Grade-of-Service in an Operational Public Safety Radio System
Once the system is built and operational, the actual working grade-of-service can be determined based on system measurements. The parameters you will want are:
- number of calls (in a period of time)
- number of calls queued
- duration of queueing delay
Note that grade-of-service is not an instantaneous measure. It is something that is measured over a period of time (like a month, or a year). So it is not something you are going to measure quickly during an acceptance process by plugging in a special “GoS” test set. To have a meaningful GoS figure you will need to calculate the value based on actual system traffic over an extended period of time.
With this in mind, plan ahead when procuring a system by specifying the system statistics that you require. These statistics will also be vital during the system acceptance phase.
Once the actual GoS has been determined, what can be done if is worse than you expected/required? There are two things that determine the the Grade of Service:
- the users and
- the number of channels.
Either the users are using the system more than originally planned for (more used or longer average transmissions), or there are not enough channels available in the system.
Adding more channels to the system will obviously cost money. If the system has not yet been accepted then who should cover that cost will depend on who calculated the required number of channels and who provided the numbers upon which the design calculations were made (e.g. number of users, average length of each transmission).
When managing an operational system, the Busy Hour statistics should be gathered on at least an annual basis (and preferably monthly or quarterly). These measured busy hour traffic figures should be plotted on a graph to display potential trends. Over time it can be expected that more users, more radios, will be added to the system. More users mean more traffic and more traffic on the same number of channels means there will be a greater probability of queueing (i.e. all channels being busy when a user presses the PTT button). From a graph of historical traffic levels of the system you can extrapolate to predict the traffic for the next year or two. When you see that predicted line crossing the system capacity level (for your targeted minimum G-o-S), it will be time to take action!
The action to be taken will likely involve making plans to add an additional channel or two to the system. Keep in mind your organizational processes for accomplishing this. Will budgets need to be requested, will a formal project team need to be established? How long will all of this take? Considering these processes will guide you in determining how far into the future you need to be looking.