This PDF file contains the front matter associated with SPIE Proceedings Volume 7137, including the Title Page, Copyright information, Table of Contents, Introduction, and the Conference Committee listing.

Future access network architectures should exhibit reliability while providing broadband Internet services. Progress on research and development of reliable wired and wireless access networks is discussed in this paper. Some emerging access network architectures and their reliability issues are explained.

The hybrid optical wireless network has gained interest for future access networks. In order to guarantee quality of service (QoS), especially for the multi-media services, resource management and scheduling mechanism become critical design issues in the hybrid optical wireless network. Various scheduling algorithms have been proposed in either optical (EPON) or wireless (WiMAX) domains. However, none of these algorithms has considered the end-to-end QoS for a connection communicating through both two domains. In this paper, we come up with an idea to modify the original signalling protocol in the optical domain, and use it to design a novel Integrated Optical Wireless Admission Control (IOW-AC) in an uplink scheduler. From both analysis and simulation, QoS requirement in terms of delay is guaranteed and the utilization of bandwidth resources is improved.

This paper proposed a scalable architecture of IPv4-IPv6 internetworking gateway based on EZchip 10Gbps network processor NP-1c. The Application Layer Gateway(ALG) of control plane can be upgraded without needing to modify the data forwarding plane.A SIP ALG of 3GPP IMS(IP Multimedia Subsystem)was implemented and tested under real China Next Generation Internet(CNGI) network environment.IPv4 SIP UEs can communicate with IPv6 SIP UEs through the gateway.

This paper goal is to provide a framework for the remote configuration and management of services for PON (Passive Optical Network) access and fiber access. Also it defines how Auto-Configuration Servers (ACS) in the network can remotely configure, troubleshoot and manage a Passive Optical Network (PON) optical network termination (ONT) with layer 3 capabilities using the CPE WAN management protocol, TR-069.

This paper overviews the recent transport network development using reconfigurable optical add drop multiplexers (ROADMs) and 40G transmission technologies. High-speed transmission technologies with the line-rate of 40G or 100G are reviewed first. Then, the advanced modulation format suitable for constructing the metro-core 40G ROADM network from the viewpoints of impairments is discussed. Finally the advanced functionalities that enhance ROADM network agility are described.

In this paper, PCE-based inter-domain path computation has been discussed. Interworking of multiple PCEs with the BRPC procedure enables not only end-to-end shortest path computation but also end-to-end diverse path computation over multiple networks. We have analyzed various PCE-based inter-domain path computation strategies using BRPC, in terms of the path computation complexity. To reduce path computation time, a simplified path computation strategy was proposed and evaluated experimentally with our prototyped PCE system. Our experimental results show good correspondence with analytical ones. The proposed strategy drastically reduced path computation time down to one-seventh in comparison with the conventional strategy.

We evaluate the connection provisioning performance of GMPLS-controlled wavelength routed networks under dynamic traffic load and using three different wavelength converter placement heuristics. Results show that a simple uniform placement heuristic matches the performance of complex heuristics under dynamic traffic assumptions.

Passive Optical Network (PON) system with characters of low-cost and high-efficiency has become a promising solution for access network in relieving the bandwidth bottleneck problem. PON system can provide high-speed optical subscriber networks with broad bandwidth and long distance transmission. In current telecommunication carriers' PON system, different service has its given priority which is configured by carriers based on common requirements. However, different users have their own requirement and inflexibility of the given default service priority causes users' inconvenience and frustration. In this paper, we propose a novel method of user controlled Port-IDentifier (Port-ID) assignment in a Gigabit-capable PON (GPON) system. This method provides users a more individualized priority determination mechanism for end users through assigning a multi-level Port-ID to different service. Then users can manage their own bandwidth based on their requirements for different service and carriers can provide a flexible service management mechanism.

In this work, we use a low-pass filter in the burst-mode receiver between PIN+TIA and LA to optimize the receive bandwidth, and finally to improve the receiver performance of uncooled Reflective-SOA-based WDM-PON significantly. In addition, an extra power budget is earned for further development.

Capacity allocation(CA) problem modelled by a new method for time-varying traffic is formulated and solved. An upper bound of the capacity saving is given by theoretical analysis. The numerical results show that approximately 25% of the total required capacity can be saved which corresponds to the theoretical analysis.

An intra-fibre wavelength conversion (IFWC) adopting ROADM-like structure for upgrading MG-OXC router is proposed. A minimum resource first assignment (MRFA) algorithm aimed at this specific structure is also developed for better performance.

This paper describes summary and recent achievements of Lambda Utility Project, which is one of the major Photonic Network R&D Projects supported by the Ministry of Internal Affairs and Communications (MIC) and National Institute of Information and Communications Technology (NICT) Japan.

This paper investigates the self-adaptive transmission control architectures (SATCA) in impairment-aware transparent WDM networks. Two approaches are proposed for the implementation of SATCA. Simulations are conducted to evaluate the benefits obtained by applying SATCA. In the transparent WDM networks, the control plane based on GMPLS protocols introduces connection intelligence into the optical networks. However, since dynamic setup/teardown of connection or fast re-route may introduce unpredictable physical impairments(i.e., fluctuation of optical power and residual dispersion) into the lightpath and in all-optical networks these physical impairments will accumulate along the lightpath, the optical signal quality and service survivablity can not be guaranteed. Therefore, the future optical networks should also have self-adaptive optical transmission ability to guarantee the physical quality of connections. Two proposed approaches to implement SATCA separately introduce extensions to routing protocols (routing-based approach) and signalling protocols (signaling-based approach) of GMPLS. For both approaches, the lightpath quality estimation (LQE) module, which is applied to evaluate the lightpath's QOT and make the compensation budgets, should be added to the optical control plane. The purpose of the simulations is to research the effect of SATCA approaches on the lightpath performance. We suppose that if OSNR or RD is out of the acceptable range, the lightpath is refused. So the blocking probability can be used for comparison among the signaling-based approach,routing-based approach and the traditional approach which is without considering physical impairments. Since the lightpath performance can be improved in the SATCA approach, the quality of lightpath is greatly guaranteed.

Multi-granularity optical cross-connects (MG-OXC) is proposed as a promising technique to reduce the cost of the optical nodes in the wavelength division multiplexing (WDM) optical networks. In this paper, we use the on-line Maximize the Number of Waveband Route (MNWR) algorithm to evaluate the performance and optimize the design of multi-granularity optical WDM networks based on two different MG-OXCs: the serial MG-OXCs and the interconnecting MG-OXCs. The numerical results show that our optimal design can significantly reduce the network complexity and the blocking probability.

In this paper, we propose and evaluate a novel joint scheduling algorithm for multiple services which can support differentiated services and bandwidth allocation with global priority of different traffic such as Triple-play: voice, video and Internet services. Simulation results show that for the higher priority traffic better delay and jitter performance can be achieved compared with the lower priority traffic.

This paper studies the problem of the implementation of distributed computing applications in local area networks. We propose a novel Dynamical Wavelength Scheduled Wavelength Division Multiplexing Passive Optical Network (WDMPON) architecture, in which the number of the available upstream wavelength channels is greatly less than that of optical network units (ONU). And we experimentally demonstrate the feasibility of the proposed WDM-PON, which enables dynamically scheduling upstream data in the time division multiplexing (TDM) and WDM scheme from ONU to optical line terminal (OLT). The Time-Wavelength Co-Allocation (TWCA) Problem is defined in the proposed WDMPON architecture to aggregate large files for distributed computing applications and three scheduling algorithms are presented to solve this problem. The significant improvement compared with the conventional TDM-over-WDM PON is illustrated through simulations.

A Postponement strategy for short lightpath teardown in wavelength-routed WDM mesh networks under dynamic traffic is proposed for the first time. This strategy postpones the teardown process of the short lightpaths when there is no traffic on them. In this way, the average hop number of lightpaths will be reduced and thus more lightpaths may be set up to reduce the overall blocking probability. Moreover, since the proposed strategy extends the average holding time of lightpaths, it can help reduce the signaling flow in the optical control plane. Continuous-time system (CTS) simulation shows that the strategy is capable of reducing the overall blocking probability and the signaling flow in the optical control plane.

