Network selection with imprecise information in heterogeneous all-IP wireless systems

Selection of an optimal service delivery network is an important problem to solve in an all IP heterogeneous wireless access network environment. Several network parameters impact the process of network selection in such an environment and ideally their precise values should be known by the decision maker. In reality, however, the exact values for many of the parameters, e.g., those related to quality of service, will not be known. Hence there is a need to develop a network selection mechanism for scenarios when some of the parameter values are less reliable or unavailable. This paper describes a novel and comprehensive network selection approach that combines parameter estimation techniques with fuzzy theory and multi attribute decision making algorithm to perform network selection. In addition the paper proposes a new concept of Confidence Level in the network rankings that leverages additional available information in the final decision process. The proposed techniques provide improved network selection in heterogeneous all IP wireless access environment.