This paper evaluates the network architectures, the innovation in optical (WDM) system design, and the related transport technologies that have enabled metropolitan networks to scale to hundreds Gb/s of capacity, and to hundreds km of reach, meeting the diverse service needs of enterprise and residential applications, while maintaining cost lower than traditional long-haul optical transport. The analysis have lead to the development of a multi-service WDM system scalable to more than 300 Gb/s, and 500 km, exceeding the previous related demonstrations in reach and network (OADM) complexity. The most important performance achievements are summarized. The paper finally discusses future important technologies that hold promise to further enhance the cost-effectiveness of metropolitan optical systems.

At present, automatic detection and topology management of network device are hotspots in network device management field. The new devices of many manufacturers have such functions. But the realization of these functions on devices is based on private protocols of their own manufacturer, and cannot cooperate with each other in the network.This article presents a GARP based network Device and Topology Registration Protocol. This protocol is based on GARP protocol defined in IEEE802.1D, and the registered attributes are network device identity and topology information. This protocol packet carries network devices’ identity and topology registration information, and spreads them within the local area network. Each Participant sends Join and Leave packets, spreading network device and topology information in the connected network. During the process, each participant can automatically detect the network device information and topology information in the network and monitoring their states’ change in real time.This article defines the frame format of GDTRP packets and the information carried on each bit, including network device information (such as device type, manufacturer, IP address, MAC address and etc.) and the link state information (topology) of network devices. It also expatiates how network devices quickly detect other devices in the same network and establish their topology according to the spreading GNDRP protocol packets.Since GDTRP protocol defined in this article is based on generic GARP protocol defined by IEEE, it provides a good scheme to solve the inter-linkage problem of network devices from different manufacturers.

The paper will present a J2EE-based architecture of integrated telecom network management system, and also will introduce the MVC(Model, View and Control) design pattern in the architecture. Using J2EE and MVC design pattern, we can easily build multiple user interfaces (included Web-based), flexible, manageable, and extensible network management system.

Currently, network administration involves repetitive and time-consuming tasks. Clustering enables network administrators to overcome hurdles when they attempt to manage their LANs. Clustering is proved to be a breakthrough for LANs. Under the cluster management model, all switches can be managed as a cluster based on a single IP address. A single command switch for each cluster would automatically discover and control all switches in a cluster, regardless of where they are located. Switch Clustering is implemented over the TCP/IP protocol. The hardware scheme is present in this paper. Then the cluster communication protocol is analyzed in detail. It involves CTP (Cluster Topology Protocol) and CMP (Cluster Management Protocol). At last we present two typical examples for Switch Clustering.

As large amount of SDH optical systems are deployed in metropolitan area network (MAN) and access network, the scale and complexity of SDH management network is increasing dramatically. It is a very challenging task to design a robust and effective element management system (EMS) for large-scale SDH network. As well known, OSI's network architecture has been adopted in traditional SDH network management system for a long time. But it has been proved that OSI stack has many limits, such as complexity of implementation, high resource consumption and low efficiency. This paper presents a method different from that of OSI to build SDH management network. The presented method is based on lightweight IP stack and layered routing scheme of OSPF. By this method, the number of network elements of the managed SDH network can be increased remarkably without loss of management performance. Analysis of other merits of this method is also given. The implementation of EMS based on this method is introduced.

This paper firstly introduces the basic framework and the character and then makes analysis and design to the WDM network's management, which is using optical control channel to transmit network management information on the basis of comparing several existing transport methods of network management information.

Since optical ring is the main topology in metro ring, it is essential to study on the medium access control protocol. In this paper, medium access control protocols in metro optical ring are discussed. We firstly classify the current protocols, then analyze performance requirements and technology attributes thoroughly. At last, several topics remained to be studied are presented.

Based on the P-cycles method and Hamiltonian cycle protection method, we propose a new resource-based cycle protection method for mesh WDM optical network, which is called RP-cycles method. And a heuristic algorithm of RP-cycles method is given. The method has many advantages such as simple configuration, independent of traffic model and having physical length limitation of cycles. Network simulations show that RP-cycles method has high capacity efficiency and short recovery time. So it is a new kind of practical optical network protection method for link fault.

Ethernet is the next big thing in the telecom industry. In the metro network, access capacity is always an issue - not enough bandwidth to offer broadband services which will generate new revenue for service providers. Ethernet has been identified as one of the technologies that can break this spell in the next few years. Then why have we not seen large scale metro Ethernet deployment from any carriers? This article examines the issues in optical Ethernet technology, and the strengths and weakness of works in progress trying to solve them.

Providing massive media streaming services is a very difficult problem for academic research and also a very constructive proposal for commercial investment. Traditional client/server model with very high cost/profit ratio and big bottlenecks, and novel P2P model with complex management, are all not good. In this paper we propose a semi-autonomous peer-to-peer-like network to provide massive media streaming services and avoid above problems. Our scheme is a tradeoff between server-systems and P2P systems. In this scheme, we make full use of P2P systems advantages, high aggregation network bandwidth and large storage spaces. But to avoid the impacts from peers unbending actions, we divide whole network serving area into many small zones and designate a seeding server for whole network area and a sub-seeding server for each network zone. Simulation results prove the advantages of the new system with good scalability and stable QoS.

An admission control algorithm is a very important component in a video server that can services large number clients simultaneously. Given the real-time requirements of each client and the fixed data transfer bandwidth, a VOD server must employ admission control algorithms to determine whether or not a new request can be admitted to the server without violating the requirements of the clients already being serviced. All prior admission control policies for VOD servers have focused on improving the storage system utilization such that a video server can service more users while satisfying the QoS requirements. For VOD service, the users can only view the video through transmission networks, so the transmission of video stream in the transmission networks is also very important. The admission control scheme presented in this paper is divided into two layers: disk subsystem admission control and network subsystem admission control. Through simulation, our admission control algorithm outperforms others significantly.

As the link and gateway between backbone network and access network, metro-network becomes next infrastructure hotpot for both telecom and datacom carriers. In the transport plane of metro network, there are many solution choices to meet the various needs of carriers and operators. One of these practical solutions is with the help of metro-oriented Multi-services Transport Platform (MSTP). Inheriting from SDH transmission system, MSTP can process packaged data traffics like Ethernet on L1/L2 layers and ATM cells as well as TDM voice traffic. There has been a trend that packet data service will be the dominate in a metro-network near future. More reliable and efficient packet processing function such as RPR and LCAS will be needed in MSTP equipment. With the development of ASON/ASTN technology, more intelligent functions originated from ASON/ASTN will be added to MSTP and ASON/ASTN will be the development direction of MSTP. With the transition process of metro-network bandwidth limit from electrical domain to optical domain, some WDM and electrical/optical hybrid cross-connection functions will be merged into MSTP as optional modules. Metro-oriented MSTP will act as major node equipment in metro-transport network.

The optical transmission of signals has several significant advantages like high transmission capacity, immunity from electromagnetic interference, and thin and low weight cables. Optical networks have a high potential for applications in buildings, but they will be widely used only if component and installation costs are reduced. One promising candidate especially for home networks is polymer optical fiber (POF) due to the low cost of the fiber and its installation, as well as significantly lower costs of all active and passive components compared to networks based on silica fibers. The present paper describes the properties of polymer optical fibers (POF), active and passive components developed in our institute for networks based on POF as well as different transmission experiments. The applications of these transmission systems for multimedia and multiservice networks are the distribution of video signals, the transmission of TV signals with carrier frequencies beyond the 3-dB bandwidth of POF, and a multimedia system for the transmission of video, audio, and data signals using wavelength division multiplexing (WDM).

The new requirements of Metropolitan Area Networks (MANs) have resulted in the emergence of new data transport technology, Resilient Packet Ring (RPR). The issue of fairness is one of the key technologies in RPR. Whereas traditional definitions for fairness are not applicable for RPR, a fairness conception for RPR based upon ingress aggregated flow is presented in this paper. Then fairness algorithms described in the draft of IEEE standard for RPR is analyzed. In order to overcome the defect under condition of unbalanced data stream, an improvement of the fairness algorithm named RPR-RE which adopts the rate estimation is produced to achieve better performance. At last, the simulation is made and the results are given to prove the conclusion that RPR-RE could eliminate the throughput oscillation approximately compared with the algorithm defined in IEEE standard draft under the scenarios with unbalanced flow. Furthermore, RPR-RE algorithm could implement the fairness between the nodes when multiple congested nodes exist in the ring.

Resilient Packet Ring (RPR) is an emerging network architecture and technology designed to meet the requirements of packet-based metropolitan area network (MAN). IEEE 802.17 RPR Draft Standard[1] defines RPR dual-ring topology discovery protocol. In this paper, we propose a new RPR multi-ring interconnected network. RPR multi-ring interconnected network has better resource availability and better bandwidth utilization than RPR dual-ring topology. So how to extend RPR dual-ring topology to RPR multi-ring interconnected network really deserves our attention. RPR multi-ring network is formed by decreasing the scale of standard RPR and interconnecting several small scale standard RPR rings together through cross-stations. The new topology discovery algorithm uses the method of multi-layer subnet topology discovery in clockwise implemented by the multi-ring automatic topology discovery module in cross-station’s MAC control layer. Furthermore, we bring forward the unit structure of tree-like bi-direction list and of topology information list. The paper mainly contains the two parts: RPR standard topology discovery algorithm and RPR multi-ring interconnected topology discovery algorithm.

Contention resolution has been a major concern in the design of burst photonic networks. Previous researches have proven that wavelength conversion is an effective technique for contention resolution. However, it is not realistic under the current technology to assume that all nodes of the network have conversion capabilities. Therefore, burst photonic networks with limited wavelength conversion lead to more practical solutions. Efficient wavelength assignment algorithm is required for such networks. As a simple but effective approach, PWA (Priority-based Wavelength Assignment) tries to avoid contentions beforehand by assigning different wavelengths to bursts sharing one or more common links, using a wavelength priority database where every wavelength is prioritized for each destination based on the history of prior transmissions. In this paper, we extend PWA to be suitable for burst photonic networks with sparse conversion or limited-range conversion and show the validation of our algorithm through simulation results.

In this paper, inefficiencies of the QoS scheme based on extra-offset-time-based priority for variable-length optical burst-switched network were pointed out, and a novel fixed-length burst assembly scheme is proposed. This scheme is not only practical, but also able to decrease the latency of bursts at the edge nodes. Compared to variable-length (traditional) scheme, a smaller blocking probability for the highest priority service and almost one third offset time delay for every class were obtained in fixed-length scheme.

As a promising switching paradigm, Optical Burst Switch (OBS) was proposed for optical networks showing more efficient and shorter delay than traditional circuit switching, and it is more feasible than optical packet switching. In OBS, burst assembly is one of the critical technologies, and will significantly affect the performance of networks. In this paper, we focus on the study of burst assembly mechanism, and propose a new burst assembly algorithm, which reduces the self-similarity of IP traffic, enhances the data burst utilization and avoids the bandwidth contention due to sending burst data synchronously at different ingress nodes.

