A high-speed emergency lane for the information superhighway
With emergency services and disaster response teams reliant on the latest data, a traffic jam on the information superhighway can be fatal. In an effort to circumvent any digital roadblocks, researchers at Rochester Institute of Technology (RIT) have come up with a kind of internet express lane that would allow emergency responders to access large data files quickly in life-threatening situations.
The internet can often be a bit like an umbrella – when you need it most, you're least likely to have access to it. During hurricanes, earthquakes, and other natural or manmade disasters, networks can quickly become jammed as members of the public call loved ones, text, post tweets, or share images and videos.
This influx of activity can quickly overload what may already be a damaged system. Not being able to get a message to friends and family during an emergency might cause some anxious moments, but for modern emergency services that often rely heavily on the ability to send large files of data quickly and reliably, vital information not getting through could be the difference between life and death.
"Sharing data on the internet during an emergency is like trying to drive a jet down the street at rush hour," says Jennifer Schneider, the Eugene H. Fram Chair in Applied Critical Thinking at RIT and co-principal investigator on the project. "A lot of the critical information is too big and data heavy for the existing internet pipeline."
The problem is in the way the internet handles information using protocols like Border Gateway Protocol (BGP) or Open Shortest Path First (OSPF), which are complicated and can result in packets of data being delayed or lost completely when links or routers fail.
The RIT team's approach is called the Multi Node Label Routing (MNLR) protocol and doesn't rely on routes discovered by BGP or OSPF. MNLR is a stripped-down protocol that runs underneath these usual internet protocols and is what is known as an immediate failover mechanism. That is, a backup mechanism to immediately establish an alternative route if a link or node fails during a transfer. Because it runs below the existing internet protocols, it can move large amounts of data without interfering with normal internet traffic.
It's like having a dedicated ambulance lane on a crowded motorway to allow emergency services to get through quickly even if traffic is backed up. However, since we're talking about data, a better way to look at it is to go back to the 20th century telephone system.
Not so long ago, making a landline telephone call was a time-consuming affair with conversations having to be routed through switchboards or complex electromechanical exchanges. But in special circumstances, a direct line could be laid on, which was essentially a hardwired link between one telephone and another. This was useless for general calls, but for ensuring point A could get through to point B, it couldn't be beat.
MNLR works in a similar fashion. It uses labels assigned to the various nodes along the path between the two points, and if the ideal path is interrupted, the protocol backs up and looks for a new path along the node's tier. If that one is blocked, it takes a step back and tries again until it finds a clear path. It's a method that would soon fail if it was used generally between two unspecified destinations, but as a fixed emergency pathfinder, it has advantages.
"The new protocol is actually of very low complexity compared to the current routing protocols, including BGP and OSPF," says Nirmala Shenoy, a professor in RIT's Information Sciences and Technologies Department and principal investigator of the project. "This is because the labels and protocols leverage the connectivity relationship that exists among routers, which are already sitting on a nice structure."
During a demonstration this month, MNLR was raced over the US Global Environment for Network Innovation (GENI) against BGP across 27 nodes in a simulation of a network consisting of an incident control center, a 911 call center, and the office of emergency management. Where the BGP took 150 seconds to recover from a broken data link, MNLR took under 30 seconds, indicating that MNLR is more reliable than current protocols as well as faster.
"While BGP has a recommended default keep alive message interval of 60 seconds, MNLR is not so constrained," says Shenoy. "In fact, MNLR can detect failure with one missing keep alive message as the failure or topology change information will be flooded internet wide, which can be expected in certain cases with BGP."
The team is now working to enhance the MNLR protocol with an eye on field testing and eventual deployment in emergency scenarios.