Unmanned aerial systems or unmanned aerial vehicles (UAS or UAV’s, the terms are interchangeable) are a recent addition to the modern battlespace. These devices are used to provide real time C4ISR (Command, Control, Communications, Computers, Intelligence, Surveillance and Reconnaissance) and close air support. These systems can be man portable, launched from boats, ships, other aircraft or they can be land-based. These systems can also be used by law enforcement and other first responders in dynamic scenarios, such as monitoring wildfires which is highly dangerous for manned aircraft. One of the known vulnerabilities of these systems is network security. There are several instances of adversaries hacking UAV’s.
Reg Austin describes an unmanned aerial system UAS as a system comprised of subsystems which include the aircraft (often referred to as a UAV or unmanned air vehicle), its payloads, the control station(s) (and, often, other remote stations), aircraft launch and recovery sub-systems where applicable, support sub-systems, communication sub-systems, transport sub-systems, etc (Austin p. 36). These systems work together synergistically to improve the overall tactical picture and cut through the fog of war.
Missions and Capabilities of Unmanned Aerial Vehicles
Unmanned aerial vehicles are used for many different missions. The missions are normally categorized as dull, dirty, or dangerous. The aircraft can monitor wildfires in real time, directing assets to areas that are about to flash over or directing firefighting teams to take cover because of changing winds. Ash and smoke are highly destructive to aircraft engines. Due to the fact that these aircraft are unmanned, if the aircraft should suffer engine failure there is no crew at risk.
UAV’s can also be used for more mundane purposes. Some examples include fish finding, monitoring crops for disease, proper irrigation and crop dusting, one of the newest uses of these systems is aerial photography. The potential use of UAV’s is always growing. The graph below shows the growth of UAV’s in different markets.
Figure 1 Source: Aviation Week
This growth has made the importance of network security a topic of utmost importance.
Notorious UAV Security Events
There have been several high profile events involving UAV network security. In 2009 a US official revealed that Iraqi insurgents were intercepting the feeds from UAV’s. In 2011, the Iranians were able to gain control of a US Navy RQ-170. A year later in 2012 a college team is able to do the same.
According to a 2009 article by Mike Mount and carried by CNN, Iranian backed Shiite militias were able to monitor real-time live feeds from Predator UAV’s. An unnamed defense department source was cited as stating that the intrusion software was an open market program from Russia called SkyGrabber, which can be downloaded from the internet.
The Department of Defense (DOD) is on record as stating that no operations or missions were compromised. These statements read as blatantly false and self-serving. The DOD has also admitted that a similar situation existed in Afghanistan, but has not revealed any details.
In 2011, the Iranian Revolutionary Guard Corps reportedly hacked into a highly secretive RQ-170 Sentinel (Axe, 2012). This aircraft has an estimated unit cost of $6 million USD. The exact method of the hack is unknown, however it is suspected that the Iranians were able to use intelligence gathered from multiple sources. This analysis is from the Iranians themselves, as they have released maintenance records of the aircraft. It is from this multi-source intelligence that the Iranians learned that the narrow-band communications relay used by the RQ-170 is not encrypted.
Figure 2 Source: US Air Force
In May of 2012, the Department of Homeland Security “dared” a team from the University of Texas at Austin (UTA) to hack a UAV, because it was “impossible”. In June, not even a month after the challenge was given the team, using $1,000 worth of equipment, was able to spoof the GPS system and take control of the aircraft from over a kilometer away (UTexas.EDU). This was a blatant demonstration of the vulnerabilities of command and control systems. GPS signals operate at a known frequency and by overpowering the signal from the real command center the UTA team gained control of the aircraft.
Network Security for UAV’s
According to a presentation by Vic Patel of the FAA titled “UAS Networked Access Security”, a joint security sub-group is writing security procedures using Federal Information Processing Standards (FIPS 140) and NIST 800 series Special Publications. This is designed to develop standardized security for all UA systems so that multiple devices operating within close proximity are not broadcasting mutually disruptive interference.
One of the issues that the scenarios previously described have in common is that all of the command and control links were not encrypted. In the case of UTA, they were able to use a directional antenna and broadcast a stronger signal, which over-rode the normal inputs. This was possible because the aircraft was controlled through GPS and the team was able to use signal analysis to then broadcast a stronger signal and gain control.
While this may sound trivial the DoD has acknowledged that securing the network feeds is problematic because the data needs to go to multiple locations and encryption puts an unnecessary delay on those broadcasts. This delay was seen as unacceptable. This author questions the analysis used to reach that conclusion. A delay of a few seconds already exists due to limitations imposed by the speed of light, the distances involved, data bandwidth and the processing power of the communications satellites being used. An additional delay caused by encrypting the network would not place an impossible burden on the network.
The Navy has an interesting solution to this issue. Instead of using the uplink for data and avionics, the mission is uploaded into the aircraft and operator’s do not “fly” the aircraft, instead waypoints are uploaded individually or in batches. This is through an en encrypted communications relay and therefore does not have the vulnerabilities of real-time flight controls. This also allows more bandwidth to be used for imagery data. Real-time flight controls are also not real-time; they are near real-time, because of the time lag described above. By having the aircraft fly by waypoints the aircraft has direct control of its flight surfaces in real time. This allows the aircraft to fly more efficiently and without the potential for pilot error.