Synchronization technology is one of the key technologies in high speed digital communication systems. The bit rate of an EPON system reaches 1.25 GB/s. The receiver in OLT must work in burst mode. Traditional phase-locked-loop technology can not work perfectly in this case and other burst synchronization technologies may be either difficult or expensive to realize such function. In this paper, a Single Bit Burst Synchronization (SBBS) technology, which solves the problems mentioned above, is presented. Because this method gives up the key words detection, it is independent of the protocol used. So it can be used in any PON systems, such as Ethernet PON, ATM-PON, etc. In order to analyze the performance of SBBS technology, we implement a 1.25 GB/s EPON burst receiver as an example, and the results of computer simulation are also provided.

A newly broadband access technology comes into being with the combination of IP and access technique, which is considered as EPON (Ethernet passive optical network). EPON system resolves the limitation of point-to-point, instead uses the topology architecture of point-to-multipoint. It extends the fiber cable to the last mile, so we can get the end-to-end network system with high efficiency, well expansibility and low maintenance cost. There are many key techniques and problems in EPON system, for example, CDR (Clock and data recovery), RTT (Round-Trip Time), DBA (Dynamic Bandwidth Allocation) and ULSLE (Upper-Layer Shared LAN Emulation). The essential character of MPCP (Multi-Point Control Protocol) used in EPON is to schedule the transmission of the upstream data packets, and to avoid data collision. A central counter is applied to synchronize the upstream data through a time-stamp flag. A REPORT control frame is transmitted, which give the length of the different priority queues. On receiving this control frame, central office device modifies the bandwidth allocation in the grant table. The function relating to the ULSLE protocol is similar with IEEE802.1D Bridge, and also includes point-to-multipoint, therefore so many problems are resolved, such as the frame reflection, the communication among different ONUs.

A dynamic bandwidth allocation policy to better support the TDM services in EPON is proposed. The analytical and simulation results prove that the proposed scheme does best in what it is designed to do.

This paper contains a comprehensive review of the various Passive Optical Network (PON) technologies in the marketplace today, namely APON, EPON and GPON, and draws an in-depth comparison between them. Following a review of the history of the various PON flavors, as well as the service requirements set forth by service providers, the emerging Gigabit PON (GPON) technology is examined in detail. System performance between the various protocols is compared using efficiency and scalability factors, and conclusions are drawn as to the overall throughput efficiency of and cost influence on the solution. GPON carries a two-fold promise of both higher bit rates and higher efficiency when carrying multiple services over the PON. It offers a scalable framing structure from 622Mb/s to 2.5Gb/s, as well as support for asymmetric bit rates, exceptionally high bandwidth utilization for any type of service and a variant of standard ITU-T Generic Framing Protocol (GFP) encapsulation of any type of service (both TDM and packet) onto a synchronous transport protocol. It is shown that in the worst-case scenario, based upon the most conservative assumptions regarding traffic distribution, GPON is substantially more efficient, with an overall efficiency of 93% compared to 71% with APON and 49% with EPON. Using a more detailed analysis based upon a traffic model provided by the service providers within the full service access networks (FSAN) consortium, it is shown in quantitative terms that GPON offers exceptionally higher bandwidth for the entire range of applications when compared to both APON and, especially, EPON, resulting in substantially lower cost per bit and a much faster payback period.

In implementing EPON systems, one of key challenges is designing bandwidth scheduling scheme to achieve network efficiency and meet various QoS requirements. This paper proposes a novel scheduling algorithm over EPON based on sliding window mechanism, which divides a transmission circle into two sections and introduces bandwidth sliding window substituting for the conception of fixed maximum window. Computer simulation results show that the performance of the proposed algorithm is effectively improved in terms of packet delay and upstream channel utilization.

On solving the access network bottleneck problem (usually known as the last mile problem), Ethernet passive optical network (EPON) is paid a great many attentions. Many works have been done on TDMA-based EPON system. Architecture of a WDM-based EPON system is proposed in this paper. The proposed system consists of one OLT, 1:32 Splitter/combiner, 32 ONUs and a Network Management System (NMS). The OLT includes a Service Node Interface (SNI), an Ethernet switch, an ODN Interface and a Network Management agent, while the ONU includes a User Side Interface (USI), an Ethernet Frame Forwarding, an ODN Interface and a Network Management agent. The downstream data are broadcast from the OLT to all ONUs with 1.25Gbps of line speed. Each ONU send the upstream data encapsulated in Ethernet frames up to OLT with specific wavelength and the 125Mbps of line speed. The upstream signals from all ONUs are aggregated to one fiber by the 1:32 combiner. When the aggregated stream arrives at the OLT, it is de-multiplexed and converted into 32 pairs of signals for more processing. As a result, all ONUs can send data to the OLT simultaneously. Experimental results show that our WDM-based EPON system works well.

The traffic grooming functionality of two optical packet switched (OPS) node architectures that use a switched wavelength infrastructure are investigated. Three different application scenarios are considered: single-hop grooming with edge-OPS only, multi-hop grooming with edge-OPS only and grooming with edge and core OPS. The advantages and disadvantages of the first two application scenarios are explained by a case study. The third scenario also considers the effects of differing core-OPS placement in different network topologies. One key factor investigated is the effect of aggregation granularity on the grooming performance in the core-OPS scenario.

A genetic algorithm is proposed with permutation based chromosome presentation and roulette wheel selection to solve traffic grooming problems in WDM ring network. The parameters of the algorithm are evaluated by calculating of large amount of traffic patterns at different conditions. Four methods were developed to improve the algorithm, which can be used combining with each other. Effects of them on the algorithm are studied via computer simulations. The results show that they can all make the algorithm more powerful to reduce the number of add-drop multiplexers or wavelengths required in a network.

Integrated routing is a routing approach to support the peer interconnection model in IP over WDM networks. To have a better network link load distribution and network usage in IP over WDM networks, in which network nodes may have the ability to handle traffic in fine granularities, it is important to take into account the combined routing at the IP and WDM layers. Based upon this, this paper develops an algorithm for integrated dynamic routing of bandwidth guaranteed paths in IP over WDM networks. For newly dynamic arriving requests, as the developed algorithm takes into account the combined topology and resource usage information at the IP and WDM layers, and the routing procedure makes full use of the statistic information of the users’ bandwidth requirement and considers carefully both the routing cost and the corresponding length of the routing path, thus a better link load balancing and network usage can be achieved. Simulation results show that the developed scheme performs well in terms of performance metrics such as the number of rejected demands and the network link load balancing.

With the Internet Protocol (IP) being the dominant protocol for new optical network services, there is increasing pressure to optimize the network infrastructure and protocols for IP traffic. This paper presents a technology for constructing a novel architecture for IP over photonic systems. WDM network has been launched by the concept of wavelength routing. The principle is that high-speed data flows, which consist of many time-division multiplexed channels, are associated with specific optical wavelengths. Thus, they are routed through the optical network by means of their wavelengths. Thus, wavelength routing consents the realization of OXCs working in an efficient way. However, MPLS-based forwarding is still an electronic solution. Therefore, even though MPLS needs to establish a closed domain to utilize a new lower-layer technology, it is still useful to incorporate optical technology herein to further increase throughputs for large-scale Internet networks. Moreover, provisioning wavelength capabilities based on MPLS has emerged, that is , multi-protocol lambda switching (MPλS). This system tries to merge the functionalities of the wavelength switching, SONET mux/demux, and IP routing into one layer, and is sometimes known as the concept of optical MPLS and MPLambdaS. Its extension Generalized-MPLS(GMPLS) is also discussed.

When lightpaths are routed in the planning stage of transparent optical networks, the textbook approach is to use algorithms that try to minimize the overall number of wavelengths used in the . We demonstrate that this method cannot be expected to minimize actual costs when the marginal cost of instlling more wavelengths is a declining function of the number of wavelengths already installed, as is frequently the case. We further demonstrate how cost optimization can theoretically be improved with algorithms based on Prim’s algorithm. Finally, we test this theory with simulaion on a series of actual network topologies, which confirm the theoretical analysis.

No Abstract Available

Optical monitoring methods need to evolve from simple optical power and spectra measurements as transparent and all-optical networks begin to be deployed. End-to-end verification of signal integrity and definitive fault localization will need to be assured by in-network monitoring of the noise, jitter and distortion of optical channels affected by optical amplifier impairments, optical nonlinearities, chromatic dispersion, polarization dispersion, and local effects of filtering, crosstalk and multipath interference in network elements. Several novel measurement techniques are described, illustrating the opportunities ahead for advanced optical monitoring.

In this paper, we will discuss the network architectures and the enabling technologies to realize a robust multi-wavelength optical access network. Various network topologies employing the dense WDM or coarse WDM technique; as well as several network enabling technologies such as data re-modulation technique, fiber link protection and traffic restoration, etc. will be described. Besides, some future trends and applications of multi-wavelength optical access networks will also be discussed.

Nowadays, Broadband service subscribers have increased exponentially and have almost saturated in Korea. Several types of solutions for broadband service applied to the field. Among several types of broadband services, most of subscribers provided xDSL service like ADSL or VDSL. Usually, they who live in an apartment provided Internet service by Ntopia network as FTTC structure that is a dormant network in economical view at KT. Under competitive telecom environment for new services like video, we faced with needing to expand or rebuild portions of our access networks, are looking for ways to provide any service that competitors might offer presently or in the near future. In order to look for new business model like FTTH service, we consider deploying optical access network. In spite of numerous benefits of PON until now, we cannot believe that PON is the best solution in Korea. Because we already deployed optical access network of ring type feeder cable and have densely population of subscribers that mainly distributed inside 6km from central office. So we try to utilize an existing Ntopia network for FTTH service under optical access environment. Despite of such situations, we try to deploy PON solution in the field as FTTC or FTTH architecture. Therefore we analyze PON structure in comparison with AON structure in order to look for optimized structure in Korea. At first, we describe the existing optical access networks and network architecture briefly. Secondly we investigate the cost of building optical access networks by modeling cost functions on AON and PON structure which based on Ethernet protocol, and analyze two different network architectures according to different deployment scenarios: Urban, small town, rural. Finally we suggest the economic and best solution with PON structure to optimize to optical access environment of KT.

Ethernet passive optical network (EPON), which represents the convergence of low-cost, high-bandwidth and supporting multiple services, appears to be one of the best candidates for the next-generation access network. The work of standardizing EPON as a solution for access network is still underway in the IEEE802.3ah Ethernet in the first mile (EFM) task force. The final release is expected in 2004. Up to now, there has been no standard application specific integrated circuit (ASIC) chip available which fulfills the functions of media access control (MAC) layer of EPON. The MAC layer in EPON system has many functions, such as point-to-point emulation (P2PE), Ethernet MAC functionality, multi-point control protocol (MPCP), network operation, administration and maintenance (OAM) and link security. To implement those functions mentioned above, an embedded real-time operating system (RTOS) and a flexible programmable logic device (PLD) with an embedded processor are used. The software and hardware functions in MAC layer are realized through programming embedded microprocessor and field programmable gate array(FPGA). Finally, some experimental results are given in this paper. The method stated here can provide a valuable reference for developing EPON MAC layer ASIC.

