Reliable mission critical voice communications
By Tim Thompson
By Tim Thompson
Just about everyone knows someone who works in public safety. For me, I have two close family members who are first responders. My brother-in-law is a firefighter in Fort Worth, Texas. His son, my nephew, is a paramedic. They both rely on mission critical voice communications to save lives every day, which is something I think about in my work at NIST’s (National Institute of Standards and Technology) Public Safety Communications Research division (PSCR) in Boulder, Colo.
First responders have relied on radios as their primary means of voice communications for several decades. Radio systems are extremely reliable and their tower sites are hardened with bulletproof shelters. Backup power systems enable radios to continue working through unforeseen disasters, either natural or man-made.
In addition to the radio, my nephew carries a smartphone to communicate with ambulance dispatchers. Modern smartphones have many new data-rich features not currently available on radio devices. For example, they often have location-based services such as GPS navigation with up-to-date traffic, online weather reports and radar. But some areas my nephew serves in the Fort Worth metroplex do not have reliable cellular coverage. In those situations, he uses the radio.
First responders see the advantages of migrating from radio to smartphones. However, they like the simple yet powerful “push to talk” function in typical radios: Pressing a single button enables you to broadcast to whoever is using the channel, and releasing the button puts you in receiving mode. Many are concerned about the reliability and capabilities of smartphones’ push-to-talk communications. Plus, there is no guarantee of ubiquitous coverage.
Typical commercial cell tower sites aren’t hardened to the level of radio sites and may not have bulletproof shelters or backup power — making smartphone tower sites more vulnerable to extreme events. Furthermore, cellular devices are not built to handle environmental conditions like radio handsets are; a term we often use at PSCR is “ruggedized” equipment. Additionally, smartphone push-to-talk communications are being further developed but do not currently meet the requirements of a “mission critical” system. Before first responders embrace the new technology, they need to know that smartphones will perform to the same level of confidence they have grown to expect from radio communications devices.
Measuring system performance
First responders need reliable communications to respond quickly to an emergency. So, several of us on PSCR’s Mission Critical Voice team are developing methods to equitably measure the performance of voice communications systems such as radio and push-to-talk over cellular networks.
Our measurement methods team comprises five electronics engineers and three mathematicians. The electronics engineers — myself among them — provide voice communications and systems knowledge, whereas the mathematicians focus on more of the data-processing analysis and measurement uncertainty calculations. Although we have different backgrounds and skill sets, we rely on one another’s expertise to accomplish great things for public safety. We also collaborate with public safety practitioners in both Boston, Mass., and Boulder.
Currently, our group is developing a series of measurement methods to quantify the performance of voice communication systems with the end-user experience — in this case, the first responder — in mind. These quality-of-experience measurements differ from traditional quality-of-service measurements because they focus on the external events that describe the user interaction with the system.
For example, end-to-end access time is a measurement based on the receiving user hearing an intelligible voice. In contrast, quality-of-service measurements focus on the technical, internal, system-specific measurements and may not be a good indicator of the actual user experience.
Previously, we developed a mission critical voice measurement that quantifies mouth-to-ear latency, which is the time it takes speech input into a voice communication transmit device to be output from the receiving device. Some members of our team developed an audio in/audio out mouth-to-ear latency measurement system. Measurements are made using only the speech going into and out of a communications system, so they provide a fair platform for the comparisons of mouth-to-ear latency across any voice communications technologies, such as radio and cellular. This was the first mission critical voice measurement method we developed, and we have since published a paper, test results data and software.
A closer look at access delay
The second measurement method, which we are currently developing, is the aforementioned end-to-end access time. End-to-end access time describes how long it takes from a sender’s push-to-talk button push to the moment a receiving user hears intelligible speech. This time comprises two sources: mouth-to-ear latency and access delay. Access delay characterizes the time it takes for a system to assign a channel upon a push-to-talk request, as well as the time for the communications devices to turn on all required systems to transmit and receive an intelligible voice.
Excessive access delay could cause a first responder to not be able to communicate effectively during a crisis if the system takes too long to grant a channel. The first responder may not be able to address victims’ needs in a timely manner and may get very frustrated while waiting on the system. Even mere hundreds of milliseconds can make the difference between life and severe injury or death during an emergency response.
From a user experience point of view, access delay can be defined as the minimum length of time a user must wait between pressing a push-to-talk button on a communications device and starting to speak to ensure that the start of the message is not lost. The key to the measurement then becomes determining if the start of the message is lost or not. PSCR Division Chief Dereck Orr uses this example: A first responder pushes the button, speaks into the radio, and says, “Don’t shoot.” Imagine the impact if the first part of that message was lost.
This implies the need for audio intelligibility to be directly tied to any notion of access; therefore, we decided to seek out an audio intelligibility expert. Thankfully, we found Steve Voran, an electrical engineer for the Institute for Telecommunications Sciences. Voran’s decades of research in audio intelligibility has been instrumental in helping us develop this measurement method.
A learning experience
Suffice to say, it’s been a learning experience. We had many challenges in developing the end-to-end access time measurement method, which also uses software to quantify end-to-end access time for radio communication systems. It turns out, much more exhaustive testing is required to determine speech intelligibility across multiple speakers, both male and female.
Despite the challenges, we stuck with the program and have published a paper, test results data and software (bit.ly/36xNYW2). Researchers performing push-to-talk measurements can read the paper, view the test results and download the code to further assist them in their research efforts.
Additionally, we are working with a private company to deploy an upper 700 MHz radio system with a repeater located on Green Mountain, on the west side of NIST’s Boulder campus. A radio repeater is similar to a base station in a wireless network and provides the communications infrastructure necessary to support radio calls over a wide geographic area. The antennas for the radio repeater are located on a 24.4-metre (80-foot) tower at an elevation of 1,768 metres (5,800 feet) above sea level. The tower provides coverage around the city of Boulder and beyond.
The new radio system will be used solely for testing purposes once fully commissioned and it will support Project 25 (P25) Phase 1 and 2, the land mobile radio standard set by the Telecommunications Industry Association (TIA), as well as conventional operations. The capabilities of the new radio system allow us to test the various flavours of P25 digital radio so we can compare their performance in push-to-talk voice scenarios.
First responders need devices they can trust. They may migrate from radio to smartphones but should feel confident about the reliability and capabilities of cellular push-to-talk communications. Here at NIST PSCR, we are committed to helping ensure first responders have communications technologies that allow them to perform their jobs and make the world a safer place.
* Reprinted with permission of the National Institute of Standards and Technology (NIST). Originally published in Taking Measure, the official NIST blog here: bit.ly/2PPdSib.
For further reading, visit the links below:
- Mission Critical Voice QoE Mouth-to-Ear Latency Measurement Methods:
- Mission Critical Voice QoE Access Time Measurement Methods:
Tim Thompson has been an electronics engineer at NIST since 2016.