All-optical virtual private network (VPN), which offers dedicated optical channels to connect users within a VPN group, is considered a promising approach to efficient internetworking with low latency and enhanced security implemented in the physical layer. On the other hand, time-division multiplexed (TDM) / wavelength-division multiplexed (WDM) network architecture based on a feeder-ring with access-tree topology, is considered a pragmatic migration scenario from current TDM-PONs to future WDM-PONs and a potential convergence scheme for access and metropolitan networks, due to its efficiently shared hardware and bandwidth resources. All-optical VPN internetworking in such a metro-access integrated structure is expected to cover a wider service area and therefore is highly desirable. In this paper, we present a TDM/WDM metro-access integrated network supporting all-optical VPN internetworking among ONUs in different sub- PONs based on orthogonal differential-phase-shift keying (DPSK) / amplitude-shift keying (ASK) modulation format. In each ONU, no laser but a single Mach-Zehnder modulator (MZM) is needed for the upstream and VPN signal generation, which is cost-effective. Experiments and simulations are performed to verify its feasibility as a potential solution to the future access service.

In this paper, two novel protection and restoration schemes were proposed by using the BC and BR schemes which were originally designed to reduce the bursts loss probability due to contention in OBS network. The 1+1 protection scheme in LOBS can be easily carried out using the MPLS 1+1 protection scheme in the 2.5 layer and burst cloning scheme in the OBS layer. By introducing some new OSPF-TE opaque LSA messages, the burst retransmission scheme can be transplanted into the restoration scheme in LOBS network. Furthermore, the dynamic quality of resilience mechanism to inherit these two schemes according to the QoS requirement of the traffic was also investigated. Experimental study using a LOBS network test bed verified that the proposed schemes suited well in the LOBS network.

We propose and experimentally demonstrate a novel coding scheme that can eliminate extinction ratio limitation in a WDM-PON system using RZ-ASK/DPSK re-modulation format corresponding to downstream/ upstream signals. Rayleigh and stimulated Brillouin backscattering effects are simulated and analyzed in the deployment of single fiber bidirectional transmission probably using in practical application.

Under the current and foreseeable limitations of optical technology, performance of the next generation optical network is mainly hampered at the network node by resource contention. Existing work addresses contention resolution mostly in the wavelength domain (as wavelength conversion, WC), the time domain (buffering by using fiber delay lines, FDLs, the space domain (as deflection routing), and the burst domain (as segmentation). In this paper, a new architecture of optical networks-the server view based on optical networks is proposed. From the point of this new architecture, the network can be modeled as a server system with three type servers-the access server, the node server and the link server. The network performances such as cost, energy consume and network capacity can be affected by the capability of these three type servers. Unlike existing research activities that are focused on the node server design, we present the Dual fiber link (the link server solution) and IP calking (the access server solution) to ease the resource contention. To evaluate the performance of the proposed methods, computer simulations are conducted. The simulation results show these solutions can reduce the contention effectively with existing commercial devices.

Based on the service plane (SP) of Adaptive Multi-Service Optical Network (AMSON), this article designs and implements optical service level agreement (O-SLA),which meets the clients' requirements and improves the performance of the optical network. Our Adaptive Multi-Service Optical Network (AMSON) testbed is based on the framework of Automatic Switched Optical Network (ASON), and introduces a service plane (SP) into it as an independent functional plane. SP receives and processes the service requests submitted by the clients. An O-SLA is a formal contract between an optical network service provider and a service subscriber that contains various service levels and diverse clients' requirements. We focuses on designing and implementing O-SLA on SP of AMSON by introducing O-SLA to Bandwidth on demand (BoD). BoD is a kind of service initialized by clients in SP and can be adjusted dynamically. We classify three levels for the contents of O-SLA that enable the clients to get three kinds of authorization. To conduct this O-SLA, the network operators should publish relevant information to SP. According to this information and O-SLA, the related modules of SP can lay out proper instructions to CP. Then appropriate operation will be done by CP and the results of it will be returned to SP. Some simulations have been done and results are given, which illustrate that the implementation of O-SLA combined with BoD on SP of AMSON induces the blocking rate and improves the utilization of the network resource.

Service plane has been proposed to meet the increasing complexity in the NGON. In this paper, an in-depth research on self-adaptation service plane extension in NGON is made. Firstly, through the analysis of network developing trend and future service characteristics, the principle of introducing service plane is cleared. Then the self-adaptation technology is explained for the purpose of adapting to the development of optical network. Subsequently through the description of service self-adaptation capacity, the self-adaptation framework of service-oriented NGON is built. After that, according to the extraction of service plane functional requirements, service plane functional elements are built and the flow of access services is illustrated. Through this paper, a reference is provided for more deeply research of service optical network.

We derive a set of spread spectrum sequences from chaotic dynamical system to support more simultaneous users for OCDMA-WDM-PON system, named Two-level chaotic Logistic-map sequences. The BER performance of the proposed scheme in comparison to conventional methods is investigated.

The wavelength-division multiplexing passive optical network (WDM-PON) is considered to be a promising method to overlay or upgrade the existing optical access network. A centralized WDM-PON in which the downstream wavelength is reused to modulate the upstream has been studied comprehensively, for same wavelength operation of both downstream and upstream, colorless operation and source-free of optical network units (ONU) can be realized in such systems. We propose a scheme of centralized WDM-PON with non-return-to-zero/continuous-wave (NRZ/CW) polarization multiplexed downstream and remodulated upstream using a CW injection-locked Fabry-Perot (FP) laser. This scheme can eliminate the crosstalk of downstream data on the upstream and get higher extinction ratio of received downstream signal. Experiment with 22.2km transmission of 10-Gb/s downstream signal and 622-Mb/s upstream signal was realized using a commercial FP laser, with power penalties of 0.2dB and 0dB respectively. 2.5-Gb/s upstream signal using an injection locked FP laser was also investigated, and 61 modes within 25.4nm span were locked with side-mode suppression ratio of above 27dB.

T-MPLS is a connection-oriented packet switched solution for transport networks. In this paper, we investigate the signaling mechanism in T-MPLS control plane with BRR-N and SRR strategies and simulate it in OPNET modeler. The standardization process is pushed by ITU-T and IETF. T-MPLS is much easier to achieve a convergent packet platform for any L2 protocol. It focuses on the packet transport application that adheres to the standard groups. With the proposed mechanisms, the performance was deeply studied in this paper for the controlling overhead and the block probability of the transport network. The realization of the T-MPLS network router model and its associated structure is also presented to show the detail functional components. The convergent packet transporting platform for any L2 and L3 protocols is attractive for future telecom applications. Keywords: T-MPLS, backward reservation, forward reservation, packet transport

Abstract: Ethernet is the most important Local Area Network (LAN) technology since more than 90% data traffic in access layer is carried on Ethernet. From 10M to 10G, the improving Ethernet technology can be not only used in LAN, but also a good choice for MAN even WAN. MAN are always constructed in ring topology because the ring network could provide resilient path protection by using less resource (fibre or cable) than other network topologies. In layer 2 data networks, spanning tree protocol (STP) is always used to protect transmit link and preventing the formation of logic loop in networks. However, STP cannot guarantee the efficiency of service convergence when link fault happened. In fact, convergent time of networks with STP is about several minutes. Though Rapid Spanning Tree Protocol (RSTP) and Multi-Spanning Tree Protocol (MSTP) improve the STP technology, they still need a couple of seconds to achieve convergence, and can not provide sub-50ms protection switching. This paper presents a novel rapid ring protection method (RRPM) for carrier Ethernet. Unlike other link-fault detection method, it adopts distributed algorithm to detect link fault rapidly (sub-50ms). When networks restore from link fault, it can revert to the original working state. RRPM can provide single ring protection and interconnected ring protection without the formation of super loop. In normal operation, the master node blocks the secondary port for all non-RRPM Ethernet frames belonging to the given RRPM Ring, thereby avoiding a loop in the ring. When link fault happens, the node on which the failure happens moves from the "ring normal" state to the "ring fault" state. It also sends "link down" frame immediately to other nodes and blocks broken port and flushes its forwarding database. Those who receive "link down" frame will flush forwarding database and master node should unblock its secondary port. When the failure restores, the whole ring will revert to the normal state. That is block secondary port on master node for loop avoidance. Ethernet rings may be interconnected through dual nodes with a shared link. In this case, we should fix that which ring on the share link is ctrl-ring, and only ctrl-ring can deal with the failure happens on the share link. Carrier Ethernet is an alternative for traditional MAN technologies. Ethernet ring network which can make CE more reliable is one of the hottest topics in recently years. RRPM can provide protection switching in sub-50ms and process link fault on single ring and interconnected rings.

