Abstract:
Radio Resource Management (RRM) becomes a challenging job for network providers when it comes to provide flexible, higher bandwidth services, and maintaining the best system capacity, which are the distinguishing features of 3G communication systems. Call dropping, non-availability and poor network performance can lead to reduced revenue and customer dissatisfaction. To create harmony between expectations of service providers and customers, the concept of SLA can be a potential solution. Service Level Agreement allows service provider to differentiate itself from its competitors and to offer different levels of service guarantees to its customers based on the amount they pay. The basic idea behind underlying research is to manage these resources efficiently and dynamically in order to maintain QoS requirements according to SLA.
The concept of intelligent agent is used to give autonomy to the base station so that it can efficiently maintain quality of service as specified in SLA as the congestion occurs. The idea is to minimize the probability of occurrence of congestion or to control the factors causing congestion before its occurrence. Using Case Based Reasoning, an Artificial Intelligence technique, agent catalogs experience into "cases" and matches the current problem to the experience. It finds solution to the new problem by analyzing previously solved problems, called cases, or adapting old solutions to meet new demands. Now the agent can continuously monitor service performance and apply a suitable policy to prevent or reduce congestion.
In this work, a case retrieval and adaptation approach is illustrated. Retrieval approach efficiently retrieves a case, from historic records in CBR, which is closest to the current network situation. The adaptation approach adapts a previously stored case to the current network situation. Consequently, a best solution is retrieved by learning from previous network situations plus the corresponding actions taken.
