Sustaining Negotiated QoS in Connection Admission Control for ATM Networks Using Fuzzy Logic Techniques

The objective of connection admission control (CAC) is to keep the network load moderate to achieve a performance objective associated with Quality of Service (QoS). Cell Loss Ratio (CLR), a key QoS parameter in ATM networks, is essential for proper network resources dimensioning, congestion control, bandwidth allocation and routing. In this research, we employed fuzzy logic technique in Statistical Connection Admission Control (S-CAC)--a CAC employing multiplexing of the bandwidth between the peak cell rate and the sustained (average) cell rate. The fuzzy technique consists of an input stage, a processing stage, and an output stage. We defined the rules with "max-min" inference method in which the output membership function is given the truth value generated by the premise. The results was defuzzified to a crisp value using the "centroid" method which favors the rule with the output of the greatest area, and the result thereafter charted to compare it operation. Keywords: ATM, Fuzzy Logic, Connection Admission Control (CAC), B-ISDN, QoS.