presented at various different venues in June 2021, including:

May 2021

Ericsson Network Deployment

This month we installed Ericsson 4G/5G RAN equipments at the Lake Wheeler site (LWL) and the Core equipments at the AERPAW Network Control Center (NCC) at NCSU Centennial campus, and integrated with the AERPAW fiber backplane (backhaul) between the LWL site and AERPAW NCC. The RAN includes a) one 2-sector radio at 3.7GHz band with two external antennas, b) two radios in 1.7/2.1 GHz FDD band with builtin antennas that makes two overlaid sectors, all installed on the tower, and c) a 6U 10KW enclosure on the ground with a BBU and Router that connects to the backhaul (AERPAW backplane switch at the LWL). Similarly, the Core network at the AERPAW NCC includes a) an EPC Dell server, b) an App Dell server, and c) a router that connects to the AERPAW switch for the other end of the backhaul and the AERPAW Gateway for remote access.

The current deployment configuration is two sectors of 5G NSA network (Option B) with two 5G TDD sectors in 3.7GHz band overlaid with two LTE FDD sectors in 1.7/2.1 GHz band. Alternatively, the network can be configured to all 4G sectors (Option A), i.e., two LTE TDD sectors in 3.7 GHz band and two LTE FDD sectors in 1.7/2.1 GHz band, or simultaneous 4G & 5G one sector (Option C) with one 4G LTE sector at 3.7GHz overlaid with a 5G NSA sector with 5G NR in 3.7 GHz band and an LTE in 1.7/2.1 GHz band. After the initial deployment of all the equipment, both RAN and Core, the remote access through AERPAW Gateway and the transport network backhaul connectivity between the Core and RAN through AERPAW backplane switches were established and the testing is in progress. After testing the transport network and the remote connectivity, the 5G wireless network testing with 5G devices will follow.

April 2021

Field Testing Continues

This month we continue to expand drone tests and check accuracy/reproducibility. The first figure shows a Large AERPAW Multicopter (LAM) with a portable node during the data collection phase for an IMU notch filter. Our current fleet of UAVs stretches their arms: three LAMs, one Small AERPAW Multicopter (SAM), and the DJI Phantom 4 we use for taking videos. All three LAMs (as well as the SAM and DJI) have been flight tested at this time. The below images also include the variable altitude 3D pattern superimposed over our initial square-off tests. The video for the second flight in Lake Wheeler is provided. This month we also completed the MAVLink Filter testing: first on the emulator, then rover, then SAM, and finally on the LAM. All tests passed successfully

March 2021

Phase 1 Vehicles: Field Testing

We recently tested our Phase 1 vehicles in the field. For this particular test, we used one rover, one Small AERPAW Multicopter (SAM), one portable node (PN), and two Large AERPAW Multicopters (LAMs). Our portable node includes an Intel NUC computer and a USRP B205mini. The SAM is used to test the software on a lower-stakes vehicle before testing it on the LAM. The tests included manual tests as well as fully autonomous tests where the rover, SAM and LAM were fully under the control of the PN: the safety pilots were present for “just in case”. The results (shown in the videos) show good control both for the rover and for the LAM. One of the figures shows the consistency of the trajectory of two consecutive runs of the LAM.  The rovers, the SAM, and the LAM were commanded to traverse a square with a side of 10m, each command being relative to the previous one (takeoff, 10m north, 10m west, 10m south, 10m east, land). Although this is a simple mission, the test runs through all the relevant instructions we plan to support on the testbed in Phase 1.

February 2021

Drone Classification Experiments by University of Louisville

In Fall 2020, the AERPAW team hosted researchers from the University of Louisville (UoL) to carry out experiments in the Lake Wheeler site related to drone detection and classification. Olusiji Medaiyese from UoL visited NC State University and worked with AERPAW pilots to collect data from various different drones and other RF sources using a high sampling oscilloscope from Keysight that was made available to them for their experiments. After post-processing the data collected from the Lake Wheeler site, Medaiyese submitted the following journal:

  • Olusiji Medaiyese, Martins Ezuma, Adrian P. Lauf, and Ismail Guvenc, “Wavelet Transform Analytics for RF-Based UAV Detection and Identification System Using Machine Learning”, submitted to Elsevier Journal of Pervasive and Mobile Computing, Feb. 2021 [Paper].