Ethernet Passive Optical Network (EPON) aims at providing full services to end users over a single platform, which combines the low cost Ethernet equipment and economic fiber infrastructure, appears to be the best candidate for the future broad access network. Due to Ethernet has a seamless connection with modern IP network, EPON has shown excellent performance when transmitting data service. However, a key technical challenge for EPON vendors lies in enhancing Ethernet’s capabilities to support real-time services, such as VoIP, live video and other TDM services, e .g. POTS and T1/E1. In order to guarantee the required QoS, many new technologies are introduced, either common to all IP based network or dedicated to EPON network. Firstly, a short overview of EPON technology is given. Then, two prevailing QoS architectures, Intserv and Diffserv, are presented in detail. Both their advantages and disadvantages are also discussed. Subsequently, the light is turned to a certain DBA algorithm, which is more efficient in EPON. Both the features of the algorithm and the main ideas are described in particular. At last, reserving bandwidth for TDM service, which is simple but not very efficient, is also proposed.

A method for quantitatively analyzing the survivability of EPON access network is presented. The relationship between the survivability of network and the intensity of network damage is obtained. Meanwhile, the comparison of survivability among three kinds of backup of EPON access network is put forward. The simulation result shows that, when the coverage area is the same, the survivability of EPON with Node-Fiber Backup is better than that with Trunk-Fiber Backup, and when the number and distribution of ONU are the same, the survivability of EPON with Bus-Fiber Backup is better than that with Node-Fiber Backup. The result is useful not only for forecasting and evaluating the survivability of EPON access networks, but also for EPON access networks topology design.

Computer simulation is used to evaluate the performance of cascaded multiplexers/ demultiplexers in optical networks. The intersymbol interference (ISI) of non-return-to-zero (NRZ), return-to-zero (RZ) and carrier-suppressed return-to-zero (CSRZ) formats caused by a series of AWG filters is simulated. We also consider the RZ formats with different duty circles transmitted in the optical networks. In order to suppress the power penalty caused by AWGs, a novel method is given out in this article. A synchronous clock is used to modulate the received optical signal so that enhance the clock part in the frequency field of the signal. This method can reduce the power penalty of AWGs and help optical signal to pass through more optical network elements in large area.

It is well known that traditional routing protocols such as OSPF have mature and efficient mechanism to deal with the convergence problem. But, it is different in hierarchical routing ASON network. The optical network has its own distinct routing features, topology aggregation and abstraction used in inter-domain routing make the view of network topology different to each routing layer, all these factors demand a creative way to converge and synchronize the hierarchical network. This paper will depict the architecture of ASON hierarchical routing and the requirement of inter-domain routing. Based on the fundamental depiction, the factors that affect the convergence and synchronization of ASON network will be focused on and critical issues of the reasonable layout of routing layers, topology synchronization between layers, the triggering of topology update etc will be worked over. In this paper, we try to bring forward a kind of parallel topology convergence mechanism depending on the exchange of topology abstracts information, which is suitable for inter-domain routing in multiple layers and domains ASON network. The corresponding performance analysis will also be presented. Also we analyze the update triggering with relative threshold which will be useful to reduce the diffuse range and amount of the update messages in hierarchical routing. These new methods will decrease the convergence time efficiently and improve the efficiency of topology refresh.

This paper studies design and provision problem of OVPN over ASON. Some special issues are dedicated, including address translation, access authentication, OVPN membership management. UNI framework which supports OVPN over ASON is presented. And procedures of OVPN operation are designed.

Optical Virtual Private Network (OVPN) is an increasingly important application of Automatic Switched Optical Network (ASON). To meet the demands of today's different types of network services, dynamic OVPN technique is studied and two new functions are brought forward, which are dynamic connection management, and fast protection/restoration. We realize dynamic OVPN technique in the first ASON testbed in China, based on GMPLS signaling mechanism. The dynamic OVPN technique makes the optical network able to provide customers with reliable and flexible services by more functional and intelligent methods.

This paper presents automatic switched optical networks (ASONs) test-bed based on GMPLS made by Tsinghua University in details. The test-bed contains three plans: transport, control and management. The transport plane is formed with four optical crossconnects (OXCs) and one optical adding and dropping multiplexer (OADM); the control plane is constructed with signaling, routine, link management and UNI modules which are based on GMPLS protocols, and the simplified management plane is developed for the purpose of switched permanent connection. The test-bed possesses changeable topologies and is suitable for researches in optical network with different purposes. The architecture of the test-bed and modules for control plane are presented in the paper. Some experiments within the test-bed are also described.

As we all know, the increasing speed of data traffic is exceeding gradually from voice in today’s communication network. The main reason is the explosive of Internet. The controversy with IP over ATM/SDH/Optical becomes hotter and hotter, Many people in the telecommunication field are doubt: HOW LONG WOULD SDH/SONET BE PROLONGED? WHAT KIND OF SDH EQUIPMENTS COULD BE USED IN THE NETWORK? With the analysis from several aspects: services in the network, new development with SDH technology, market in transport equipment, This paper is considered that the SDH with some new features would be predominant transport technology in the recent years.

SDH/SONET technology is the main manner of information transport. With the increasingly mature related technologies, the combination of data and optical networks become closer. At the same time, the tremendous success and rise of data traffic have put the focus on the protocols and framing technologies for mapping data over SDH/SONET in fiber optical networks. A first attempt at a more efficient high-speed WAN protocol is packet over SDH/SONET, in which IP data packets are conveyed through using the link-layer point-to-point protocol and being encapsulated into byte-stuffed HDLC-like frames. LAPS also using byte-stuffed HDLC offers a simple method of providing Ethernet LAN extension over a public WAN. A generic framing procedure provides an efficient and protocol-agnostic frame delineation and encapsulation mechanism for transporting a variety of protocols over SDH/SONET tributaries. Protocol signals may be PDU-oriented, block-coded, or a constant bit stream. In this paper, we review several emerging technologies for the transport of packet services, and compare these framing technologies known as POS, LAPS and GFP. Finally we present several applications of these technologies.

This paper reviews the recent progress and standardization activities on Ethernet over SONET/SDH technologies, focusing on the needs of the young metropolitan-area network. Furthermore, this paper reports the latest technologies such as LCAS, LAPS/GFP, and MPLS that meet the requests of MAN. Finally a solution for SDH system is given.

This article introduces technical achievement, equipment exploitation, experimental test and project application of the national item "32×10Gb/s SDH Wavelength Division Multiplex system" which is undertook by Wuhan Research Institute of Post and Telecommunications.

This paper discusses and analyzes the designing of SDH embedded with RPR and its applications in MAN. The main contents of this paper include: First of all, we discuss the disadvantages of the traditional Ethernet module embedded in SDH node while it carrying dada service in perspective of network organization, of QoS guarantee and network security. Secondly, we introduce the designing of SDH embedded with RPR, which can resolve the above problems. And finally, this paper puts emphases on analyzing the application models of this kind of SDH nodes in the metro area transport network, and the performance optimization for data services provided by the nodes.

As a result of data traffic’s exponential growth, network is currently evolving from fixed circuit switched services to dynamic packet switched services, which has brought unprecedented changes to the existing transport infrastructure. It is generally agreed that automatic switched optical network (ASON) is one of the promising solutions for the next generation optical networks. In this paper, we present the results of our experimental tests and economic analysis on ASON. The intention of this paper is to present our perspective, in terms of evolution strategy toward ASON, on next generation optical networks. It is shown through experimental tests that the performance of current Pre-standard ASON enabled equipments satisfies the basic requirements of network operators and is ready for initial deployment. The results of the economic analysis show that network operators can be benefit from the deployment of ASON from three sides. Firstly, ASON can reduce the CAPEX for network expanding by integrating multiple ADM & DCS into one box. Secondly, ASON can reduce the OPEX for network operation by introducing automatic resource control scheme. Finally, ASON can increase margin revenue by providing new optical network services such as Bandwidth on Demand, optical VPN etc. Finally, the evolution strategy is proposed as our perspective toward next generation optical networks. We hope the evolution strategy introduced may be helpful for the network operators to gracefully migrate their fixed ring based legacy networks to next generation dynamic mesh based network.

Wavelength Division Multiplexed (WDM) networks that route optical connections using intelligent optical cross-connects (OXCs) is firmly established as the core constituent of next generation networks. Rapid failure recovery is fundamental to building reliable transport networks. Mesh restoration promises cost effective failure recovery compared with legacy ring networks, and is now seeing large-scale deployment. Many carriers are migrating away from SONET ring restoration for their core transport networks and replacing it with mesh restoration through “intelligent” O-E-O cross-connects (XC). The mesh restoration is typically provided via two fiber-disjoint paths: a service path and a restoration path. this scheme can restore any single link failure or node failure. And by used shared mesh restoration, although every service route is assigned a restoration route, no dedicated capacity needs to be reserved for the restoration route, resulting in capacity savings. The restoration approach we propose is Centralized Pre-computing, Local Distributed Optimization, and Shared Disjoint-backup Path. This approach combines the merits of centralized and distributed solutions. It avoids the scalability issues of centralized solutions by using a distributed control plane for disjoint service path computation and restoration path provisioning. Moreover, if the service routes of two demands are disjoint, no single failure will affect both demands simultaneously. This means that the restoration routes of these two demands can share link capacities, because these two routes will not be activated at the same time. So we can say, this restoration capacity sharing approach achieves low restoration capacity and fast restoration speed, while requiring few control plane changes.

With the explosive growth of data services, SDH/SONET network, which is designed for voice traffic, is required to be optimized for data services. SDH-based Multi-Service Transport Platform (MSTP), which greatly improves the data capabilities of SDH networks, has been attracting much interest of vendors as well as network operators. As a leading operator in China, we have conducted a field trial study and extensive experiments on MSTP. In this paper, we present the results of our experiments. It is verified via extensive experiments that the products from some leading vendors conform to the standards. In addition, it is shown that the performance of the new value-added metro Ethernet services is highly satisfied, i.e., with tight QOS guarantees and flexible bandwidth that fits the user’s requirement exactly. We have also performed experiments for multi-vendors interoperability, and the results are highly satisfied, paving the way for large scale deployment of 2G MSTP. Our trial work showed MSTP can be used to optimize the existing metro networks, i.e., to improve the utilization of the fiber and the bandwidth that are used to carry the ADSL traffic and the IP traffic. In addition, MSTP presents a new array of opportunities for operators to increase their revenues.

The Global Seamless Network (GSN) Demonstrator is presented, a joint effort of system vendors and Deutsche Telekom Group R&D to demonstrate network functions and management integration and enable, for the first time, experiences with a carrier grade Automatically Switched Transport Network (ASTN) implementation and the envisaged main ASTN clients, IP and Ethernet. For end-to-end monitoring capability, integrating the view on the ASTN and Ethernet-MAN configuration, an UMS (Upper Monitoring System) is being developed. Furthermore broadband application were implemented to visualise the network functions. The ASTN backbone consists of four cross connects and an ULH-WDM system with 3x 10Gbit/s channels (OCh) between Berlin and Darmstadt, whereby each OCh is treated as a virtual fibre.

