Use of a fuzzy logic scheduler to improve quality of service on cellular networks

摘要

An efficient and fair packet scheduler is very important to guaranty the cellular network quality of service (QoS). In general, classic packet schedulers follow static rules based on priorities. Unfortunately, static rules may cause unfair resource distribution. We present and analyze FuzSy, a cellular network packet scheduler based on the quality of service and fuzzy logic. Instead of following static prioritization rules, it prioritizes traffic classes that have a higher risk of not reaching QoS requirements. Its dynamic priority changes any traffic class in case of performance decrease risk. Based on network metadata, FuzSy uses fuzzy logic to decide about prioritization. We conducted simulations in a 4G network. FuzSy achieved QoS requirements and network fairness for all classes. One of FuzSy's significant advantages is the fairness guaranty that solutions contained in other works do not garantee. We show that FuzSy can be used in any kind of cellular network because it is straightforward to have its parameters adjusted. In the future 5G networks, for example, FuzSy may be helpful due to a large number of heterogeneous connected devices and different kinds of applications. The FuzSy scheduler use may achieve the QoS, and the fairness guarantees since its concept is independent of technology, and it is usable in any kind of network.