By David Layher WAØPCC
CVARC Communications – Activities Director
AREDN© (Amateur Radio Emergency Data Network) is a collaboration of radio amateurs that repurpose commercially available WIFI equipment to create high-speed ad hoc data networks.
The conversion of a commercial WIFI access point into an AREDN node (each AREDN radio on a network is called a node) is done by flashing/reprograming the hardware with the AREDN firmware. This is the real heart of AREDN. For a deeper dive into all things AREDN, visit arednmesh.org.
For several months, a small group of local hams have been working together to develop the skills and hardware to use AREDN for a race-type deployment, with the goal of having a deployable mesh network for use in emergency situations. A further goal is to have permanent nodes installed at critical service locations such as hospitals, police and fire departments, and Emergency Management (EMA).
We finally had enough experience and equipment to attempt our first nonemergency deployment.
We chose the Especially for You (EFY) race/run/walk through downtown Cedar Rapids on Sunday, October 6. The goal was to deploy several modes with streaming video from locations along the course. This was an aggressive attempt due to the topography of the race, downtown = tall buildings. Since AREDN is on the 2.4 and 5.8 GHz bands, it is line of sight. Buildings, trees,
power lines, and people can/will degrade the signal.
To make the EFY event work, we would need at least one high node that could be reached by the others. Thankfully, Mercy Medical Center (MMC) is the main sponsor of the event and the location of the start/finish. They graciously allowed us to put a temporary node on the roof of their building. At this point, let me offer a very special thank you to Chuck in MMC maintenance for all his assistance in placing, adjusting, and taking down the node.
Before we go on, a bit about the AREDN node hardware. The AREDN radios are powered by 12 to 24 vdc. Recent radios are MIMO, Multi In Multi Out, simultaneously transmitting/receiving with both vertical and horizontal polarization. Each node is a radio, and antennas and many devices have both radio and antennas in the same package. The radios are relatively low power (~0.5 watts). Nodes with external gain antennas have more. There are three basic antenna types: dish, sector and omni. Dish antennas keep the beam fairly narrow and are good for longer distances. Sector antennas typically cover a 45-to-120-degree area. Omni antennas have 360-degree coverage. Due to the topography of downtown and the use of all temporary nodes, the EFY deployment was going to be a point-to-point network configuration. This eliminated one of the most valuable attributes of AREDN, a self-discovery, self-healing, mesh network. We were now in a
single point of failure configuration which Murphy would use against us. Our main node for the network was a radio with an omni antenna on the roof of MMC. This node was powered by a solar/battery supply. Due to the extreme down angle from this radio to the start/finish area, a second sector radio was angled down toward the start/finish. Fortunately, this node performed flawlessly. We chose to have a node on the PCI parking garage at 2nd Avenue and 10th Street. This node would have a camera and would use a sector node with line-of-sight to the MMC omni. This node would capture the participants making the turn from 10th Street on to 2nd Avenue.
Our next node was at the rest station on 1st Avenue by the Veterans Memorial Building. This node would also have a camera to show runners going out and back. This location did not have line of sight to MMC. This sector antenna node would need a relay.
There was a node with video at the Halfway Point of the 5K. This location presented the largest challenge, as it did not have line of sight to MMC and there were trees and buildings that significantly narrowed the line of sight to the McGrath relay. A dish node was used for the location. A relay was needed and it would be a challenge. McGrath Amphitheater Relay – This location was set up to receive from the 1st Avenue node and the Halfway Point node, via sector radios aimed accordingly, and then send the data on to the MMC location via a dish. This location was our most complex radio setup and was the potential failure point for two other locations. Murphy was loving it.
A quick note about co-located nodes. These nodes are connected through a switch and the AREDN software recognizes this and uses the hard wire connection to pass data between the
nodes, bypassing any rf between the nodes. At the Start/Finish area were two nodes: one near the Medical Tent across 10th Street from MMC and the TV node next to the Hall-Perrine Cancer Center (HPCC). The Med Tent location had two radios, one sector radio pointing to the MMC roof and a second sector radio pointing to the HPCC location. It also had a camera. The HPCC location was the final stop in the network. It had a sector radio to receive from the Med Tent, a camera, and a 54-inch screen to display the video feeds. A laptop computer attached to the network running Blue Iris software allowed all video feeds to be displayed on screen at the same time. The team did a dress rehearsal on the Saturday morning before the event. Several hams that were new to AREDN came to assist with the exercise. We did a quick shakedown of the hardware for each node location, then the group divided up into smaller groups to set up the nodes at their race day locations. The rehearsal went well and most of the hams that helped with the rehearsal volunteered to man the node on race day.
Race Day: Set up was done pre-dawn (it was dark outside) and that added complications to the set up that did not happen during rehearsal. Also, race day was windy. After the network was up, we discovered that the McGrath Relay nodes needed to be re-aimed and because of the wind, one of the radios fell over and quit working. Since this was such a short event, there was not time to recover. The relay issues caused the feeds from the Vets Memorial rest stop and the Halfway point to be very spotty. Did I mention it was windy? A power cable drooped through the line of sight of the 2nd Avenue node to the MMC omni. It was not an issue during rehearsal. With the wind, the power line was waving and it sporadically caused the feed from the node to drop. A little extra height at the 2nd Avenue node would have prevented the disruption. Murphy anyone? Not enough height, again. The Med Tent to HPCC connection was rock solid during rehearsal. However, on race day, there were thousands of people on 10th Street between the nodes and that occasionally caused the feed to drop. A few feet of additional height on the nodes would have easily corrected the issue. Again, short event, no time. We were able to have full network connectivity for a few short periods of time and it showed that the system was capable of the
needed bandwidth.
Post-Event Reflection: In many ways, this deployment was a success. All the hardware worked as it should. Any failures were due to human error. A team of hams got together to do something we have not done before. And yes, we learned from our mistakes. We plan to do a smaller and hopefully better AREDN deployment at the Reindeer Run in December. Anyone that would like to participate is welcome. The AREDN group meets on the first Saturday morning of each month at the First Lutheran Church at 3rd Avenue and 10th Street SE. For more information you can contact me, David Layher, at wa0pcc@gmail.com.
Thank you to all these hams that helped with the EFY deployment:
WØGSK Greg Kuhlman, KDØCGN Kraig Wilford, KØCTA Steve Hammer, KF9MT Rickie Sexton, NØSFH Neil Johnson, KBØQCF Steve Nyberg, NØXZK Dan Hoffman, WAØFBX Erv Hoffman, and
KD6UCE Greg Yehling.