The UoL team will make the collected data publicly available in the future and cited it in their accepted journal paper. An example drone dataset that was posted by the AERPAW team in the past is available at [Dataset].

This experiment fits under bring your own device (BYOD) type AERPAW experiments. AERPAW is expected to be “generally available” in the coming months and will be broadly accessible to experimenters with programmable drones and programmable wireless equipment. On a case-by-case basis, the AERPAW team can also work with the interested researchers on BYOD-type experiments.  

Drone and Rover Experiments

We have been testing our AERPAW UAV and rover in the Lake Wheeler site, both of which have the same interface for controlling their trajectory. Below you can find a video from our recent testing, where the rover is controlled in manual control (the pilot has control), automatic control (autopilot has control), as well as autonomous control (the portable node has control). All tests went well and we are getting ready for supporting external experiments with AERPAW UAVs and rovers in the coming months. 

January 2021

AERPAW team worked on the manual PID tuning process of the large AERPAW UAV (the “Fat Bird”). The manual control was tightened for better responsiveness as well as reducing the limits on the maximum roll and pitch as the default was too aggressive for a drone of this size. Subsequently, the team took it out for a spin to get a feel of the result and it was deemed as flying well. The UAV breached the geofence in the end, and initiated (and completed) a return to launch, landing where it took off from.

Ground Control and Portable Node Prototype

The below figures show our current ground base station set up with a table that includes two laptops (one for setting and monitoring the UAV, the other for controlling and monitoring the companion computer).  Internet access is currently available as one of the two laptops has a cellular modem and creates a WiFi hotspot. Sun glare is a perennial problem.

The portable node prototype was manually put together after a packaging design process. RF testing of all the components has been carried out (experimental SDR, monitoring SDR, backhaul cellular modem).

AERPAW Presentations

AERPAW team recently carried out the following presentations: 

  1. M. U. Uddin Chowdhury, C. K. Anjinappa, I. Guvenc, M. L. Sichitiu, O. Ozdemir, U. Bhattacherjee, R. Dutta, V. Marojevic, and B. Floyd, “A Taxonomy and Survey on Experimentation Scenarios for Aerial Advanced Wireless Testbed Platforms”, in Proc. IEEE Aerospace Conf., Mar. 2021. [Website][Slides]
  2. I. Guvenc, “AERPAW: Aerial Experimentation and Research Platform on Advanced Wireless”, IEEE Internet of Things Vertical and Topical Summit, Jan. 14, 2021. [Website][Slides]
  3. I. Guvenc, “AERPAW: Aerial Experimentation and Research Platform on Advanced Wireless”, 2nd Buffalo Day for 5G and Wireless Internet of Things, Nov. 20, 2020. [Website][Youtube][Slides]

A list of AERPAW’s publications and academic outreach can be found on our Wiki.

December 2020

During Dec. 2020, six Terragraph radios from Facebook have been installed at CC8, CC10, CC11, and CC12, forming a daisy chain connectivity. They are being tested and expected to be available for as-is experimentation in the coming months. We have also finished installing a Keysight RF sensor on CC2 and successfully performed initial testing. We are expecting to provide remote connectivity to it during the coming months.

November 2020

AERPAW purpose-built light pole getting installed in the Centennial Campus

October 2020

The first flight of a drone for Lake Wheeler Road Field Laboratories (LWRFL)

July 2020

AERPAW LWRFL Construction: Over the last few months, in a previously empty field, a hut and a tower went up, to house respectively the site distribution facility and the first radio node in the Lake Wheeler Road Field Laboratories for AERPAW. A few pictures from various points in the process are included below: the last one shows a radar drone detection experiment in progress.

March 2020