Closed all-optical cycles in wavelength division multiplexing (WDM) optical networks may generate lasing resulting in adverse effects to network performance. In this paper, causes and physical conditions for both all-optical idle cycles and all-optical leakage cycles are analyzed by using optical spectrum of the network. A five-node experimental optical network is built. Multiple lasing peaks are observed to co-exist and compete in a fiber. Affects of the filtering characteristic mismatch of optical nodes to all-optical lasing cycles are discussed. We point out that multiple closed leakage cycles may occur in one optical channel in the case of bandwidth mismatch. Two lasing peaks of leakage cycles are observed in one channel and they can co-exist with the signal wavelength. Finally, harms of lasing cycles are discussed and methods are proposed for eliminating them.

The concept of Next Generation Network (NGN) was conceived around 1998 as an integrated solution to combine the quality and features of the PSTN with the low cost and routing flexibility of the Internet to provide a single infrastructure for the future public network. This carrier grade Internet solution calls for the creation of a consolidated, packet transport and switching infrastructure and the development of a flexible, open, software switch (softswitch) to handle voice telephony as well as multimedia services. Almost all the telecom equipment manufacturers as well as some Internet equipment vendors immediately subscribed to this vision and joined the race to create convergent products for the NGN market.

To improve the reliability of the increasing network two disjoint paths should be found between a given source and a given destination. The problem of finding two minimum cost node-disjoint/edge-disjoint paths with different costs in a directed network can be formulated as a linear integer programming problem minimizing the sum of the costs on the edges in two paths, which is strongly NP-complete problem. Linear relaxation programming which relaxes the integer variables in the original programming is often applied to solve this NP problem. Comparing with linear relaxation programming, Lagrangean relaxation affords a lower bound of the objective value of original programming. Based on this a Lagrangean relaxation method for solving two disjoint paths is presented after a mathematical programming model of the problem is established. By using a modified subgradient optimization technology a new algorithm to solve the Lagrangean relaxation is put forward. The complexity of the proposed algorithm is as same as the Dijkstra’s algorithm (O(n²)). The efficiency of this algorithm is demonstrated by test examples.

The demand to carry large amount of data as fast and as reliably as possible is continuously increasing, in parallel with formation of the information society. Nowadays, Fiber optic cables using Wavelength Division Multiplexing (WDM) are used to provide huge transport capability to support such demand effectively and economically. Each link is able to carry huge amount of traffic, thus a possible failure causes loss of tremendous data. So network survivability is highly required for such a high speed and huge capacity network. The operation of the modern telecommunication network is based on a multi-layer structure. The major reason of this multi-layer is the multiplexing hierarchy for efficient use of transmission facilities and the convenience of service and facility management. The most common example of multi-layer telecommunication network is the ATM over SDH over WDM configuration. This work investigates to the survivability of such multi-layer networks, and an integrated spare capacity dimensioning scheme, joint network design (JND) method, is proposed. Compared with single layer design (SLD) method, JND is bandwidth efficient and cost effective. The configuration problem of multi-layer fault tolerant networks is formulated as an Integer Linear Programming model. We also present a genetic algorithm approach for large networks whose size makes the ILP model difficult to use.

We consider the problem of designing an optical survivable mesh network with undetermined topology. In this class of problems, only the geographic locations of n network nodes and the traffic demand between these node-pairs are given. The objective is to optimize the total network cost that is characterized by three parameters for each link, namely, the installation cost (α), the bandwidth cost (β) and the equipment cost (γ). We tackle this NP-hard problem by incorporating a modified drop algorithm (MDA) or a genetic algorithm (GA) into the previously developed heuristic called SCAPE. We have previously shown that joint optimization of the topology design, working- and spare-capacity planning using MDA for a 10-node network can result in 20% cost reduction over separate optimization planning. Here, we compare the performance of GA versus MDA with respect to 10 different sets of cost parameters. It is shown that GA slightly outperforms MDA when β is dominant while MDA outperforms GA by about 20% when α is dominant. By eliminating the nodal degree effect, we also show that the joint optimization of MDA with SCAPE result in a 14% net improvement.

In WDM wavelength routed networks, prior to a data transfer, a call setup procedure is required to reserve a wavelength path between the source-destination node pairs. A distributed approach to a connection setup can achieve a very high speed, while improving the reliability and reducing the implementation cost of the networks. However, along with many advantages, several major challenges have been posed by the distributed scheme in how the management and allocation of wavelength could be efficiently carried out. In this thesis, we apply a distributed wavelength assignment algorithm named priority based wavelength assignment (PWA) that was originally proposed for the use in burst switched optical networks to the problem of reserving wavelengths of path reservation protocols in the distributed control optical networks. Instead of assigning wavelengths randomly, this approach lets each node select the “safest” wavelengths based on the information of wavelength utilization history, thus unnecessary future contention is prevented. The simulation results presented in this paper show that the proposed protocol can enhance the performance of the system without introducing any apparent drawbacks.

This paper proposed a novel RWA algorithm based on load balance (RWALB) to solve the dynamic wavelength routing problem in new generation optical networks based on the MPLmS paradigm. In RWALB algorithm, the cost of each link is adjusted dynamically as the overall network status changing. By applying RWALB algorithm the best load balance can be produced in the overall network wavelength resource, the average blocking probability can be reduced significantly and the wavelength resources can be utilized efficiently. Through numeric simulation on NSF and EON network, it is shown that the network performance is significantly improved compared with ordinary RWA algorithm that applies shortest-path algorithm and first-fit (FF) wavelength assignment strategy.

As known, the QoS routing under multiple constraints, which can be additive, multiplicative and concave, is a NP-complete problem, and is difficult to get a polynomial-time solution. Currently, many heuristic algorithms have been proposed to solve this problem, where the path distance usually acts as the most important factor during the routing decision-making. However, most of heuristic algorithms are based on linear path distance, which has the greatest shortcoming, i.e., the search of feasible solutions is inefficient. Besides, they also consider the link sharing interference and load balancing. To improve their performance, this paper proposes a novel QoS routing for multiple constraints based on the nonlinear path distance, which is called MIS-LB. At first, this paper represents the additive and multiplicative QoS metrics in the uniform metric, i.e. the nonlinear path distance. Then, the paper also brings the link sharing and load balancing into the uniform metric. Based on the shortest-first criterion, MIS-LB can not only find feasible paths as the normal routing algorithm (e.g., Dijkstra algorithm), but also adjust the link sharing, and balance the network loads between multiple feasible paths. So compared with TAMCRA and H_MCOP, the simulation results show that MIS-LB can further improve network performance.

In this paper, we examine the QoS issues in the intelligent optical network (ION), and specially focus on the QoS implementation scheme associated with the control plane. Referring to the QoS models in IP networks, we propose the optical IntServ QoS model using in metropolitan optical networks, and the optical DiffServ QoS model using in core optical networks. The networks to support optical IntServ QoS model must deal with each traffic request for QoS assurance, then compute routes and establish lightpaths. In the optical DiffServ QoS model, core networks only accept the limited classes of QoS requests. Therefore traffic flows with similar QoS requirements from the metropolitan networks must be aggregated, and then core networks provide tunnels for them. It increases the scalability of networks. In the end, this paper presents the generalized control plane architecture of the node in ION, which is based on extended GMPLS and able to guarantee subscriber’s QoS using two optical QoS models.

The optimization of multicast routing is discussed and the delay-constrained minimum cost multicast tree is constructed by simulated annealing algorithm. A model of multicast routing is presented and the result of simulation is shown. Compared with genetic algorithm , simulated annealing algorithm has the ability to get the optimal solution in a short time. The result indicates that the simulated annealing algorithm has characteristics such as high-efficient in multicast routing.

The future of TMN in ITU-T will be heavily influenced by new telecom technologies, such as IP, and associated management needs. To meet this challenge, SG 4 has adopted two approaches. Current focus of ITU-T SG4 is common working methods for specifying protocol-neutral TMN requirements and information/models and identification of key management technologies to meet future needs. ITU-T SG4 has successfully set up the CORBA framework. The core is re-using the CORBA Common Object Services. ITU-T defines a set of CORBA interfaces for the CORBA generic information model. These interfaces are translated manually from a set of M.3100 GDMO managed object classes following the TMN CORBA framework and guidelines. Although TMN CORBA framework has finished, it still raises new issues, such as implementation conformance and complexity of additional services. It is remarkable that China has made great progress on CORBA-based network management standardization. ITU-T adopts some correlative technologies. SG4 agreed to define the Integrated Management of Hybrid Circuit Switched and Packet/IP Networks. SG4 also make progress in tML, Qos, ETS.

We work on the network resource allocation issue concerning network management system function based on market-oriented mechanism. The scheme is to model the telecommunication network resources as trading goods in which the various network components could be owned by different competitive, real-world entities. This is a multidisciplinary framework concentrating on the similarity between resource allocation in network environment and the market mechanism in economic theory. By taking an economic (market-based and game theoretic) approach in routing of communication network, we study the dynamic behavior under game-theoretic framework in allocating network resources. Based on the prior work of Gibney and Jennings, we apply concepts of utility and fitness to the market mechanism with an intention to close the gap between experiment environment and real world situation.

This paper concerns about optical switching transport network planning methodologies and the associated economic perspective analysis. Mathematical expressions for algorithms to determine network elements in various scenarios such the segmented networks as well as the hybrid-granularity demands are developed for modeling network architectures.

Dual-ring networks are known to provide higher throughputs and reliability than other competing ring technologies. Resilient Packet Ring (RPR) is a Media Access Control (MAC) layer protocol that operates over a double counter-rotating ring network topology. RPR is designed to enhance Synchronous Digital Hierarchy(SDH) in order to handle data traffic more efficiently. Since Intelligent Protection Switching(IPS) is one of the key technologies in ring networks, RPR provides two intelligent protection algorithms: steering and wrapping. While wrapping in RPR in essence inherits the automatic protection switching(APS) algorithm of SDH, it also wastes the bandwidth on the wrapping ringlets and may result in severe congestion. Whereas steering in RPR provides high bandwidth utilization, its switching speed is low because that it is indeed a high layer’s restoration algorithm. In this paper, an effective algorithm integrated self-healing(ISH) algorithm for RPR is proposed, which synthesizes the merits of the two algorithms by transporting healing signal and computing routing in MAC layer. At last, the performance of ISH algorithm is analyzed and simulated.

In the last five years, the traffic growth rate in China has been extremely fast. By 2005, the number of wired telephone customers is estimated to reach 220 to 260 million, while the number of expected cellular customers will reach 260 to 290 million. To meet these challenges, we will continue evolving with more wavelengths and higher speed. By evolving point-to-point WDM systems to OTN/ASON systems, we can eliminate the throughput bottleneck of network nodes caused by electronics, provide optical-layer bandwidth- management capability, provide scalability (which allows continuous traffic growth and network expansion), and provide reconfigurability (which allows semi-dynamic and dynamic optical networking). We can also simplify and speed up provisioning of high-speed circuits and services and offer fast network protection and restoration on the order of tens or hundreds of milliseconds to guarantee excellent network and service survivability. The CMCC (China Mobile Communication Company) will build its OTN network towards the ASON. The CMCC’s long-haul national network utilizing OXC has clearly becomes an intelligent network. It offers end-to-end point-and-click provisioning, shared mesh restoration with a few tens to a couple of hundred msec restoration times, re-provisioning of connections in the event of double failures and network capacity that is not optimally used. In this paper, first we present the CMCC network situation, The network planning tool will be introduced, Then we compare ring with mesh solution in terms of the cost, network performance, protection and restoration, network re-optimization. At last we derive a desired conclusion.

