Abstract
The ever-increasing number of new customers joining the power grid is impacting the delay performance in smart grids, as packets generated by numerous Electrical Devices (EDs) are required to be transported through several interconnected networks. Thus, the high volume of traffic generated in Neighbourhood Area Networks (NANs) is causing delay due to link congestion. In order to minimize this, Wide Area Networks (WANs) combined with an efficient routing strategy is a possible solution. In this paper, a topological design model that adapts to the faster routes is proposed for achieving minimum end-to-end delay. Routers are modelled as G/G/b(l) queueing systems with First Input First Out (FIFO) discipline. A Dynamic Fastest Routing Strategy (DFRS) for such a network with n number of nodes and m number of links that enables the packets to choose the best routes from source to destination is presented. It takes into account the level of clogging over each link, based on: i) the average threshold time of transmitting a packet over a link, ii) the Quality of Service (QoS) expected from the WAN, and iii) the performance of forwarding packets over sprightly routes. Expressions for mean packet delay (including waiting, transmission, and propagation time) are derived as a function of maximum number of packets, traffic intensity, death rate, capacity, number of channels, and the latency of the link. The simulation results are compared with theoretical expressions and wire-line networks that employ classical G/G/1 queueing system for an Open Shortest Path First (OSPF) routing scheme. It is noted that the proposed model is simple to implement and minimizes the average packet delay in WANs, and meets smart-grid standards.