LALRED Algorithm for Congestion Avoidance in Wired Networks

Some applications such as audio and video conferencing require a network to provide QoS guarantee. End- 2-End delay is one of the prominent factors in QoS. Packets after crossing the routers queue arrive to destination node. Thus with guaranteeing the queuing delay in routers the network will be able to guarantee End-2-End delay. Furthermore developers can contract service level agreement (SLA) intelligently. In order to guarantee queuing delay, congestion control algorithms can be used in routers. Furthermore providers can contract service level agreement (SLA) intelligently. Congestion control algorithms are a solution to guarantee queuing delay. A learning automata (LA) is an automaton that interacts with a random environment, having as its goal the task of learning the optimal action based on its acquired experience. Here , we present a LALRED algorithm for congestion avoidance in wired networks. The main aim of this algorithm is to optimize the value of the average size of the queue used for congestion avoidance and to consequently reduce the total loss of packets at the queue and also reduces the Queue delay. We achieve this by applying the LA algorithm at the gateways and by discretizing the probabilities of the corresponding actions of the congestion-avoidance algorithm. In Every Iteration the LALRED, chooses the action which is having the highest estimate vector. This algorithm reduces the number of packet losses at the gateway and also reduces the queue delay.