Detection rate is vital to intrusion detection. We propose a new search algorithm base on probability to speed up the process rate for a novel compound intrusion detection system (CIDS). We employ an improved Bayesian decision theorem to build this compound model. The improved Bayesian decision theorem brings four profits to this model. The first is to eliminate the flaws of a narrow definition for normal patterns and intrusion patterns. The second is to extend the known intrusions patterns to novel intrusions patterns. The third is to reduce risks that detecting intrusion brings to system. The last is to offer a method to build a compound intrusion detection model that integrates misuse intrusion detection system (MIDS) and anomaly intrusion detection system (AIDS). During the experiment of this model, we find that different system calls sequences have different probabilities. So the sequences with high probabilities are compared prior to an observed sequence, which is the foundation of our new search algorithm. We evaluate the performance of the new algorithm using numerical results, and the results show this new algorithm increases the detection rate.

In this paper, data mining technologies are used to analyze and extract features that can distinguish normal activities from intrusions. Based on the common model CIDF, we present an IDS framework with an embedded data mining module to improve accuracy of IDS. Three subsystems (including monitor system, data process system and decision-making system) in the framework are introduced respectively. Using experiments on mining network connection features, we present a decision-tree classification algorithm, which uses data set of network connection features as training data set to build decision tree. Using system behaviors as new samples and testing their attributes on the decision tree can recognize anomalies and unknown intrusions accurately.

In a modern computer system, intrusion detection has become an essential and critical component. Data mining generally refers to the process of extracting models from large stores of data. The intrusion detection system first apply data mining programs to audit data to compute frequent patterns, extract features, and then use classification algorithms to compute detection models. The most important step of this process is to determine relations between fields in the database records to construct features. The standard association rules have not enough expressiveness. Intrusion detection system can extract the association rule with negations and with varying support thresholds to get better performance rather than extract the standard association rule.

Effective management of computer networks has become a more and more difficult job because of the rapid development of the network systems. Fault identification is to find where is the problem of the network and what is it. Data mining generally refers to the process of extracting models from large stores of data. We can use data mining techniques to help us in the fault identification task. Existing approaches of fault identification are introduced and a new approach of fault identification is proposed. This approach improves MSDD algorithm but it need more computation. So some new techniques are used to increase the efficiency.

In this article, we mainly focus on the integrated dynamic survivability in IP/MPLS/WDM networks under the peer-modeled architecture. We systematically investigated various integrated dynamic protection and restoration schemes combining knowledge of resource and topology information in both the IP/MPLS and WDM layers. To allocate the restorable bandwidth-guaranteed paths, we propose an integrated routing algorithm: Load-Balanced Physical Resource Minimum Hop (LBPRMH) integrated routing algorithm. In this algorithm, we present a new link cost function used in Dijkstra’s algorithm considering both physical resource and load balancing factors and employ a parameter (relevant to traffic intensity) to control the influence of two kind of information. When network load is low (traffic intensity is small), the main factor in the proposed link cost function is physical resource, which leads to high resource utilization. When network load is high, the idea of load balance becomes more important in routing, which avoids connection requests conflict and increases the efficiency of resource usage. Then extensive simulations are done to analyze and compare the dynamic survival performance of various integrated survival strategies. Resource utilization, protection efficiency, restoration efficiency are employed as metrics to evaluate the survivable performance. We demonstrate that the proposed integrated routing algorithm performs better than some existing approaches in terms of the above three metrics. The simulation results show the high resource utilization in optical layer recovery with wavelength granularity and the high restoration flexibility in IP/MPLS layer recovery with finer granularity. If the recovery action is implemented first in lower layer and then higher layer if the first action failed, the integrated protection and restoration schemes can combine the advantages of both layers and perform better than recovery schemes in either single layer.

P-cycle based protection has been proved to be an effective way for WDM networks’ survivability design. In IP over WDM networks, p-cycle can also be constructed through GMPLS technology to protect a number of IP links. In this paper, we demonstrate that there are some drawbacks associated with the current p-cycle searching algorithms, e.g., 1) none of the them take the hop-count constraint into account, which means some requests may not able to be protected by their associated p-cycles; 2) to protect those failures occurred on the straddling links, the algorithms demand the traffic to be split in half over each p-cycle segment, which may not be a resource efficient solution, and the routing procedure may be prohibited by hop-count constraints, also, it may introduce extra link propagation delays and causes IP packets to arrive out of order. We propose an improved object function, and give a novel p-cycle searching algorithm. Simulation results obtained from CERNET-like physical topology and the national backbone networks show that the developed algorithm performs well in terms of performance metrics such as the connection blocking probability and the success rate of the protection.

This paper presents a distributed scheme for link failure recovery in mesh optical networks, based on the use of network hierarchical spanning trees. The scheme intends to maximize restorability in a network with known working and spare capacities. The hierarchical protection tree (p-tree) provides the hierarchical layering of the network. The straddling links that are not located on the tree are protected through tree branches to the higher layer Parent nodes. The links on the tree are protected by links to backup parent nodes. We study the problem of finding the most optimized network tree to achieve maximum restorability, and present heuristics for finding the best tree. Our algorithm includes two steps: Selection of the best root node for the tree, and construction of the tree based on distribution of tree ID labels among the nodes. Each node selects a primary parent node and a backup parent node, and constructs pre-determined protection paths accordingly. In case of failure, all connections on a link are switched quickly to the protection path as a bundle. We perform restorability analysis for several real and arbitrary long-haul networks and show that our scheme provides excellent network restorability along with exceptional scalability and maintainability.

The survivability of IP over WDM network gains importance as the volume of traffic to be transported by the network keeps growing. In this article we investigate the restoration schemes for handling dynamic traffic demand failures in IP/MPLS over WDM networks. We classify the various restoration schemes into two broad categories: recovery at the lowest layer and recovery at the highest layer. Six specific restoration schemes will be introduced and analyzed. Simulation will be performed to compare the network utilization of the six restoration schemes.

As the digital network is expanding so quickly, many applications are widely used in the social living, especially wireless LAN and IP telephone. The DTE(Data Terminal Equipment) based on Ethernet are also used in many place, these equipments have low power cost, but commonly they have very high requirements in stability and reliability of power supply. This article brings forward a system solution project of long-distance centralize power supply, which can be controlled by network, according to the 802.3af standard. And also presents the key technology in this solution.

Nowadays VoIP (Voice over IP) develops quickly. H.323 is the most popular and mature architecture for VoIP, in which Gateway is one of the most important components. In this paper, two main aspects of the Gateway are discussed. They are the processing of simultaneous initialization of multiple calls and the compatibility with other H.323 standard endpoints and different H.323 versions. At last, the test of our Gateway is shown and the result is discussed.

With broadband access network development, the Ethernet technology is more and more applied access network now. It is different from the private network -LAN. The differences lie in four points: customer management, safety management, service management and count-fee management. This paper mainly discusses the safety management related questions. Safety management means that the access network must secure the customer data safety, isolate the broad message which brings the customer private information, such as ARP, DHCP, and protect key equipment from attack. Virtue LAN (VLAN) technology can restrict network broadcast flow. We can config each customer port with a VLAN, so each customer is isolated with others. The IP address bound with VLAN ID can be routed rightly. But this technology brings another question: IP address shortage. VLAN aggregation technology can solve this problem well. Such a mechanism provides several advantages over traditional IPv4 addressing architectures employed in large switched LANs today. With VLAN aggregation technology, we introduce the notion of sub-VLANs and super-VLANs, a much more optimal approach to IP addressing can be realized. This paper will expatiate the VLAN aggregation model and its implementation in Ethernet access network. It is obvious that the customers in different sub-VLANs can not communication to each other because the ARP packet is isolated. Proxy ARP can enable the communication among them. This paper will also expatiate the proxy ARP model and its implementation in Ethernet access network.

This paper investigates the problem of dynamically establishing dependable connections in traffic-grooming wavelength division multiplexing (WDM) mesh networks, and proposes a dynamic shared sub-path protection (SSPP) scheme for multi-granularity traffic. To establish a dependable connection, SSPP searches a primary path for each connection request, and then it segments the found path into several equal-length sub-paths, and computes their corresponding backup paths, respectively. In SSPP, if two sub-paths are fiber-disjoint then their backup paths can share backup resources to obtain the optimal spare capacity. Based on dynamic traffic with different load, the performance of SSPP has been investigated via simulations. The results show that SSPP can make the tradeoffs between the resource utilization and the restoration time.

Efficiently grooming low-rate traffic flows into high-capacity lightpaths will improve the network throughput and resource utilization. In this paper, we study the traffic -grooming algorithm for WDM mesh networks. A novel dynamic grooming graph is proposed, which models the number of transceivers per node and the number of wavelength per fiber in addition to the combined information of IP layer and WDM layer, while taking into account the constraints of wavelength continuity in optical domain. Based on the grooming graph, a new dynamic traffic-grooming algorithm, namely Integrated Grooming Algorithm (IGA), is developed to provide an efficient solution to the traffic-grooming problem in WDM optical mesh networks. This algorithm is evaluated via simulations and results reveal that satisfactory performance of the proposed algorithms can be achieved.

Optical fiber links, due to sharing some common physical resources, e.g., fiber cable, conduit and Right of Way, have certain failure-dependent. The term "Shared Risk Link Groups (SRLG)" is introduced to describe the relationship between links with shared risk; SRLG diversity has become the important constraint of protection design problem. This paper addresses SRLG-constraint-based shared-path protection design problem with differentiated reliability. Via SRLG conditional failure probability concept, we propose an efficient PSD-SPP (Partial SRLG-disjoint Shared-Path Protection) algorithm to solve the protection design problem with differentiated reliability under the consideration of links failure-dependence. Simulation result shows that this algorithm not only efficiently satisfies the user-specific requirement, but also effectively decreases the network-blocking ratio.

In this paper, firstly, we introduce the protocol of VoIP, and analyze the process of a call establishment based on H.323, including RAS, H.225, and H.245 phases. Then, we focus on implementing ALG, which allow an application on a host in one address realm to connect to its counterpart running on a host in different realm transparently. An ALG may interact with NAT to set up state, use NAT state information, modify application specific payload and perform whatever else necessary to get the application running across disparate address realm. In addition, this section presents many important data structures and algorithms, for example, the algorithm to check for match of remote IP address and port, the algorithm for classifier to look at both outbound as well as an inbound packets, and etc. Finally, the paper describes and analyzes the results of experiments, and we obtain some significant parameters, namely, call setup latency and packet forwarding latency. The experiments and applications demonstrate it feasible for ALG to realize NAT for VoIP.