Security of the network should be placed above an unnoticeable delay in the relay of data. If a network is unsecure and the enemy has access to it what exactly is gained? Tactical intelligence that is not secret is not usable as the enemy can easily change their operating methods and cause allied forces to believe false data and either be at a different location or an ambush.
UAS Security in a Civilian Environment
In the previous examples, we have examined the issues in the intelligence and military realms. The civilian realm is just as vulnerable, if not more so than government sectors. Businesses that invest a large portion of operating capital into a UAS and then have that aircraft stolen or damaged may not recover from the financial damage caused and go out of business. This could ripple through the economy causing other negative effects and disruptions to other businesses in the sector.
One potential example of this is a fishing boat that purchases a ScanEagle system. This system was designed and optimized to be launched and recovered from small boats in order to track bait balls and large pellagics. By using a non-secure network, a competitor could take control of the aircraft and use the imaging for their own purposes or if they are unable to do that, they could simply turn off the engine, causing the competing business to lose the aircraft and the investment represented by the system. The financial impact to the business cannot be understated.
ScanEagle launch from Mk V SOC
Law enforcement and emergency services also operate in the civilian realm. Kai Daniel and Christian Wietfeld wrote in a paper for the IEEE that “Unmanned Aerial Vehicles (UAV) enable the in-depth reconnaissance and surveillance of major incidents. Uncontrolled emissions of liquid or gaseous contaminants in cases of volcanic eruptions, large fires, industrial incidents or terrorist attacks can be analyzed by utilizing UAV.”
The data gathered by unsecured networks in a dynamic situation can be used by criminals or other actors to impugn the character of evidence gathered by these systems thereby affecting the outcome of judiciary proceedings. Arsonists could alter visual records of a fire recording thereby creating reasonable doubt in the eyes of a jury.
Conclusion and Recommendations
Due to the myriad risks and vulnerabilities posed by unsecured UAS networks, it is vital that ICAO, FAA and private operators implement technologies and regulations that ensure the security of the systems and networks. While network security seems like a trivial issue, we can see that it has directly lead to the loss of a $6 million USD aircraft. That aircraft was an asset that was highly important to national security. While a ScanEagle operated by a private fishing charter does not rise to that level of importance on a national scale. It is important to the owner of the system and those who are financially supported by the owner. Different systems will, by their very nature require different technologies. A smaller system cannot be expected to use the technologies required by larger systems. This must be reflected in any rules or regulations adopted.
Other issues that need to be addressed are what encryption methodologies are going to be used. Whether this is single key, double key or split key encryption. Each of these methods has its own strengths and weaknesses that must be analyzed to ensure that system reliability and operability is not impaired. If the network security induces unnecessary lag that can endanger innocent lives and property, methodologies and procedures need to be put in place that mitigate or eliminate those risks. Security should be one of the key considerations before purchasing or operating a UAS.
Procedures also need to be consistently applied in the applicable sector. Regulations should be different for different economic sectors. As an example, procedures for using UAV’s for security surveillance of a nuclear power plant should be much different than those for fishing or firefighting.
In conclusion, regulations, network security and standard operating procedures need to work synergistically to enhance the use of these systems. This synergy will build public trust and enhance operational capabilities. This is beneficial for the businesses affected and the economy in general.
References
Austin, R. (2010). Unmanned aircraft systems UAVs design, development and deployment. Chichester, West Sussex: Wiley.
Axe, D. (2012). Nah, Iran Probably Didn’t Hack CIA’s Stealth Drone. Wired. Retrieved from: http://www.wired.com/2012/04/iran-drone-hack/.
Daniel, K., & Wietfeld, C. (2011). Using Public Network Infrastructures for UAV Remote Sensing in Civilian Security Operations. Presented at the IEEE Conference on Technologies for Homeland Security, Waltham, MA.
George, S. (2015). FAA UAS Cyber Security Initiatives [Presentation]. Retrieved from http://csrc.nist.gov/groups/SMA/ispab/documents/minutes/2015-02/2015-feb_george-ispab.pdf
Mount, M., & Quijano, E. (2009). Iraqi insurgents hacked Predator drone feeds, U.S. official indicates. CNN. Retrieved from http://www.cnn.com/2009/US/12/17/drone.video.hacked/.
Patel, V. (2014). UAS Networked Access Security [Presentation]. Retrieved from http://www.icao.int/safety/acp/ACPWGF/ACP-WG-M-21/IP03%20WGM%2021%20UAS%20Networked%20Access%20Security.pdf
Shachtman, N. (2011). Exclusive: Computer Virus Hits U.S. Drone Fleet. Wired. Retrieved from http://www.wired.com/2011/10/virus-hits-drone-fleet/.
University of Texas. (2012). Cockrell School Researchers Demonstrate First Successful “Spoofing” of UAVs. Retrieved from University of Texas, Austin Cockrell School of Engineering website: http://www.engr.utexas.edu/features/humphreysspoofing
Figures
1. Warwick, G. (2014, January 1). AUVSI – Precision Agriculture Will lead Civil UAS. Aviation Week.
2. http://www.af.mil/AboutUs/FactSheets/Display/tabid/224/Article/104547/rq-170-sentinel.aspx
3. http://www.news.navy.mil/view_single.asp?id=551