Embedded expert system for cognitive congestion and flow control in WLANs

This chapter introduces an embedded expert system for congestion and flow control of delay sensitive real-time traffic in Wireless Local Area Networks (WLANs). The expert system is based on the fuzzy set theory. It adjusts transceivers' traffic flow(s) for prevailing network conditions to achieve maximum throughput in required application dependent delay limits. In wireless networks delay and throughput are very much dependent on the packet size, packet transmission interval, and the node connection density. Therefore, the expert system on the destination node monitors congestion by measuring an average one-way delay and a change of one-way delay of the incoming traffic. Thereafter it adjusts packet size and transmission interval of the source node by transmitting a control command to the source. A linguistic decision making model of the expert system is described by linguistic relations. The linguistic relations form a rule base that is converted into numerical equations for transceiver's computational efficiency. The developed congestion and flow control method does packet size definition by at most 56 computations. In the system level, the feedback control increases only lightly communicational load by transmitting application level acknowledgements after every 200 received packets. The model was validated by simulating User Datagram Protocol (UDP) traffic in OMNeT++ network simulator. The achieved results demonstrate that the developed expert system is able to regulate packet sizes and transmission intervals to the prevailing application dependent optimum level very fast, accurately and with minimal overshoot and to increase overall throughput of the network. Even if this work is mainly motivated by the congestion and flow control of WLAN systems and the simulations and results were performed for the IEEE 802.11b system, the approach and the techniques are not limited to these systems, but are easily applicable for other Packet Switched Access Networks (PSANs), too