With the overwhelming success of Ethernet technology in local access networks (LANs), owing to its ease of management and its much lower line rate cost, there has been a continuous effort in migrating Ethernet technology into metro-area networks (MANs) and wide-area networks (WANs). However, current Ethernet technology has to overcome the scalability and reliability deficiencies in order to be recognized as a viable MAN/WAN technology. Recently these concerns are being addressed under the framework of Generalized-MPLS (GMPLS) and its traffic-engineering (TE) intelligence. In this paper we propose an optical Ethernet solution viable for MAN/WAN based on the concept of a reconfigurable Ethernet node, integrating reconfigurable optical cross-connects (OXCs) with traditional Ethernet switches. We subsequently proposed a dynamic TE scheme that utilizes both the simplicity of all-optical switching and the flexibility of optical-electrical-optical (OEO) traffic grooming realized by the proposed optical Ethernet node. We showed through simulations that with our proposed TE scheme, the number of OEO transceivers needed by such an optical Ethernet node is only about 20% of a fully-populated one in the extreme case (very high load) and about only 10% in normal load conditions, without degrading the performance. This actually implies a tremendous cost saving on Ethernet OEO transceivers, and on the other hand indicates that a highly efficient and scalable MAN/WAN Ethernet network system could be well achieved by using Ethernet switches of only modest switching capacity (as legacy Ethernet switches have) along with reconfigurable OXCs. To demonstrate and test our proposed ideas we have built a GMPLS testbed using MEMS-based optical cross-bar switches and legacy Gigabit Ethernet switches. The testbed consists of three core nodes and six edge nodes. Performance experiments have been conducted on the testbed as well.
Reconfigurable optical network, the way of the future, can encounter serious packet loss problem during optical level reconfiguration. This paper presents a novel buffering scheme for GMPLS optical networks that minimizes packet loss during reconfiguration of optical nodes, consisting of both a reconfigurable optical cross-connects (OXCs), or an optical add-drop multiplexers (OADMs), and a packet-capable layer-two switch [1]. In this scheme, buffering takes place whenever an optical path is altered due to either a node reconfiguration or protection/restoration. Packets are buffered in a distributed manner starting from the first O/E/O-capable node before the reconfigured optical node all the way to the ingress node, if needed. In the worst-case scenario, where the path is all-optical or buffers are full along the path, all incoming packets are buffered at the ingress node, or the edge router. Buffering is carried out at layer two so that the interruption of service is kept at a minimum (less than 50ms, if possible). Furthermore, if buffers are overflowed, packets dropping will be carried out in accordance with the QoS levels, which further ensure that QoS of the network is maintained. It is shown by both analysis and simulation methods that this scheme performs quite well on either a general local area network, such as the UMBC WDM optical testbed, or the MCI backbone network. It is also cost-effective in that, even at the worst-case scenario, the required buffer is quite small.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.