Triangular and trapezoidal shapes are adopted for both input
and output membership functions and are shown in Fig. 2.
Table I shows the linguistic rules for the fuzzy logic adaptive
weight factor estimator. The design of the fuzzy linguistic rules
takes into consideration of the following conditions:
• Under severe congestion: the arrival rate of the input traffic
increases rapidly, the queue length and traffic load at the output
link are normally high, then the adaptive weight factor should be
low or moderateJow, to reduce the available free bandwidth.
• Under a light load condition: the queue length is low, the input
traffic and the outgoing link load are low, the adaptive weight
factor should be high, so that the proposed CAe can increase
the amount of available free bandwidth in order to accept more
connections.
• Under moderate lnad condition: the weight factor is medium
so-that the. proposed CAe keeps accepting new connections as
long as the QoS requirements are satisfied.
B. Congestion Control
Fig. 1(b) shows the proposed fuzzy logic congestion controller.
It consists of a fuzzy logic predictor and a fuzzy logic
target utilization factor generator.
Separate output queues for different categories of services, i.e.
the CBR/VBR queue and the ABR queue, are considered. Since
CBR and VBR traffic have distinct tolerances for delay, jitter and
cell loss ratio, they are served at a higher priority than ABR traffic.
The predictor predicts the ABR queue one round-trip delay
in advance. This predicted queue value, together with the total
queue growth rate and current ABR queue length are provided
to the fuzzy logic factor generator which produces a target utilization
factor (Tf ). This factor varies the target ABR capacity
dynamically according to the buffer condition one round-trip delay
in advance.