The problem of routing and wavelength assignment (RWA) in WDM (wavelength division multiplexing ) networks is addressed in this paper. Most of the attention has been devoted to RWA algorithm. In this article, some concepts on RWA are studied first. Then we studied the associated research problems and challenges on RWA. Various RWA approaches are examined and compared. Finally, we proposed a new RWA algorithm based on multi-objective. In this new algorithm, we consider multiple network optimizing objectives to setup a lightpath with maximize profit and shortest path under the limited resources. By comparing and analyzing, the proposed algorithm is much better than the algorithms, which only consider one optimizing objective. This new RWA algorithm may be useful to a network maker.

In WDM networks, the bandwidth request of a traffic stream is usually much lower than the capacity of a wavelength. Traffic grooming can aggregate low-rate connections onto high-capacity lightpaths to make efficient use of the bandwidth. But most of the researches related to traffic grooming focused on ring networks, and fewer concerns are about the dependable traffic grooming of low-rate connections in WDM mesh networks. In our study, we present a Hamiltonian Cycle Protection based Traffic Grooming algorithm (HCPTG) considering both the survivability and traffic grooming in WDM mesh networks. Simulation results show that good performance can be achieved in terms of capacity efficiency.

As a new solution of all optical networks, Optical burst switching (OBS) has been wildly studied in recently years for its high traffic throughput and high resource utilization at current technology level. In this paper we address some issues of efficient contention resolution: burst segmentation and deflection routing. Under the Just-Enough-Time (JET) protocol based OBS networks architecture, control packet is the guider for corresponding data burst during the transmission. Taking these into consideration, here we propose a novel segmentation strategy - current burst segmentation (CBS) and a modified deflection routing strategy fit for JET based OBSN. Through analysis and simulation, it is shown that these new strategies perform well.

All-optical networks are emerging as a promising technology and have absorbed many eyes. However they are intrinsically different from traditional electro-optical networks, so their security is also different. The countermeasures are more important because of the high data rates of the networks and must be reconsidered carefully. This paper discusses the still unstudied security and new possible countermeasures.

In this paper, according to the ITU-T recommendation G.8080, we present a novel architecture of RC based on the extended GMPLS, which consists of interface component, OSPF-TE component, route computation component and routing information database. The RC is emulated in the ASON simulation software developed by us, where the OSPF is extended as the routing protocol and GMPLS RSVP-TE is adopted as signalling protocol. In the simulation, the RWA scheme, the fixed shortest path routing way and first fit wavelength algorithm, is used. The simulation results of RC have been presented in terms of flooding time (FT), flooding overhead (FO) and route computing time. The FT and FO just have a very little change under different traffic load, which indicates that re-flooding is rare and the RC works efficiently.

The rapid progress on optical performance monitoring technology makes it possible for optical network designers to guarantee the QoS quality of lightpaths in Wavelength Division Multiplexing (WDM) network. This article describes the basic idea of online QoS guarantee scheme, focusing on the distributed optical noise figure calculation algorithm and optimal design heuristic.

The emergence of intelligent optical network (ION) not only introduces the function of intelligent connection control, but also makes some new-style optical layer services possible. There are many kinds of survivability schemes applied in ION with mesh topology. According to the recovery degree of traffic, differentiated resilience service (DRS) can provide an appropriate survivability scheme from them. In this paper, a differentiated resilience service scheme in intelligent optical network is proposed. We firstly discuss some relative problems of DRS, and then define some DRS degrees in terms of recovery time for traffics. At last, the DRS protection/restoration modules of control plane in ION are designed and analyzed. Thus a DRS implementation method in ION is formed.

At this point in technology's evolution, the simplicity, elegance, extensibility, and broad compatibility of the Internet protocol suite has made it the automatic choice for most forms of communication. The attempts at resolution of this apparent dichotomy consist of a collection of technologies and philosophies known as Quality of Service. In an IP network, QoS defines the ability to compensate for traffic characteristics without compromising average throughput. Clearly, optimizing QoS performance for all traffic types on an IP network presents a daunting challenge. To partially address this challenge, several Internet Engineering Task Force groups have been working on standardized approaches for IP-based QoS technologies. The IETF’s approaches fall into four categories: prioritization using differentiated services, reservation using integrated services, label switching using multi-protocol label switching, bandwidth management using the subnet bandwidth manager. Differentiated services classify per-hop behaviors on the basis of a Diffserv code point attached to the type of service byte in each packet’s IP header. This DSCP approach represents a form of soft QoS that rather coarsely classifies services through packet marking. The differentiated QoS routing in GMPLS-based IP/WDM Networks are a promising candidate for the next generation optical Internet networks. By using a unified control plane, such networks make more efficient usage of network resources both at the IP layer and the WDM optical layer. In this paper, we consider prioritized routing of bandwidth-guaranteed Label Switched paths (LSPs) providing service differentiation between classes of high and normal priority traffic. The QoS delay requirements are assumed to be translated into bandwidth and O-E-O conversion requirements. We present a graphical representation of the integrated network state which is different from other conventional graphical representations in that it models the cost of usage of ports and o-e-o conventions in addition to the usage of wavelength and bandwidth resources. We then develop a threshold-protection-based routing algorithm which admits high-priority LSPs in preference over normal-priority LSPs and satisfies the bandwidth and o-e-o constraint requirements. Through extensive simulation experiments, we approve the service differentiation and effectiveness of our algorithm.

In this paper, we describe a solution of how EPON support multi-service access through layer 2 and its protection restoration scheme, which are fully compliant with current EPON standard. Through the class-of-service mechanisms built in and flexible protection restoration scheme, EPON can provide a QOS guarantee and high reliability while keep the simplicity and cost effectiveness of Ethernet.

The authors propose a multiwavelength bi-directional optical crossconnect (B-OXC) structure based on 1x16 multiplexer/demultiplexer (MUX/DMUX) and 2x2 optical switches. The 2x2 B-OXC system is implemented using one pair of array waveguide grating (AWG) based MUX/DMUX and all channels can be cross-connect independently in either direction. The bit error rate (BER) performance of the bi-directional WDM add/drop multiplexer is experimentally studied for a data rate of 2.5 Gb/s per channel, providing an overall capacity of 40 Gbit/s (2.5 Gb/s x 16 chs) in a 400 Km transmission. The worst channel crosstalk level is -37 dB and the typical switching time is 10 ms. We found that the performance of the B-OXC signals is only 1.0 dB degraded by a backward propagating signal.

Fault location in WDM all-optical networks is a very important problem in practical network operation and administration. In this paper a simple and efficient method based on causal chain and set theory is proposed for the problem of fault location according to the current alarms in one network. A detailed fault location algorithm for single point fault is implemented based on the algorithm theory mentioned above. This method combines the network configuration model with the network fault model using object-oriented idea. Some tags and attributs related to fault and alarm are extended in configuration model which reduces the memory space and quickens the seeking. A WDM all-optical network connection management system using this algorithm is applied to a three-node WDM network. The testing results show that the proposed algorithms are correct and feasible. The average time for fault location is 9 seconds and the average time for restoration is 14 seconds.

A novel, simple and controllable bandwidth allocation scheme, called bandwidth on Demand (BoD), is proposed for optical burst-switched network in this paper. It assures every service class to consume bandwidth no more than its granted percentage by examining the bandwidth usage of each class periodically. The burst loss ratios under arbitrary input loads are theoretically analyzed as well as numerically simulated. The theoretical descriptions accord with the simulated results very well. The scheme can be used as an effective scheme of bandwidth allocation and management.

The conservative cascades process with multipliers generated by symmetric β distribution is introduced as an EPON upstream traffic model, which can catch the amplitude burst and temporal burst behaviors of real traffic simultaneously. The traffic matching and synthesizing methods are illustrated through an example.

This paper proposes a new static routing algorithm applying Traffic Engineering, which integrates Multiprotocol Label Switching (MPLS) and Quality of Service (QoS) Routing. Because of using MPLS, centralized control is applied to the transmission paths of different service type in the algorithm. At the same time, to select LSP based on the state of networks and the requirements of QoS, the algorithm can make the resource using globally optimal. It avoids the traditional routings’ shortage that the network congestion is produced by the disequilibrium of resource using. United object strategic in the algorithm can produce effective projects for the problem of satisfying Multi-requirement in one routing count, which is NP-hard. Finally the paper proves that the algorithm is feasible and preferable by computer simulation and theoretical deduction.

Differentiated Service was proposed by IETF in order to ensure various Quality of Service requirements of a wide range of Internet applications. Due to different service levels of traffic, traditional routing architectures can’t do a competent job in future networks and new Constrain-Based Routing mechanism which can deal with classified traffic should be constituted. We put forward a novel static routing scheme which can (1) meet different requirements of routing traffic; (2) maximize operating benefit of ISPs while satisfying restrictions of traffic and resource; (3) optimize network resource allocation. The formular model of this algorithm was proposed and simulation results showed that the proposed method was feasible and advantaged.

All factors of Dense Wavelength Division Multiplexed (DWDM) Wireless Optical Networks relate with one another. Among them, signaling is based on the premise of getting high efficiency of optical bandwidth utilizing, and also with ultra low latency and the lowest rate of data to be dropped. Networks running under the direction of Preliminary Survey with Amending Algorithm (PSAA) may get high bandwidth utilizing efficiency with high Quality of Service (QoS). Our proposed architecture of Area Coverage Technology (ACT) is formed from Substantive Wavelength Channels Distributor (SWCD) allowing cooperation with Distributed Controller (DC). Finally, some excellent network performance is revealed by simulation and experiment.

Free Space Optical technology (FSO) is transparent to communication protocols, the most efficient way to utilize it is wireless optical network. Founded on unique characteristics of carriers, the proposed network system disposes every link fitting-in bit-rate with appropriate frame, then coordinate the topology, address, route technology with management technology as well as message technology, so it has the high communication capacity with being controlled delay time from every Class of service (CoS). Ultra low packet loss rate and ultra low latency lay the foundation of it serving for different communication modes with high Quality of Service (QoS). All works will be done from the physical layer to the network layer.

Nowadays NGN (Next Generation Network) is the hotspot for discussion and research in IT section. The NGN core technology is the network control technology. The key goal of NGN is to realize the network convergence and evolution. Referring to overlay network model core on Softswitch technology, circuit switch network and IP network convergence realized. Referring to the optical transmission network core on ASTN/ASON, service layer (i.e. IP layer) and optical transmission convergence realized. Together with the distributing feature of NGN network control technology, on NGN platform, overview of combining Softswitch and ASTN/ASON control technology, the solution whether IP should be the NGN core carrier platform attracts general attention, and this is also a QoS problem on NGN end to end. This solution produces the significant practical meaning on equipment development, network deployment, network design and optimization, especially on realizing present network smooth evolving to the NGN. This is why this paper puts forward the research topic on the NGN network control technology. This paper introduces basics on NGN network control technology, then proposes NGN network control reference model, at the same time describes a realizable network structure of NGN. Based on above, from the view of function realization, NGN network control technology is discussed and its work mechanism is analyzed.

A multi-time scale congestion control mechanism has been proposed in this paper, which utilizes the characteristic of self-similar in the network. Based on the theory of the alpha-stable processes, a prediction model is brought forward and a frame of multi-time scale is designed. The aim of all of those is try to avoid the congestion of network in statistically. Moreover, the frame of the new congestion control mechanism is described and the working procedure of the new congestion control mechanism is explained.

