Grade-of-Service During the Radio System Planning Stage
In Part 1 of this series, Grade of Service (GoS) was explained in a simple fashion. There are two phases in a radio system’s life when Grade-of-Service needs to be considered – and approached in different ways. The first time Grade-of-Service comes up is in the initial planning and procurement process. The second time GoS comes is in the system is operational.
In the procurement process (an RFP for example) the Grade-of-Service will tell potential vendors a very important part of your requirements. At this point the Grade-of-Service is a theoretical concept. Mathematics and industry best practices will be applied to design a system that will meet your needs. The target Grade-of-Service percentage will allow (through traffic engineering principles) the number of channels that the system requires, to be determined.
The mathematics of traffic engineering are well proven, but for good results you need good starting numbers. These starting numbers will be your historical measured radio traffic (if available), that can be extrapolated to estimate future radio traffic. The two key traffic parameters are:
Calls per hour (where a call refers to each individual transmission, each press of the push-to-talk (PTT) button)
the average call duration or how many seconds you talk on each transmission (more accurately, it is how long until the PTT is released)
Public Safety radio traffic is not going to be constant, a consistent number of calls each and every hour. The call rate has cycles – daily cycles, weekly cycles and annual cycles. In each of these periods there will be busier and quieter times. So which hour’s traffic do you use when designing a system, figuring out how many radio channels are required? Radio systems are normally designed based on the traffic during the Average Busy Hour. The number of calls per hour will refer specifically to the number of calls in the busiest hour over some period of time. That period of time would typically be a month, but in theory it could be a year or just a week. It is important that what ever period is used, this “busy hour” definition is documented and understood by both the agency and vendor.
By definition, most of the time the system will handle less traffic than this average busy hour, so the GoS will usually be better (i.e. a lower probability of a user having to wait for an available channel). Because the design calculations are based on statistical probabilities, even a system which has been properly designed for its busy hour, can not guarantee that queueing on the system will not exceed the design target. There will also be random events that could happen at anytime and perhaps even turn your agency traditional “quietest time” into its busiest period.
Given the target Grade-of-Service and a prediction of the radio traffic, the number of radio channels required for your radio system can be calculated. Keep in mind that the better the Grade-of-Service (the lower the percent-queueing number), the greater the cost of the system. A perfect grade of service – no possibility of blocking at all, is simply not realistic from either a financial or licensing standpoint. The only way to guarantee “no blocking” is if you had a separate radio channel for each user.