Applying High Availability Design and Parallel Redundancy Protocol (PRP) in Safety Critical Wide Area Networks

Abstract

Networks which protect the safety of human lives place special emphasis on network availability and survivability. The nation’s Air Traffic Control (ATC) and First Responder public safety networks used by police departments, fire and rescue, and emergency medical teams are examples of networks that require high availability and survivability. The term mission critical network is often used to describe the characteristics of networks which protect the safety of human lives. There is not a universally accepted standard definition of the term, but much literature on the subject typically identifies three salient characteristics: • Highly Secure • Highly Available • Highly Survivable Highly secure is an important characteristic and needed to design a safety critical network, but the focus of this paper is availability and survivability. It should be noted that mission critical safety networks are private networks and should not be confused with the public Internet simply because they use IP. A private network in itself does not constitute a mission critical network, but it is a significant characteristic of a mission critical network due to the security and performance benefits it supports. The security benefit is risk mitigation from external threats because only authorized internal users can access the network. The performance benefit is similar in that only authorized users have access to the network and their network usage does not have to compete for bandwidth with other external users. Availability and Survivability are related, but they are not the same thing. Availability is simply a measure of the time the network is operating compared to the total time it should be operating. Availability is defined as Uptime divided by Uptime plus Downtime. This same reference defines Survivability as the capability of a system (or network in this case) to perform its mission recognizing that failures are going to occur. As will be explained later in this paper, survivability considers catastrophic events that cannot be easily predicted in an inherent availability model. Specifically, this paper focuses on the availability and survivability of the Wide Area Network (WAN) terrestrial core backbone component of safety critical networks. Much literature on public safety networks for First Responders is devoted to the wireless radio networks including Land Mobile Radio (LMR), P.25 packet radio, cellular telephony and evolution towards broadband 4G Long Term Evolution (LTE) wireless networks. Air Traffic Control networks rely on other wireless forms of communication including narrow-band Air-to- Ground (aircraft to ground based controller) voice and data links in the Very High Frequency (VHF) spectrum. All of these wireless forms of communication rely on a terrestrial core backbone for backhauling and distributing information to the right place. The terrestrial core backbone is a foundational building block for other safety critical network components. This paper also describes some of the differences between legacy Time Division Multiplexing (TDM) technology and modern Internet Protocol (IP) packet switched technology. Historically, networks such as the nation’s Air Traffic Control (ATC) network have relied on point-to-point TDM technology.