For the past few years, telecom companies have been working diligently to provide us with pseudo-broadband Internet connections over copper (DSL) and cable (cable modem). I use the term "pseudo-broadband" because the existing telecom infrastructure can only provide speeds of up to 1.5 megabits per second. (In theory, cable modem can provide up to 2.5 megabits per second, but in reality nobody obtains these speeds because the shared aspects of cable modem results in lower speeds.) No doubt improvements will be made over the next few years to squeeze more out of copper and cable, but it doesn't matter, because fiber to the home is coming, and it will be here faster than most people predict. In case you're wondering, FTTH provides download speeds of up to 155 megabits per second -- that's 100 times faster than the pseudo-broadband DSL and cable modem connections. Can you say instantaneous data transfer? Can you say, video on demand? SBC and Bellsouth are two of the telecom giants pioneering FTTH. The initial markets are new residential construction, because you don't have to dig up streets in an existing neighborhood to lay the fiber optic cable. SBC plans to wire 6,000 homes in a community in San Francisco by late next year. Initial net connections will only be about 5 MB/second -- far from the theoretical maximum of 155 MB/second, but still blazingly fast compared to DSL and cable modem. BellSouth is also pioneering FTTH with a trial project involving more than 400 people in the Atlanta suburb of Dunwoody. These individuals have Internet connections of about 10 MB/second! No doubt there will be stumbles along the way to providing FTTH. No doubt there are challenges to making FTTH cost effective. No doubt it will take years before most residences in the world have true broadband Internet access.

To cater for the burst characteristic of IP traffic, WDM networks are expected to reconfigure the virtual topologies corresponding to the real-time changes of the traffic flow. This paper introduces the elementary algorithms of the virtual topology reconfiguration. Compared to each other, we point out their advantages and disadvantages. Then our paper focuses on the tradeoff between the benefits and costs brought by the reconfiguration of the virtual topologies. We present a evaluation function and two thresholds, namely a lower and a higher. Our presented approach balances between the benefits and costs due to the reconfiguration of the virtual topology, increasing the benefits as much as possible while guaranteeing the low costs, with a distinguished characteristic of taking the state of the network into consideration.

To provide visual communication over enterprise Intranet, the video conference system in H.323 has been proposed as a suitable architecture to take the place of circuit-switched telephony model. However, managing video conference system will be complicated due to the real-time monitoring and reporting. This paper presents some research on the network management of H.323 Video conference system, and introduces the standards about this system, such as ITU-T H.341 and H.350 recommendation, and then gives some advices on network management design for video conference system with the considering of the real-time feature.

The QoS measurement and monitoring framework proposed for Diff-Serv consists of three conceptual modules, the QoS data acquisition agent, the Qos data correlating agent and the QoS data analysis agent. The Common Object Request Broker Architecture (CORBA) is used when implementing the frame. The constructing of the measurement package, the transmitting rule of active measurement packages and path measurement is studied. A abbreviated QoS data correlating and analyzing is introduced.

This paper discusses the specific realization of SDH protection which based on ITU-T recommendation G.841, and gives detailed flow of APS protocol for some representation status of MS shared protection rings (not defined in G.841), and reinforces the content of G.841.

Most schemes of quantum key remote distribution available mainly deal with the quantum repeater based on quantum entanglement swapping. However, as the quantum repeater has many insurmountable shortcomings, this paper, relying on the existing fairly mature technique, proposes a highly efficient QKD scheme with authentication mechanism from a new perspective of scheme not based on entanglement swapping, overcoming the low efficiency of the entanglement swapping scheme. By means of the quantum authentication mechanism, this scheme is capable of preventing and detecting any active attacks from any point in the quantum network. The scheme proposed has realized authentication and key distribution in the quantum channel at the same time, providing the authentication mechanism of this type of remote key distribution network with a reference. This paper first briefly analyzes the shortcomings of the quantum repeater scheme based on entanglement swapping and introduces the idea of highly efficient quantum key negotiation by means of the pseudo random sequence. Then a key distribution with authentication protocol based on shared secret is presented followed by a description of the application of this scheme in the quantum key distribution network.

The rapid development of the data service has brought great challenge to the traditional telecommunication network. The integrated broadband MAN with large information capability, high subscriber density and multiple sorts of services, is becoming the hot point and competitive focus of the telecommunication field. Resilient packet ring (RPR) is a new metropolitan area network (MAN) technology supporting data transfer among stations interconnected in a dual-ring configuration. Key features of RPR include: Low-delay guaranteed rate, bounded delay committed rate, and best effort service classes; 50 millisecond service restoration following a single station or link failure; Ring capacity reuse downstream of unicast destination (spatial reuse); Weighted fair access to available ring capacity; Point-of-congestion aware flow control (allowing per destination queuing in client). But RPR is belong to layer 2 technology. SDH have already become the dominant optical transport network technology, and it is considered to be one of foundation for the physical layer of the broadband IP network. But SDH is taken as a voice-fit technology. This paper fully make use of RPR & SDH advantages to provide The integrated broadband MAN solution, this is the RPR over MSTP System. The architecture of RPR over MSTP and key technologies of realization of RPR over MSTP node is provided. The main technical specification of RPR over MSTP node is also provided.

This paper analyses the characteristics and architecture of EPON system. We probe two management methods. Based on EPON management resources, information models could be defined in dependent of deployment technology. In the end we discuss functional architecture of EPON management system.

In this paper, an analysis of networking and the possible applications of the automatically switched optical network (ASON) in the telecommunication network are provided. The following contents are included: First, several networking applications of ASON network is studied, and differences between them are compared; Second, the possible applications of these models in the telecommunication network are analyzed. Finally, how to evolution to ASON and the incremental services provided by ASON is discussed.

The need to provide QoS-guaranteed services in the WDM optical networks is becoming increasingly important because of a variety of candidate client networks (e.g., IP, ATM, SONET/SDH) and the requirement for QoS-delivery within the transport layers. This article addresses the QoS problem and presents a framework of QoS provisioning in the WDM optical network. We first characterize the QoS problem in the WDM optical network by comparing with that in the traditional networks. Then we propose a QoS service model in the optical domain called extended differentiated optical services (E-DoS) model based on a set of optical parameters that captures the quality, the reliability and the priority of an optical connection. Each component of the E-DoS model has been analyzed in detail in this article.

This paper addresses QoS provisioning mechanisms in the WDM optical networks. With the appearance of metropolitan optical network, a hierarchical metro and wide area optical network will be envisioned in the near future. This hierarchical optical transport network is often divided into optical domains by geography, administration and technology, which usually employ different QoS routing algorithms and policies. To provide end-to-end optical QoS is becoming a new challenge for the optical network design. In this paper, we first give an overview of issues on the QoS provisioning in data, control and management planes of the WDM optical network. And then three provisioning approaches are analyzed and compared. Finally, we propose an agent-based hybrid centralized/distributed QoS provisioning mechanism based on the concept of domain agent. This agent-based hybrid mechanism employs centralized approach in the domain and distributed approach between domains. It offers scalability and intra-domain optimal QoS routing. It also keeps independence and interoperability between domains.

Group multicasting is a kind of communication mechanism whereby each member of a group sends messages to all the other members of the same group. Group multicast routing algorithms capable of satisfying quality of service (QoS) requirements of multimedia applications are essential for high-speed networks. We present a heuristic algorithm for group multicast routing with end to end delay constraint. Source-specific routing trees for each member are generated in our algorithm, which satisfy member’s bandwidth and end to end delay requirements. Simulations over random network were carried out to compare proposed algorithm performance with Low and Song’s. The experimental results show that our proposed algorithm performs better in terms of network cost and ability in constructing feasible multicast trees for group members. Moreover, our algorithm achieves good performance in balancing traffic, which can avoid link blocking and enhance the network behavior efficiently.

Recently, there is an increasing interests in providing differentiated services in Internet. However, research efforts have almost exclusively focused on routers by improving their policies of packet scheduling and queue management. There has been much less work on transport level approaches to support differentiated services. The mechanism presented by Chang-Biao Xu, DSAS-TCP and MulTCP are the only pieces of the works in this direction known to the authors. Up to now, there is no paper to discuss the interrelation between these mechanisms. Regarding throughput as TCP criteria to support proportional-differentiated-services (PDS), this paper deeply explores the variants of AIMD(a,b)-based TCP congestion control and their effect on differentiated services, and presents a transport level approach for TCP to support PDS, namely PDS_TCP which can be obtained by introducing weighted factor to a or b of AIMD(a,b)-based TCP congestion control. PDS_TCP also takes into account the influence of slow start for timeout. From the analysis, this paper draws the conclusion that the existing mechanisms are only variants of PDS_TCP. For the example of PDS_TCP, the principles, implementation and simulation results of PDS_a_TCP are discussed in detail. The theory analysis and simulation results show that the proposed mechanism PDS_TCP can be implemented with lower additional overheads and support controlled PDS very well without the loss of flexibility.

A Differentiated Service (Diffserv) model is proposed to support Differentiated Services end-to-end and several function terms are explained in detail in this paper. In order to implement Diffserv for Optical Burst Switched (OBS) networks, we present a kind of Diffserv code-point in control packet. A series of Diffserv rules are studied in detail, including the overwriting and update of Diffserv code-point and the scalability of Diffserv model. Also several issues which need to be improved are proposed.

For the integrated optical Internet model, we designed a novel dynamic integrated routing algorithm with Quality of Service and fairness guarantee that employs adaptive unified routing strategies according to the priority and granularity of the traffic connection request. Performance simulation results of this algorithm showed that our adaptive approach outperforms in terms of network throughput performance than traditional overlay sequential routing scheme such as optical-layer-first routing (OFR) and electrical-layer-first routing (EFR) schemes. This algorithm gets the better trade-off of QoS and fairness of routing for integrated IP/WDM networks.

Constraint-based routing algorithms are a key component to meet various requirements subjects to a set of constraints for bandwidth requests in the GMPLS-based optical networks. The purpose of GMPLS control plane is to provide an intelligent automatic end-to-end label switched path provisioning/signaling framework to support arbitrary bandwidth traffic. We developed a detailed simulation platform for an IP-over-MPLS-over-WDM network, so that extensive simulations are carried out to evaluate the network performance of various routing algorithms and routing strategies with RSVP-TE or CR-LDP signaling protocol in GMPLS-based IP/MPLS optical networks in terms of blocking probability. The simulation results show that the signaling protocol improves the network performance more significantly than multi-fiber configuration and CR-LDP signaling mechanism outperforms RSVP-TE in all cases. Furthermore, the LCF routing strategy also has better performance than LCL strategy.

A MAC scheme of HORNET based on PRI and node structure with variable optical buffer is reported. When collision is detected in an AP, PRI of this coming packet is compared with that of the packet being transmitted, if priority of the former is low, it will be routed into a variable optical buffer to loop for the latter. If its priority is high, the buffer will be closed, and the coming packet will be transmitted undisturbedly. On the other hand, the packet being transmitted is terminated, and the transmitter sends out a jamming signal. The downstream AP recognizes the incomplete packet by the presence of the jamming signal and pulls it off the ring. The AP can reschedule the transmission of the packet for a later time. Thus it can be ensured that the packet with high PRI will be transmitted first when conflict emerges.