Although Gigabit T-MPLS technology has not been widely used for transmission network, it is attracting more and more attention. The design of high-speed data switching is the key technology in T-MPLS. In this paper, one method that implements 4 ports Gigabit T-MPLS switch chip on FPGA which is used for TDM over T-MPLS exchanging is introduced, and the simulation and verification results will be given in the conclusions. It researches the requirements of transport plane in T-MPLS and analyses the architecture, and demonstrates a design of TDM in Transport MPLS based on FPGA through a testbed which is used to explore and implement the concepts of Transport MPLS.

Grid is a resource management technology which provides infrastructures to execute applications. Grid manages resources and provides an infrastructure composed of computing, storage and other resources. While Grid assumes existence of network which connects resources, network used to be considered as a non-manageable resource. However, in these several years, an idea and implementations to use bandwidth of network as a manageable resource have been proposed and demonstrated. Using this kind of schemes together with optical circuit (path) switching technology, a high-performance, energy-efficient and low-cost IT infrastructure can be realized. In this paper, the G-lambda project and its experiments, in which lambdas (optical paths) are reserved and used by requests from applications and middleware, are introduced.

Nowadays scheduling jobs on parallel computers or grid components in grid environment has becoming more and more important. Detailed knowledge of the individual user preference is a necessary factor to be taken into account for developing appropriate scheduling strategies. Other than minimizing makespan [1], workload and the balancing of resource utilization is also an important aspect for scheduling. In the premise of satisfying the preference of individual user, previous approaches mainly focused on minimizing makespan, however workload balancing is rarely optimized. Therefore some of resources may leave underused while other resources are overused. In this paper, a new dynamic scheduling scheme is proposed, using the packing and partial cloning, and the expected QoS can be obtained. This scheme has been further evaluated by simulations and shown to yield good results compared with the former studies.

Optical grids can integrate geographically distributed resources more efficiently to support applications because of characteristics of optical networks. The essential aspect of optical grid applications is scheduling. By scheduling, multiple optical grid applications from different users can be executed on one optical grid. This means multiple users can use one optical grid simultaneously. In this paper, we propose three new algorithms for the scheduling of multiple optical grid applications onto one optical grid, and one algorithm among these three can make the scheduling order dynamic. In these algorithms, communication contention has been incorporated into scheduling, and arrival time of multiple optical grid applications is assumed to abide by Poisson distribution to simulate multiple users. Our objects of scheduling are not only to make scheduling span short, but also to achieve low unfairness. Impacts of different algorithms are comparatively investigated by simulations.

Optical Grid has been thought as a promising technology to support large-scale data-intensive distributed applications. For such a system with so many heterogeneous resources involved, faults seem to be inevitable. A fault in Optical Grid may lead to the failure of the application. Therefore, fault-tolerant policies are necessary to improve the performance of Optical Grid and satisfy the SLA requirements of the applications. In this paper, we propose several fault-tolerant policies for Optical Grid. The performances of different polices are compared by simulation.

Performance of a resource-efficient provisioning framework (RPF) is evaluated in this paper in wavelength-continuous optical networks under static traffic. The structure of the survivable routing problem is unraveled by an integer linear programming (ILP) model. Numerical results from direct ILP approaches show that the RPF still achieves more than 10% capacity savings over the traditional provisioning framework when subject to the wavelength-continuity constraint. Next, Resource-Efficient Two-stEp Routing subject to Wavelength-ConTinuity constraint (RETER-WCT) algorithm is developed to overcome the drawback of the ILP approach.

In WDM optical networks, a low speed traffic is often multiplexed or switched into a high speed lightpaths. In common measures, a lightpath established in traffic grooming does not consider transmission impairment. It transports traffic optically as long as possible. But optical medium is not ideal, transmission impairment like ASE noise, which degrades the signal to noise ratio, exists. To maintain an acceptable SNR level at the receiver side, the number of optical amplifiers in a lightpath should not exceed a Maximum value (M). A segment between two optical amplifiers referred as a span. In this paper, we adds span constraints to the traffic grooming problem, and present an heuristic algorithm to bring up a series of lightpaths explicitly to accommodate such traffic. From the numerical result, it could be found that it guarantees each traffic to be successfully routed. It also achieves good performance of wavelength efficiency and lightpath efficiency.

The traffic grooming based on Adaptive Immune Evolutionary Algorithm (AIEA) is proposed. The process of AIEA for traffic grooming problem was also given simply. Computer simulations and comparison with other heuristic algorithms were performed. The results show AIEA has better resource utilization, good convergence rate and global optimization features in WDM Mesh networks. Simulation results show AIEA has better resource utilization, average jumping delay time and good convergence.

Transparent optical WDM networks have been expected to be the most effective solution to achieve high throughput transmission, in which a GMPLS control plane is used to manage both IP layer and the optical transport layer. However, current GMPLS does not take the influence of physical impairments during data transmission process into account. This paper proposes a signaling-based approach to implement self-adaptive transmission control (SATC) in impairment-aware transparent WDM networks. The function of standard control plane of automatic switched optical network is enhanced to realize this aim. Two evaluation schemes are proposed for the implementation of this approach. Simulations are conducted for proof of benefits gained by introducing SATC.

In this paper, the reliability and MTTF (Mean Time to Failure) for multi-ring RPR (Resilient Packet Ring) are calculated on the conditions of single-link failures, double-link failures and no failure, respectively. The parameters such as the total number of stations N, the number of the sub-rings R, and the distribution of N_i which represents the number of the stations in the i-th sub-ring (1≤i≤R) are contained in the formulas. The relationship between the value of the reliability/MTTF and the parameters N, R and N_i is analyzed. The result shows that reliability/MTTF of the RPR multi-rings is increasing while the variance of N_i is decreasing. It is also proved that the value of the reliability/MTTF is maximum when N_i=N_j ( i ≠j and 1≤i, j≤R) by using Lagrange multipliers method, i.e. the condition of the optimal reliability of multi-ring RPR is satisfied when var(N_i) =0.

This paper focuses on minimizing the capacity of DiffServ/MPLS networks taking into account two widely deployed bandwidth constraint models: the maximum allocation model (MAM) and the Russion doll model (RDM). With the given physical network and traffic demands, by optimizing Label Switched Paths (LSPs) resource allocation, route distribution, our designs aim at minimizing the link capacity of the total network. The optimization problems are formulated as integer linear programming (ILP) models first, and then, since the ILP formulation is not adapted for large-scale network, two novel fast algorithms are proposed. Numerous simulation results for a variety of networks indicate that the performances of our proposed algorithms are close to the optimal results.

IP packet based services such as high speed internet, IP voice and IP video will be widely deployed in telecom network, which make transport network evolution to packet transport network. Characteristics of transport network and requirements of packet transport network are analyzed, T-MPLS/MPLS-TP based PTN technology is given and it will be used in metro (access, aggregation and core) network.

It is because the MSTP based on SDH has developed to be regional multi-service communication network come from the need of LAN-interlink and the new data traffic that the study of encapsulation protocol of MSTP data service over SDH has become especially important. This paper firstly analyzes three kinds of encapsulation protocols which are based on SDH to transport data. The three protocols are GFP, LAPS and PPP, and the study emphases on the encapsulation mode, frame structure, function mode and currency disposal process of the GFP. At last it introduces the GFP's applications in GPON.

