The usefulness of any communications network depends on successful calls. Assuming that no calls fail for technical reasons, the success rate depends on the amount of traffic on the channel and how well it is managed. Channel occupancy is a measure of the amount of traffic carried and is affected by:
1. the number of subscribers (mobiles on a radio network)
2. the number of calls per subscriber in a given time
3. the average duration of each message.
The occupancy of any system varies hourly, daily and possibly seasonally. To be successful the network must be able to cope with the traffic during the periods of highest occupancy and, when considering occupancy it is usual to examine the conditions over one hour, the ‘busy hour’. The occupancy of a channel can be expressed as a percentage: nct
Channel occupancy = 36 %where
n = number of subscribers (on a mobile radio system, mobiles on watch)
c = call rate (number of calls per subscriber in 1 hour)
t = average message duration in seconds
Occupancy may also be expressed in erlangs where continuous traffic, i.e. 100% occupancy over a specified period, is 1 erlang. Less than 100% occupancy is expressed as a decimal fraction, e.g. 30% occupancy is 0.3 erlang. Demand may also be expressed in traffic units, A:
A Ct erlangs =Twhere C is the number of calls in time T and t is the mean call duration.
The amount of congestion occurring on a system can be defined by the grade of service it offers. Grade of service is expressed as a
200 percentage in terms of the probability of meeting either congestion or waiting more than a specified delay before a call can be established. Erlang’s formula is:
AN/N!Grade of service = 1+ A+ A2/2!+ A3/3!+···+A/N! where A is the number of traffic units (erlangs) and N is the number of outlets (trunked channels).The probability that delayed calls will have to wait more than a specified time (W>t)is:
P(W > t)
=
exp
−
(C − A)t H
where
A = traffic units (erlangs)
C = number of trunked channels
H = mean call duration in busy hour
1. the number of users is infinite
2. the intervals between calls are random
3. call durations are random
4. call set-up times are negligible
5. delayed calls are queued so that they are dealt with on a first come, first served basis.
The number of potential radio users is continuously growing and any technique that makes more effective use of the finite radio spectrum is welcome. Whenever technology has permitted, higher frequencies and reduced channel spacings have been employed in the attempt to meet the demand. A method of making more effective use of the available spectrum is trunking.