Smart Grid Test Bed using OPNET and Power Line Communication

Abstract

Recently, the government of the Republic of Korea (ROK) is preparing a mid-/long-term promotional plan having a 'new energy business' in the negawatt market as its core idea. Both the Smart Grids and Micro grids which can manage a small-scale power system of various distributed energy resources that directly produce electric power using new renewable energies for the power distribution, in addition to the Energy Storage System (ESS), and Energy Management System (EMS), are the linchpins of the plan. With such a basis, the ROK government expects to generate more profits and jobs from the ESS- and EMS-based Smart Grids and electric vehicle industry, where some new business-convergent models can be found. Smart Grid is a technology that allows an efficient power usage by collecting the information concerning the condition of power line and power usages status by grafting ICT technology onto power grids. The Smart Grids are largely divided into 5 major areas: the Smart Power Grid, an intelligent power grid established through integration of IT technology; the Smart Place, a living environment where two-way communications between a power supplier and a consumer are possible; the Smart Electricity Service that provides a power transaction service and maintains the Total Operation Center (TOC); the Smart Transportation which is enabled with power charging technology and its infrastructure; and the Smart Renewable that can be achieved through quality improvement of new/renewable energies and their stable links in the power grid. In this paper, the test bed experiments have been conducted focusing on the Smart Power Grid and Smart Place. One of the major problems that the network researchers and operators face today is that currently they can hardly find scalable test beds for the Smart grids. Since it is often impossible for them to conduct experiments directly on a network or build analytical models for the complex systems, they mostly rely on the simulations or the reduced-scale test beds. However, in most cases, designing a reduced-scale experiment that can replace a large-scale experiment with a fair degree of precision is highly unlikely as the results obtain from the alternative experiment could be falsifying and also it is possible that the errors will not be found until the protocols or the applications are actually used in internet operation. Thus, in this paper, a scalable test bed has been constructed by simulating the Power Line Communication (PLC)-based environment largely debated overseas and the Zigbee- and WiMax-based Smart Grid environments that have been presented by the Korean-model Smart Grid with the same scenario in the physical layer and datalink layer. Also, as for the transport layer, the Reliable User Datagram Protocol (RUDP) has been proposed instead of existing TCP and UDP. RUDP refers to reliable UDP. In general, TCP is reliable but slower in speed, and UDP is faster but lacks reliability whereas RUDP has both benefits. One of the TCP's characteristics is that correlation between server and client will be established. That is, a certain connection management is needed whether it is the server or the client. However, RUDP emerged for the reason that it allows efficient communications between clients. First of all, the TCP-based communications between clients in a general Client/Server structure achieve reliability by transmitting data through system server but due to TCP's lower speed compared to UDP and its heavy load to the server, RUDP, which has both the advantage of UDP (speed) and of TCP (reliability), has appeared. One other strong point is that for TCP, one side has to standby being a server but UDP allows immediate communication without doing so. For these reasons, RUDP has been proposed as a suitable protocol for the Smart Grid environment in this paper. In the transport layer, the simulation results showed RUDP's faster speed compared to UDP, and better reliability against TCP. In addition to above, an AMI application for the PLCs has been proposed for the application layer by implementing two scenarios operating in a 220V environment as a foundational technology, followed by performance evaluation of proposed AMI application in which an Intelligent Agent has been embedded. The results showed that the AMI application with Intelligent Agent supplemented the weakness of PLCs. Also for the AMI application, the vulnerable security problem of PLC has been improved by enabling DDoS attack detection function.