In this paper, we propose a novel multi-modular optimal capacity algorithm for SDH (Synchronous Digital Hierarchy) mesh networks. Our objective is to minimize total capital cost of the SDH mesh networks with modularity. First, the optimization problem is formulated as an integer linear programming (ILP) model. Then considering the time complexity of ILP formulation for large-scale networks, a fast heuristic algorithm named Multi-Modular Optimal Capacity (MMOC) is proposed. Several network models, which serve as test cases, are provided and the numerical results show our heuristic algorithm has low time complexity and good performance (in terms of network cost) near to the ILP results.

Much progress has been made in the development of capabilities for enabling rapid provisioning and bandwidth-on-demand services, including at the physical layer, and in the control and management planes. Here we summarize work which examines key factors driving the economies of such services for carriers and their customers.

Self Organizing approaches for Optical Burst Switching (OBS) Networks are suggested at different layers like Application Layer, Transport Layer, Networks Layer and OBS Layer. Self Organization at control plane layer level is also suggested. Detailed methodology of Self Organizing paradigms is explained. Experiments are conducted on OBS Testbed and results are discussed. In conclusion, it is suggested that in OBS networks, layer based approach should be adopted and cross layer design approach should be used for Self Organization to achieve maximum redundancy.

The ever-growing complexity of communication networking has been enhanced by the huge capacity of optical networks, raising the need for new architectures. The WDM layer is now asked to be aware of emerging limitations of the physical layer and QoS requirements of the upper layers. This situation suggests that at least part of the network be designed for self-organization. The paper discusses the merits of swarm intelligence, and especially Ant Colony Optimization, to support a self-organized control plane of an optical network.

The Shannon capacity limit of multimode fibers with asymmetric impulse responses is calculated for a 220 m link with a 10GBASE-LRM standard link power budget. Results show that a 1:2 asymmetric Bi-mode case allows largely bandwidth independent behavior near the maximum Shannon capacity.

Based on Simplified Finite State Machine (SFSM) as an improved model, an approach for passive testing consistency is proposed. The improved model is introduced and the algorithm complexity is analysed. Also, this approach is implemented with the EPON Tester, and proved to be feasible to analyze Multi-Point Control Protocol (MPCP) consistency.

This paper presents an overview on millimeter wave radio over fiber systems based on millimeter wave generation by optical frequency multiplication (OFM). The disadvantage of original OFM technique is checked and a novel bi-directional mm-RoF system based on new OFM technique using dual-drive MZM as high-order optical side-modes generator is proposed in principle and experimentally verified by a bi-directional mm-RoF system which not only produces 38 GHz information bearing signal in downstream, but also provides 40GHz local reference for down-conversion of upstream signal. Finally the advantage of OFM technique in reducing fading effect of fiber dispersion is theoretically proved.

In the traditional EPON network, optical signal from one ONU can not reach other ONUs. So ONUs can not directly transmit packets to other ONUs .The packets must be transferred by the OLT and it consumes both upstream bandwidth and downstream bandwidth. The bandwidth utilization is low and becomes lower when there are more packets among ONUs. When the EPON network carries P2P (Peer-to-Peer) applications and VPN applications, there would be a great lot of packets among ONUs and the traditional EPON network meets the problem of low bandwidth utilization. In the worst situation the bandwidth utilization of traditional EPON only is 50 percent. This paper proposed a novel EPON architecture and a novel medium access control protocol to realize direct packets transmission between ONUs. In the proposed EPON we adopt a novel circled architecture in the splitter. Due to the circled-splitter, optical signals from an ONU can reach the other ONUs and packets could be directly transmitted between two ONUs. The traffic between two ONUs only consumes upstream bandwidth and the bandwidth cost is reduced by 50 percent. Moreover, this kind of directly transmission reduces the packet's latency.

This paper analyses current bandwidth schemes and proposes a novel dynamic bandwidth allocation scheme for EPON. According the scheme, we define four kinds of multimedia services such as Unsolicited Request Service (URS), Realtime Service (rt-S), Non-Real-time Service (nrt-S) and Best Effort (BE). Different kinds of services have different Quality of Service (QoS) requirements. Our scheme considers the diverse QoS request, e.g., delay for rt-S, throughput for nrt-S and fairness for BE. The simulation results show this novel scheme can ensure the quality of service (QoS) and improve bandwidth utilization.

This paper proposes a network anomaly detection method based on catastrophe theory. The network anomaly detection model based on catastrophe theory is constructed by utilizing the catastrophe characters of the cusp catastrophe model in catastrophe theory and analyzing the properties of network data included normal and abnormal data. The simulation demonstrated that with the anomaly detection algorithm, this model could detect the network anomaly not only timely, but also detect precisely.

This paper considers the fault-diagnosis problem for low-degree all-optical networks. Failure-detection schemes which subject to topological constraints are proposed. Optical probe signals are sent in parallel along a set of designed lightpaths. The network health state is derived from the results of this end-to-end test (i.e., probe syndromes). The design objectives are to minimize the diagnosis cost which is represented by the number of probes. We obtain a lower bound of the number of the probes to identify any single link failure and implement this scheme in several networks as well.

GMPLS inter-domain network control was investigated by employing BGP-based inter-domain routing and flexible signaling with loose hop expansion. Per-domain based operational policy was successfully applied to the establishment of GMPLS inter-domain LSPs.

Modeling the traffic demands accurately has important significance for network resource utilization and performance optimization. The uncertainty of demands has made the long term forecasting of traffic a particularly difficult problem. So the traffic model named as sliding scheduled lightpath demands (SSLDs) is considered in this paper. A time-disjointness maximizing algorithm is proposed, which is combined with simulated annealing (SA) algorithm to solve routing and resource assignment problem for SSLDs. This solution is evaluated and compared with two fixed routing algorithms. Simulation results show that SA algorithm has a better performance for routing and resource assignment.

In the recent past, wireless sensor networks have found their way into a wide variety of applications and systems with vastly varying requirements and characteristics. The major limit to the sensor nodes is the limit of energy and computing capability. The major consumption of the nodes energy in sensor networks occurs in the communication between nodes or between the nodes and the base station. Many current protocols only pay attention to how to reduce the energy consumption, rather than the stability during the routing process. Aimed to solve this problem, the paper produces a new kind of reliable and energy-efficient wave routing algorithm. The results of the experimentation can indicate that this new algorithm can not only reduce the energy consumption, but also ensure the stability of the routing.

In this paper, we propose a novel hybrid optical switching scheme called HyLABS (Hybrid Lightpath and Burst Switching), which has two implementing schemes including C-HyLABS and I-HyLABS. Both of the two schemes of HyLABS can control the two optical switching techniques OCS and OBS together, and adopt the former for normal traffic and the later for overloading traffic. Thus HyLABS can efficiently adapt network traffic fluctuations and achieve good performance. Then we evaluate the performance of TCP over HyLABS networks and find an issue called False Fast Retransmit (FFR) that can degrade TCP's throughput. Finally, we propose two algorithms including DRDA and EN-DRDA to address the issue of FFR. Simulation results show that both of the two algorithms can actually improve TCP's throughput performance over HyLABS networks.

In this paper, we propose and describe a new distributed reservation protocol for establishing lightpaths in multi-domain optical networks. Distributed reservation scheme is a promising solution for resource reservation and path computation element (PCE), proposed by IETF and considered as a promising scheme to deal with the issue of path computation in support of multi-domain networks, establishes architecture to solve the reservation problem in multi-domain network. Thus, we propose a novel distributed reservation mechanism (PBDR) for PCE-based multi-domain optical network to ease the inefficiency of the step by step reservation process (Source Initiated Reservation and Destination Initiated Reservation). Also, we evaluate the performance of the proposed PBDR through simulation and the result shows the efficiency and availability of the distributed reservation scheme proposed in multi-domain optical network.

