The algorithm first requires the establishment of fuzzy
sets and rules. By appropriate membership functions, the
input parameters are changed from the crisp numbers into the
fuzzy sets, which is called the fuzzification. Then the fuzzy
sets are fed into the fuzzy inference engine to be analyzed
based on the pre-defined fuzzy rules. Finally, the results of
fuzzy inference are converted back to a crisp output through
the defuzzification process. The output from the defuzzier
determines whether a handoff is necessary. Fig. 2 shows the
basic process of the fuzzy logic system