The progress of the mesh networks technology makes optical networks more complex. An adverse result is the increased network vulnerability. It’s difficult for current methods of centralized restoration to deal with the problem mentioned above. A novel survival scheme, virtual nodes scheme, is proposed in this paper. The character of network topology is considered as an important factor in this scheme. The major idea of this scheme is dividing network into some subnets, which are named as virtual nodes, and different survival approaches are adopted within each virtual node and among virtual nodes. Each virtual node not only can dispose the interior failures independently, but also can settle faults collaborating with other virtual nodes. Compared with current survival schemes, which regard each node or link as a manageable object, virtual nodes scheme take each virtual node as a manageable object. Virtual nodes scheme can find failures and restore rapidly. In addition, utilization of capacity is optimized within each virtual node. Therefore, total capacity of network is distributed efficiently. Realization of virtual nodes scheme is explained in this paper. Through theoretical analysis and computer simulation, the performance of this new scheme is demonstrated.

This paper mainly shows that to utilize the characteristics of wireless optical transmission, Space-Frequency(SF) addressing mode mating with Optical Weaving(OW) topology and route planning by employing Hybrid Methods (HM) make wireless optical network provide unique services with high Quality of Service(QoS) which can not be offered by the other means of communication

In this paper, a traffic-grooming problem for multi-granularity traffic of SDH/SONET in WDM grooming mesh networks is investigated. We propose a path select routing algorithm to solve this problem. The performances of this traffic grooming path select routing algorithm are evaluated in WDM grooming networks. Finally, we presented and compared the simulation results of this algorithm in dynamic traffic grooming WDM mesh networks with that of other algorithms.

In this paper we investigate the fairness problem of offset-time-based quality of service (QoS) scheme proposed by Qiao and Dixit in optical burst switching (OBS) networks. In the proposed schemes, QoS relies on the fact that the requests for reservation further into the future, but for practical, benchmark offset-time of data bursts at the intermediate nodes is not equal to each other. Here, a new offset-time-based QoS scheme is introduced, where data bursts are classified according to their offset-time and isolated in the wavelength domain or time domain to achieve the parallel reservation. Through simulation, it is found that this scheme achieves fairness among data bursts with different priority.

In this paper, the objective was optimizing the port number with dynamic traffic grooming of SDH/SONET WDM mesh networks to give useful referenced data to networks design and the cost control of networks. The performances of different path select routing algorithms were evaluated in WDM grooming networks by considering traffic of different bandwidth requests. Finally, the results were presented and compared with in distributed-controlled WDM mesh networks.

One of challenges to efficient utilizing network resources is to design an effective lightpath teardown strategy in IP/MPLS optical networks. In the traditional integrated routing schemes, a lightpath is established optimally for a LSP request, but may not remain optimal for future requests. Releasing the suboptimal lightpath may free wavelength resources to prepare for new optimal lightpath establishment. In this paper, we proposed and evaluated three lightpath path strategies, threshold-based period-based and no lightpath teardown strategies in a comprehensive distributed simulation platform for IP-MPLS over optical networks within the generalized MPLS (GMPLS) framework. The simulation results show that TBLTS and PBLTS achieve shorter-length paths and at low load, and TBLTS outperforms the other two in terms of blocking probability. In addition, CR-LDP also can get significant performance advantages than RSVP-TE signaling protocol.

In this paper, an existing protocol, i.e. TRIP (Telephony Routing over IP) is developed to provide distributed QoS when making resource reservations for VoIP services such as H.323, SIP. Enhanced LSs (location servers) are deployed in ITADs (IP Telephony Administrative Domains) to take in charge of intra-domain routing policy because of small propagation price. It is an easy way to find an IP telephone route for intra-domain VoIP media association and simultaneously possess intra-domain load balancing features. For those routing reservations bridging domains, inter-domain routing policy is responsible for finding the shortest inter-domain route with enough resources. I propose the routing preference policy based on QoS price when the session traffic is shaped by a token bucket, related QoS messages, and message cooperation.

This paper described certain problems when performing QoS active measurement. The term Test Traffic Influence Factor (TTIF), describing quantitatively influence which the test traffic compared to actual traffic on Qos parameters is defined. A kind of ideal model based on the queue theory to study TTIF is build up and a TTIF for delay is discussed. The theoretical analysis results are verified by using network simulation tool-OPNET modeler. Then certain important conclusions and advice about IP network QoS measurement are given and further research direction is directed.

Resilient Packet Ring is a new technology being standardized by IEEE 802.17 working group. This paper describes the ways to accomplish uniform traffic handling and spatial traffic handling by comparison. Since uniform shaping cannot avoid head of line blocking or decrease the jitter of bandwidth utilization on multiple chock point ringlet, this paper proposes a solution for shaping of spatial traffic handling. Virtual destination queues are introduced to the shaper to avoid HOL, and a special kind of dynamic priority is added to the scheduler to optimize bandwidth utilization. A simulation model is set up to exam the performance of VDQ aware shaper, which is the first phase to accomplish spatial shaping.

The Resilient Packet Ring (RPR) IEEE 802.17 standard is under development as a new high-speed backbone technology for metropolitan area networks (MAN) [1]. Bandwidth reclamation has been concerned in RPR specifications from draft 0.1 to draft 2.4. According to specifications, allocated bandwidth can be reused, or reclaimed, by a lower priority service class whenever the reclamation does not effect the service guarantees of any equal or higher priority classes on the local station or on any other station on the ring [2]. The class-based queuing (CBQ) algorithm is proposed to implement link-sharing [3]. A hierarchical link-sharing structure can be used to specify guidelines for the distribution of 'excess’ bandwidth [4] and it can rate-limit all classes to their allocated bandwidth. There is some sameness between the link-sharing of CBQ and bandwidth reclamation of RPR. The CBQ is a mature technology while RPR is a new technology. Given CBQ improvement and full use so as to make its thought suitable for bandwidth reclamation of RPR is the focus of our work. In this paper, we present the solution that can solve the reclamation problem, which proves to be effective by simulation.

Combining asynchronous transfer mode (ATM) over a passive optical network (APON) can provide broadband services as defined by the international telecommunications union (ITU). The medium access control (MAC) layer is of primary importance to the access scheme as in controls the flow of traffic in the access network. This paper presents a novel MAC protocol-segmented proportional spacing MAC protocol, which complies with ITU-T recommendations, is firstly designed for APON system based on the analysis of different type of bandwidth allocation algorithms. The main idea of protocol is: frame structure adopts the structure regulated by ITU; fine time division for the optical network unit (ONU) to apply bandwidth; the bandwidth’s application is not based on the T-interface but ONU, the bandwidth allocation algorithm uses segmented proportional spacing algorithm. At last, we compare our protocol to other MAC protocols, the results show that proportional spacing and segmented bandwidth allocation control the cell jitter with satisfactory and improve the system bandwidth efficiency at same time, the correlative conclusions are given finally.

This paper studies the QoS of VPN in an environment where the public network prices connection-oriented services based on source, destination and grade of service, and advertises these prices to its VPN customers (users). As different QoS technologies can produce different QoS, there are according different traffic classification rules and priority rules. The internet service provider (ISP) may need to build complex mechanisms separately for each node. In order to reduce the burden of network configuration, we need to design policy control technologies. We considers mainly directory server, policy server, policy manager and policy enforcers. Policy decision point (PDP) decide its control according to policy rules. In network, policy enforce point (PEP) decide its network controlled unit. For InterServ and DiffServ, we will adopt different policy control methods as following: (1) In InterServ, traffic uses resource reservation protocol (RSVP) to guarantee the network resource. (2) In DiffServ, policy server controls the DiffServ code points and per hop behavior (PHB), its PDP distributes information to each network node. Policy server will function as following: information searching; decision mechanism; decision delivering; auto-configuration. In order to prove the effectiveness of QoS policy control, we make the corrective simulation.

Resilient Packet Ring (RPR) is a newly developed Layer 2 access technology for ring topology based high speed network. Fairness Algorithm (FA), one of its key technologies, takes responsibility for regulating each station access to the ring. Since different methods emphasize particularly on different aspects, the RPR Work Group have tabled several proposals. This paper will discuss two of them and propose an improved algorithm, which can be seen as a generalization of the two schemes proposed in [1] and [2]. The new algorithm is a distributed algorithm, and uses a multi level feedback mechanism. Each station calculates its own fair rate to regulate its access to the ring, and sends fairness control message (FCM) with its bandwidth demand information to the whole ring. All stations keep a bandwidth demand image, which update periodically based on the information of received FCM. The image can be used for local fair rate calculation to achieve fair access. In the properties study section of this paper, we compare our algorithm with the two existing one both in theoretical method and in scenario simulation. Our algorithm has successfully resolve lack of the awareness of multi congestion points in [1] and the drawback of weakness of fault tolerance in [2].

The Gigabit Ethernet technology has already been maturity. Compared with other technology, Gigabit Ethernet technology possesses higher performance and more convenience realization, so in volume holographic storage (VHM) system, we choose Gigabit Ethernet as channel interface of volume holographic storage.

Even among the global telecommunications boom over the past decade starting early 1990s, the phenomenal rate of growth in the telecommunications sector in China is no less than a miracle. Not only that China now has the largest fixed line telecommunications network and the biggest mobile telephone market in the world, but China has also become a global power in telecommunications equipment supply and a serious contender in technology provision. This paper gives an overview on the telecommunications sector in China, with particular concerns on the policy and regulatory regime, the market structure and competition landscape, R&D activities and technical progress, and educational systems and human resource development. It is hoped that the explanation of the development process can serve as an indication of the future development of China’s telecommunications sector in a more open and globalized economy.

China Education and Research Network (CERNET) is the largest academic Internet backbone in China and one of the largest in the world. This paper will introduce the current CERNET infrastructure as well as the CERNET’s next generation plan. CERNET has more than 20,000KM dark fibers and operates its own DWDM transport network. Currently it is a 40Gbps system, providing 2.5Gbps router-to-router connectivity for the 8 CERNET GigaPops and reaches all the provincial capitals via a 155Mbps-2.5Gbps backbone. It connects more than 1200 universities and research institutions. Most connected university campus networks are 1Gbps based. The number of end users is about 12 Million. The traffic statistics will be analyzed in this paper and the future bandwidth demand will be estimated. CERNET is also evolving to the next generation Internet. It connects to Internet2 and other next generation academic experimental Internet backbones worldwide. The goals of next generation CERNET are to provide a high-speed backbone network and to offer services to meet the demands of advanced research applications. The challenges are to provide scalable, high speed, high quality network services to all the universities and schools in China and their total 320M students and pupils. The current and future plans of CERNET will also be discussed in this paper, including the nationwide DWDM system, the CWDM MAN and campus network, the unified network management system, the lambda on demand service, the IPv6 service, the multicast service, the wireless services/large scale roaming service and large scale video service.