Dynamic DWDM channel blocking and equalizing is proposed based on liquid crystal (LC) and bulk grating technology. A dispersion compensation and enhancement unit is employed so that low center-frequency shift small footprint could be achieved at the same time. The inter-channel crosstalk is also studied. With the optimized design of dispersion unit and LC module, a 32-channel, 100-GHz spacing sample is demonstrated with the insertion loss (IL) of 5.5 dB and 45dB extinction ratio. Application of our device to reconfigurable optical add-and-drop (ROADM) network is discussed

AS businesses become more dependent on IP networks, lots of real-time services are adopted and high availability in networks has become increasingly critical. With the development of carrier grade ethernet, the requirements of high speed metro ethernet device are more urgently. In order to reach the capacity of hundreds of Gbps or Tbps, most of core ethernet switches almost adopted distributed control architecture and large capacity forwarding fabric. When distributed switch works, they always have one CE and many FE. There for, it shows the feature of multicast with one sender and many receivers. It is deserved research for us how to apply reliable multicast to distributed switch inner communication system. In this paper, we present the general architecture of a distributed ethernet switch, focusing on analysis the model of internal communication subsystem. According to its character, a novel reliable multicast communication mechanism based on FEC recovery algorithm has been applied and evaluated in experiment.

One of the main focuses for network planning is on the optimization of network resources required to build a network under certain traffic demand projection. Traditionally, the inputs to this type of network planning problems are treated as deterministic. In reality, the varying traffic requirements and fluctuations in network resources can cause uncertainties in the decision models. The failure to include the uncertainties in the network design process can severely affect the feasibility and economics of the network. Therefore, it is essential to find a solution that can be insensitive to the uncertain conditions during the network planning process. As early as in the 1960's, a network planning problem with varying traffic requirements over time had been studied. Up to now, this kind of network planning problems is still being active researched, especially for the VPN network design. Another kind of network planning problems under uncertainties that has been studied actively in the past decade addresses the fluctuations in network resources. One such hotly pursued research topic is survivable network planning. It considers the design of a network under uncertainties brought by the fluctuations in topology to meet the requirement that the network remains intact up to a certain number of faults occurring anywhere in the network. Recently, the authors proposed a new planning methodology called Generalized Survivable Network that tackles the network design problem under both varying traffic requirements and fluctuations of topology. Although all the above network planning problems handle various kinds of uncertainties, it is hard to find a generic framework under more general uncertainty conditions that allows a more systematic way to solve the problems. With a unified framework, the seemingly diverse models and algorithms can be intimately related and possibly more insights and improvements can be brought out for solving the problem. This motivates us to seek a generic framework for solving the network planning problem under uncertainties. In addition to reviewing the various network planning problems involving uncertainties, we also propose that a unified framework based on robust optimization can be used to solve a rather large segment of network planning problem under uncertainties. Robust optimization is first introduced in the operations research literature and is a framework that incorporates information about the uncertainty sets for the parameters in the optimization model. Even though robust optimization is originated from tackling the uncertainty in the optimization process, it can serve as a comprehensive and suitable framework for tackling generic network planning problems under uncertainties. In this paper, we begin by explaining the main ideas behind the robust optimization approach. Then we demonstrate the capabilities of the proposed framework by giving out some examples of how the robust optimization framework can be applied to the current common network planning problems under uncertain environments. Next, we list some practical considerations for solving the network planning problem under uncertainties with the proposed framework. Finally, we conclude this article with some thoughts on the future directions for applying this framework to solve other network planning problems.

We propose a dynamic path allocation algorithm that considers all paths in the network simultaneously over a given time period. Numerical experiments demonstrate that the algorithm substantially reduces the blocking probability for path establishment.

Call Admission Control (CAC) is widely used in optical communication networks to reduce network congestion. However, the conventional CAC scheme recommended by International Telecommunication Union -Telecommunication Standardization Sector (ITU-T) has a serious deficiency under high traffic load. In this paper, the disadvantage of conventional CAC scheme is analyzed in detail, and a Self-organized Call Admission Control (SCAC) scheme is proposed to solve this disadvantage. This scheme is accord with the principle of self-organization system, so it can be easily implemented in practice. Numerical results show that the proposed scheme can improve the network performance to a great extent.

A novel framework named Service Plane is put forward in transport networks on the basis of the principle of adaptive multi-services provisioning platform. The flexible implementation of service and the expansibility of framework is what we concern most in this kind of Service Plane. Here we raise the core algorithms of bandwidth on demand service on this platform, simulation testbed and the implementation result are also presented in this paper.

To estimate the link loss rates using network tomography, this paper proposes a Fast Path-based Approach (FPA) by explicit estimation. Instead of inferring the link loss rates directly, we first infer the path loss rates from which we can easily derive the link loss rates. In addition, the path loss rates are inferred by a new estimator which is an explicit function of loss observations. The presented estimator is analytical and only requires simple arithmetic calculation. It is also proved to be consistent and have the same asymptotic variance as that of the maximum likelihood estimator (MLE). The simulation results show that the FPA can accurately estimate the link loss rates.

Open shortest path first (OSPF) protocol is commonly used as an interior gateway protocol (IGP) in MPLS and generalized MPLS (GMPLS) networks to determine the topology over which label-switched paths (LSPs) can be established. Traffic-engineering extensions (network states such as link bandwidth information, available wavelengths, signal quality, etc) have been recently enabled in OSPF (henceforth, called OSPF-TE) to support shortest path first (SPF) tree calculation upon different purposes, thus possibly achieving optimal path computation and helping improve resource utilization efficiency. Adding these features into routing phase can exploit the OSPF robustness, and no additional network component is required to manage the traffic-engineering information. However, this traffic-engineering enhancement also complicates OSPF behavior. Since network states change frequently upon the dynamic trafficengineered LSP setup and release, the network is easily driven from a stable state to unstable operating regimes. In this paper, we focus on studying the OSPF-TE stability in terms of convergence time. Convergence time is referred to the time spent by the network to go back to steady states upon any network state change. An external observation method (based on black-box method) is employed to estimate the convergence time. Several experimental test-beds are developed to emulate dynamic LSP setup/release, re-routing upon single-link failure. The experimental results show that with OSPF-TE the network requires more time to converge compared to the conventional OSPF protocol without TE extension. Especially, in case of wavelength-routed optical network (WRON), introducing per wavelength availability and wavelength continuity constraint to OSPF-TE suffers severe convergence time and a large number of advertised link state advertisements (LSAs). Our study implies that long convergence time and large number of LSAs flooded in the network might cause scalability problems in OSPF-TE and impose limitations on OSPF-TE applications. New solutions to mitigate the s convergence time and to reduce the amount of state information are desired in the future.

The interworking network of Generalized Multi-Protocol Label Switching (GMPLS) and Optical Burst Switching (OBS) is attractive network architecture for the future IP/DWDM network nowadays. In this paper, OSPF-TE extensions for multi-domain Optical Burst Switching networks connected by GMPLS controlled WDM network are proposed, the corresponding experimental results such as the advertising latency are also presented by using an OBS network testbed. The experimental results show that it works effectively on the OBS/GMPLS networks.

The future trend is to integrate a constantly increasing number of services in the GMPLS/ASON networks. Some services have very high resilience requirements, while other services have lower ones. This scenario calls for frameworks capable of provisioning for multiple services in a cost efficient manner. This article proposes a differentiated-resilience provisioning scheme applied to the GMPLS/ASON networks, which is expected to be the near- and long-term network technology thanks, among other things, to the great bandwidth capacity offered by optical devices. Finally, a critical evaluation of the state-of-the-art and future challenges facing operators and designers is given. Our numerical results show that the differentiated resilience scenario has better performance than that of dedicated and shared protection and received connections in the differentiated-resilience are 31% higher than that of shared protection and are 60% higher than that of dedicated protection.

Network service support to ensure quality of service (QoS) is a key requirement for many applications, we present a novel management plane oriented service architecture which provides network resource scheduling service in large-scale GMPLS/ASON network environment based on the service level agreement(SLA) conducted between service customers and service providers. It applies the updated service-oriented and policy translation structure with excellent expansibility and efficiency in the running process. The architecture contains four components: Application Monitor Service, SLA Management, Policy management and Service management. In addition, vertical service mapping and differentiated-resilience provisioning schemes of service level agreement(SLA) applied to the GMPLS/ASON networks are discussed, which is expected to be the near- and long-term network technology thanks, among other things, to the great bandwidth capacity offered by optical devices.

Failures in the GMPLS (Generalized Multi-Protocol Label Switching) control plane and in the data plane may independently occur. The established data plane connections such as lightpaths in WDM networks must remain operational in the event of a control plane failure. A control plane failure should not cause changes of the state of connections and resources in the data plane. In addition, the control plane needs to recover the lost state information, or acquire up-to-date state information. The engineering of GMPLS control plane failure recovery techniques must consider different failure scenarios, co-ordinations of different recovery mechanisms. In this paper, we summarize the graceful restart mechanism for the signalling protocols, then present a two-step failure recovery of the signalling protocols to improve the recover speed.

These days while the traditional Internet cannot meet the requirement of data-intensive communications in large scale escience grid applications, Optical network which is also referred to as Lambda Grid provide a simple means of achieving guaranteed high bandwidth, guaranteed latency and deterministic connection. Lots of e-science applications like e-VLBI and GTL require aggregating several hundred GB data files from distributed databases to super-computers frequently at real time. Thus minimizing the aggregation time can improve the overall system performance. We consider the problem of aggregating large data files from distributed databases to distributed computational resources on lambda grid. We modify the model of Time-Path Scheduling Problem (TPSP) which has been proposed and propose a new N-destination TPSP (NDTPSP) model. We present the proof of NDTPSP's NP-completeness. We also propose a list scheduling algorithm and a modified list scheduling algorithm for our problem. The performance of different algorithms will be compared and analyzed by simulations.

Peer-to-peer (P2P) based optical Grid which can utilize the distributed resource discovery mechanism is newly put forward to meet the increasing demand for computing and storage ability. In this paper, an improved P2P-based optical Grid architecture is proposed and demonstrated. Experimental results show that the proposed solution outperforms our previous studies in terms of resource discovery time.

Users and applications need to own greater ability to control network resources for satisfying their requirements and new service pattern. Optical grid nodes include resource, control and switching functional. Optical grid testbed is built to verify node structure and five character of optical grid on ASON network, which involve application-driven, grid service and optical resource uniting, parallel lightpath scheduling, wide resource sharing and multi-granularity bandwidth service. An optical grid node function framework and physical implement scheme is proposed based on distributed network architecture orienting on-demand resource service. Considering the proposed optical grid scheme, corresponding control mechanism is given to satisfy the resource requirement for grid applications and realize distributed resource scheduling on demand.

Divisible loads can be divided into any number independent sub-tasks, and map them on different platform to be processed in parallel jobs. Divisible load theory is introduced into parallel and distributed computing system to obtain available resources distributing on different locations for reaching processing efficiency, which can be extended to distributing multimedia and application system based grid. Light path scheduling algorithm based on DLT is proposed to realize optical resource scheduling on demand in optical grid under the requirement of intensive data applications, especially facing to parallel and distributed system. The proposed algorithm introduces the Divisible Load Theory as load distributing method and is extended for the distributed algorithm of divisible load scheduling to match multichannel application requirement of optical grid network. The proposed method deploys multiple wavelengths for original node, and builds parallel lightpaths to transmit independent divisible loads to collaborating nodes for a big task.

Two distributed service provisioning schemes based on the traditional scheme is proposed to facilitate the collaboration between traditional grid resource and optical networks. A Grid-based Reconfigurable optical network testbed is built to validate and compare those schemes. The experiment results based on the testbed prove that the performance of multi source multi path service provisioning scheme is better the traditional way. With the joint resource allocation, multi paths can be built up, and user can get data resources quickly from separated servers with guaranteed bandwidth from optical networks.

By using novel stabilizing and control technique of bit-wise phase correlation based on 1-bit delay interferometers, we demonstrated stable generation of bitwisely phase-controlled 160-Gb/s signals with different modulation formats such as carrier-suppressed return-to-zero (CS-RZ), pair-alternate-phase CS-RZ (PAP-CSRZ), π/2-altenate-phase RZ (π/2- APRZ), and periodical APRZ. Furthermore, novel performance monitoring technique for 160 Gbit/s OTDM signal using an asynchronous electro-absorption modulator (EAM) prescaler is proposed. This performance monitor can detect waveform distortion caused by chromatic dispersion, polarization mode dispersion (PMD) and optical signal-to-noise ratio (OSNR) degradation separately. And the feasibility of the monitor was examined by an optical test-bed in JGNII network.

New concepts of horizontal protection and half-mesh topology for survivability design in multicast optical virtual private networks are presented. All issues mentioned in this paper are expected to form targets for further investigations.

Future WDM optical networks are foreseen to be more dynamic and sensitive to failures. Therefore, dynamic service provisioning with network survivability becomes a critical requirement for network planning and management. The concept of protected working capacity envelope (PWCE) is attractive because it does not need protection resources to be dynamically configured thereby simplifying network management and reducing processing overheads. In this paper, we consider a p-Cycle-based Strategy for adaptive PWCE (protected working capacity envelope) design where PWCE can adaptively change with dynamically changing traffic. The blocking performance of this new strategy is compared to that of p-Cycle-based PWCE under a dynamic and non-uniform traffic model. The results show that the new strategy can adapt well to dynamic traffic demands that are non-uniform among the network nodes.

The complexity of optical networks has been researched and some complexity laws are described here. Then some metrics from complex networks, such as mean path length, node's degree distribution and node's clustering coefficient have been introduced into optical networks. The invulnerability in complex optical networks is mainly researched in this paper.

Layer 1 VPN (L1VPN) extends the notion of VPN to the optical domain to provide virtually dedicated circuit like leased lines, so that the security is more enhanced. Despite their secure gains from channel isolation, VPNs still suffer fragilities resulting from link-failures or node-failures. Extensive activities on survivability designs for wavelength-routed optical networks are proposed, including various protection and restoration schemes, but concerns on network edge are rare. Dual-homing is an effective skill to achieve survivability gains for L1VPNs. There are two dual-homing mode: Active/Standby mode and Load-Sharing mode In this paper, we investigate the problem of PE assignment, which is the key of dual-homing design and is NP-hard. We formulate it as an integer programming problem, and propose heuristic solutions. Simulation results show that the proposed solutions work in a correct and effective way and the Load-Sharing mode has higher bandwidth efficiency than Active/Standby mode.

In this paper, we propose a heuristic algorithm for solving node inclusion routing problem with additive constraints for the first time. The node inclusion routing problem is significant while a demand need to by-pass several domains and some boundary routers are assigned to be appeared in the route. The integer linear programming model is presented. Two other heuristic algorithms are evaluated. The results show that our heuristic algorithm is more effective with lower trap ratio and very close to the optimal result.

Though analyzing and summarizing the strategies of dynamic multicast traffic grooming, the strategies embody that which optimization criterion would be chose for the requirement of low-speed multicast service to select grooming routing in the grooming network, and the different strategic combination would obtain the different effect of multicast traffic grooming. On this basis, the Light Tree based Integrated Grooming (LTIG) and the Markov Finite Horizon Decision Algorithm of Shortest Path Tree are studied deeply, The research indicates that what the min-cost grooming routing calculation of LTIG adopts is MPH algorithm and Dijikstra algorithm which has higher complexity. But the Markov Finite Horizon Decision Algorithm of Shortest Path Tree has low complexity and it only seeks solutions in General network and isn't applicable to multicast traffic grooming. Subsequently, the Markov Finite Horizon Decision Algorithm of Shortest Path Tree is introduced to LTIG algorithm and a new MTGA-SPT algorithm is proposed. MTGA-SPT algorithm can resolve the route selection problem of resource node to multi-destination nodes, thus forming the light tree. Through the analysis of algorithm complexity, the complexity of MTGA-SPT algorithm mainly depends on the calculation of traffic grooming routing, and adopts the Markov Finite Horizon Decision Algorithm of Shortest Path Tree to compute min-cost grooming routing, making the complexity of MTGA-SPT algorithm descended, The complexity of Markov Finite Horizon Decision Algorithm of Shortest Path Tree is O((V-1)•W), and other step algorithm complexity is about O(V). To sum up, MTGA-SPT algorithm can reduce the time delay effectively.

Bandwidth-on-Demand (BoD) services are characteristic of dynamic bandwidth provisioning based on customers' resource requirement, which will be a must for future networks. BoD services become possible with the development of make-before-break, Virtual Concatenation (VCAT) and Link Capacity Adjustment Scheme (LCAS). In this paper, we introduce BoD services into L1VPN, thus the resource assigned to a L1VPN can be gracefully adjusted at various bandwidth granularities based on customers' requirement. And we propose a dynamic bandwidth adjustment scheme, which is compromise between make-before-break and VCAT&LCAS and mainly based on the latter. The scheme minimizes the number of distinct paths to support a connection between a source-destination pair, and uses make-beforebreak technology for re-optimization.

Two methods of reservation namely ANIR (all nodes-initiated reservation) and MSR (multi-states reservation) are proposed and compared with SIR (source-initiated reservation) and DIR (destination-initiated reservation). The four methods of reservation are simulated in NSFNET and ARPA topologies. Numeric results show that the blocking probability of ANIR is lower than SIR, but the number of packages which it generates is larger, and the blocking probability of MSR is lower than DIR and SIR under light load.

Generalized Multi-Protocol Label Switching (GMPLS) appears as a promising candidate to implement the control plane for next-generation optical networks. In GMPLS-based wavelength division multiplexing (WDM) optical networks, existing resource reservation protocol (RSVP) may fail to setup a multicast Label Switched Path (LSP) with high probability, since multiple destinations in one multicast session may reserve different wavelengths and result in reservation collision at branch nodes. In this work, we extend the existing GMPLS RSVP signaling protocol and propose an efficient scheme to set up optical multicast LSP and avoid wavelength resources collision, called Last-hop Label Set (LLS) scheme, by introducing a new object called Last-hop Label Set object in Resv message. Once the setup of a multicast LSP fails caused by wavelength reservation collision at a branch node, the branch node can re-reserves a new wavelength according to the Last-hop Label Set object within Resv messages. The numerical results indicate that the proposed LLS scheme offers lower blocking probability compared to the existing GMPLS signaling scheme.

In this paper, TCP performance over OBS networks with multiple TCP input flows is evaluated. A new burst assembly method is put forward to improve TCP performance on OBS networks. Normally, TCP packets are assembled by only one assembly queue, increasing the number of the burst assembly queues will affect the TCP performance. Simulation results show that when using time and burst-length threshold, according to burst-length threshold, different number of burst assembly queues should be chosen to get better TCP performance.

OBS is envisioned as a promising infrastructure for the next generation optical network, and TCP is likely to be the dominant transport protocol in the next generation network. Therefore, it is necessary to evaluate the performance of TCP over OBS networks. The assembly at the ingress edge nodes will impact the network performance. There have been several Fixed Assembly Period (FAP) algorithms proposed. However, the assembly period in FAP is fixed, and it can not be adjusted according to the network condition. Moreover, in FAP, the packets from different TCP sources are assembled into one burst. In that case, if such a burst is dropped, the TCP windows of the corresponding sources will shrink and the throughput will be reduced. In this paper, we introduced a flow-oriented Dynamic Assembly Period (DAP) algorithm for TCP over OBS networks. Through comparing the previous and current burst lengths, DAP can track the variation of TCP window, and update the assembly period dynamically for the next assembly. The performance of DAP is evaluated over a single TCP connection and multiple connections, respectively. The simulation results show that DAP performs better than FAP at almost the whole range of burst dropping probability.

We present a novel scheme of high bit rate ASK/FSK orthogonal labelling. The FSK transmitter is based on two external modulated lasers driven by two inverted electrical data. The generated FSK is easy for detection and has a small cross-talk to optical label signal. Simulations and latest experiments of the transmission performance of a 2.5 Gb/s ASK labelled 40 Gb/s FSK signal is also investigated.

Optical burst switching (OBS) is proposed as a high speed, flexible and transparent switching technology. It facilitates the efficient integration of both IP and WDM. And it provides statistical multiplexing gains and avoids long end to end setup time of traditional virtual circuit configuration. However, under the current and foreseeable limitations of optical technology, performance of OBS networks is mainly hampered at the network nodes by resource contention. Due to the fact that the random access memory is not available in the optical domain, there exists a real possibility that bursts may contend with one another at a switching node. Therefore, many contention resolutions are proposed, such as fiber delay lines, wavelength conversion, and deflection routing, etc. These resolutions are focus on the node design. They are increase the complexity of the node. But they can not always take some significant effect and make notable gains in cost-efficiency. In this paper, unlike these schemes are focus on the node design, we proposed a novel link architecture named dual-fiber-link (DFL) for OBS networks. We also proposed a relative media access control protocol. We have developed some scenarios to evaluate this architecture. Simulation results show that the proposed link can decrease packet drop probability in terms of orders of magnitude in OBS networks without optical buffers and wavelength conversions.

By analyzing the characteristic of optical buffer device and traffic's self-similarity we put forward optimizing configuration scheme for OPS node based on DLOB. According to configuring DLOB's loop length and the total number of DLOBs, average delay and delay jitter can be reduced which is showed in simulation experiment.

In today's distributed computing systems, a large amount files contain huge data need to be transferred to their destination as soon as possible or else the quality of these systems will be seriously affected, and these transfer requests arrived dynamically. We propose some effective heuristic algorithm to this problem with the purposes of minimizing the maximal file transmitting time, and we can get some primal results from the algorithm. However, as we known, the problem of routing and scheduling for the dynamic arriving files in the optical network has a large number of constrains and the exact solution is computationally expensive, so it is hard to get the optimal result about this problem and we can not know whether the heuristic results is good or how closed it closed to its optimal result. In order to get some more detail results, we apply the approach called Lagrangian relaxation combined with subgradient-based method and utility the heuristic result to compute the lower bound of the optimal solution, and we consider the optimal target of minimizing the maximal file transmitting complete time for it's an important aspect with the file transmitting problem. We mainly use Lagrangian relaxation (LR) to research the dynamical lager file transmitting problem. Firstly, in order to apply the LR method we formulation our dynamic file routing scheduling and distributing problem in WDM optical network into mathematic model with some corresponding constraints. Secondly, change the formulation with some added variables to let it more suitable for LR and then introduce the Lagrangian multipliers into the model to obtain the Lagrangian function. With this function we can divided it into some small independent problems that could let it be solved more easily and at last we utilize the result received from the heuristic algorithm to solve the Lagrangian multiplier problem with subgradient-based method in order to getting the sharpest possible lower bound. With the comparison of our simulation results, we can prove that the Subgradient algorithm based on LR can get better results of the file transmitting time than the heuristic algorithm, and with the theorem of Lagrangian Bounding Principle we can know that value of LR method is a lower bound on the optimal value.

In this paper, the problem of how to provide survivability to multi-domain optical mesh networks is studied. Differing from the survivable research on single-domain networks, the whole network topology is unknown for all the source nodes. According to this point, we propose a novel protection algorithm called Enhanced Local Segment- Shared Protection (ELSSP) based on the simplified virtual topology for dynamic connections in multi-domain optical mesh networks. For the dynamic connections, ELSSP assigns rational link-cost for the virtual link on the virtual topology when calculating the primary path, and provides segment-shared protection in the passing domains of the primary path. Simulation results show the good network performance of ELSSP.

After over twenty years since the demonstration of the first QKD systems in 1989 there have been two categories of QKD systems, namely fiber QKD systems and wireless QKD systems. The security of a practical QKD system over wireless optical paths is degraded due to imperfectness of devices and channel loss. Very simple attacks, such as beamsplitting attacks (BS), could jeopardize the security of polarization coding QKD systems. Among all the attacks photon number splitting (PNS) is the most threatening one against the security of polarization coding QKD systems. The upper bound for secure transmission length of free-space QKD systems based on polarization coding under PNS attacks is provided here as well as the secure transmission criterion. The theoretic results predicted in this paper accord with experimental results of Manderbach's work.

In the paper, we propose a digital working vacation scheme cyclic polling system for wavelength division multiplexing (WDM). This system consists of multi-queues (terminals) whose service rate is one of the fast speed and lower speed. the fast speed mode moves cyclically from queue to queue with an exhaustive schedule. Using the decomposition approach, each queue is analyzed as a single server queue with working vacation, which works at a different rate, rather than completely stops during the vacation polling period. Using quasi birth and death chain and matrix geometric solution method, we give some performance parameters, such as the mean number of packet in the queue, throughput, busy period etc. Finally, we analyze the effect of the parameters on the mean queue length and sojourn time through numerical examples.

The performance of WDM-based coherent time-spreading OCDMA system is studied. Upper bound and lower bound of normalized throughput are obtained with 8 wavelength channels and 127 length and 511 length Gold codes. It is shown that when all simultaneous users are equally allocated to different wavelength channels, WDM+ TS OCDMA has much better performance. However, if there is no central control to allocate wavelength channels equally, WDM+ TS OCDMA system has the in-between performance of normalized throughput.

This paper investigates the theory and process of Adaptive Immune Evolutionary Algorithm (AIEA), uses the AIEA to solve the problem of traffic-grooming in All Optical Network (AON). The result shows that the AIEA can be used in arbitrary topology and arbitrary traffic model of AON. Besides, the AIEA has more traffic-grooming ability in complex topology with more nodes than the one with fewer nodes. What's more, the algorithm has advantage in convergence than other algorithms. The result shows the AIEA has effective resource utilization, quick convergence ability, which used in arbitrary topology and traffic matrix of the network.

It researches the requirements of control plane in T-MPLS and analyses the architecture of control plane based on TMPLS testbed. With the growth of the demand for packet transport services, packet transport network is becoming the core technology for transport network, especially T-MPLS. Because of simply implement, independent control plane, complete OAM, T-MPLS becomes the most promising technology for packet transport network. The control plane is responsible for handling dynamically and in real-time network's esources in order to manage the establishment and deletion of label switch path(LSP), and for disseminating and discovering network topology and resource availability through the exchange of control (signaling and routing) messages between neighbor nodes over the data communication network (DCN). It is composed of mainly four modules, connection controller (containing signaling module), routing, resource manager, and protocol controller module. The four modules interact with each other to complete the control plane's functions. The flow of every module is designed in this paper to complete control plane's software programming.

Service differentiated survivable WDM mesh networks providing dedicated path protection and shared path protection are investigated under scheduled traffic. Scheduled routing with holding-time aware and unaware are studied for comparison. To tackle the service provisioning problem, an integer linear programming(ILP) formulation is presented for either situation to find the optimal service provisioning solution for a given set of traffic demands. Numerical results show that over 48% capacity savings are achieved when the holding-time knowledge is exploited.

Ethernet Passive Optical Network(EPON) is one of the most promising fibre-based access techniques which has been more and more widely used in today's access networks. In this paper, we proposed to integrate WLAN technology with EPON to provide wireless service in EPON network. First of all, theoretical analysis is given to support the feasibility of integrating WLAN technology into EPON system. Then, the architecture of the integrated system is descried. The key point of the plan is that an integrated ONU and WLAN BS called ONU-WLAN AP is setup in the system instead of a standard ONU. This is also the main difference between proposed system and a normal EPON system. The ONU-WLAN AP can provide both wired access and wireless access to the end users. System design and implementation is discussed in the following paragraph. Transmission experiments and results of the proposed plan are also described in this paper. The experimental results show that the average traffic throughput of the system could reach to approximate 20Mbps. Finally, conclusions are drawn out: The proposed system can work properly. The integrated system could be a costeffective plan for broadband access network.

Optical grid, the integrated computing environment based on optical network, is expected to be an efficient infrastructure to support advanced data-intensive grid applications. In optical grid, the faults of both computational and network resources are inevitable due to the large scale and high complexity of the system. With the optical network based distributed computing systems extensive applied in the processing of data, the requirement of the application failure probability have been an important indicator of the quality of application and an important aspect the operators consider. This paper will present a task-based analysis method of the application failure probability in optical grid. Then the failure probability of the entire application can be quantified, and the performance of reducing application failure probability in different backup strategies can be compared, so that the different requirements of different clients can be satisfied according to the application failure probability respectively. In optical grid, when the application based DAG (directed acyclic graph) is executed in different backup strategies, the application failure probability and the application complete time is different. This paper will propose new multi-objective differentiated services algorithm (MDSA). New application scheduling algorithm can guarantee the requirement of the failure probability and improve the network resource utilization, realize a compromise between the network operator and the application submission. Then differentiated services can be achieved in optical grid.

A new Optical Resilient Packet Ring buffer model was introduced in this paper, end-to-end delay of packet in Optical Resilient Packet Ring was analyzed. The computing equation for end-to-end average delay of packet with different service priority was proposed,while fiber-delay-lines acting as transfer buffer at node.The relationship between delay and traffic, fiber delay lines unit length is analyzed via the equation.

Optical encoders/decoders based on fiber gratings are designed and fabricated. Experiment on one interference user 2.5Gb/s symbol overlapping optical fast frequency-hopping CDMA (SO-FFH OCDMA) system is demonstrated. Multiaccess interference in SO-FFH OCDMA system is studied, and is compared with that of traditional FFH OCDMA system. Experiment result shows that multi-access interference will degrade the system performance. With one interference user, SO-FFH OCDMA can also decode data correctly at data bit rate 2.5Gb/s.

IGMP is a host-router signaling protocol, and it is used to manage the membership of Internet Protocol multicast groups for IPv4. IGMPv3 adds "source filtering" function required to support SSM. Firstly, this paper analyses the improvements and implementation of IGMPv3. Secondly, it points out the problems when applying IGMPv3 to IPTV carrying network. Then it presents the solutions by improving IGMPv3 protocol to reduce channel zapping time which is one of the important metrics affecting the QoE in IPTV service. Method to realize the end user authentication, authorization and accounting in the IP multicast security and authentication area by extending the IGMPv3 protocol is described at last.

Three methods of distributed reservation named FQRP (Forward Queue Reservation Protocol), FRP-λ (Single Wavelength Forward Reservation Protocol) and Extension of FRP in optical network are proposed in this paper. The three methods make some improvements on the traditional FRP (Forward Reservation Protocol) and BRP (Backward Reservation Protocol). They are described, analyzed and simulated in detail in this paper. The simulation results show that they do have advantages over the traditional two methods. Each has its own advantage.

We propose a novel survivable wavelength-division multiplexed-passive optical network (WDM-PON) based on an N × N cyclic array waveguide grating (AWG) and reflective semiconductor optical amplifiers (RSOAs). ONUs are grouped and connected with extra connection fibres (CFs). Protection resources are provided mutually in ONU pairs. The characteristics of the proposed survivable WDM-PON and wavelength routing scheme are analyzed. Experiments of 10- Gb/s downstream and 1.25-Gb/s upstream transmission experiments are demonstrated to verify our proposed scheme.

In this paper, we focus on the survivable traffic grooming problem under SRLG (Shared Risk Link Groups) constraints in WDM mesh networks. Our objective is to minimize the total network resource for a given traffic demands metric. An efficient simplified ILP (Integer Linear Programming) mathematical formulation is presented. Considering the computational complexity of ILP, a fast heuristic algorithm named Cross Layer Information Routing & Multi-Layer Traffic Grooming (CLIR-MLTG) is proposed and evaluated. We compare our simplified ILP model with the regular ILP model, and the CLIR-MLTG algorithm is also compared both with our simplified model and a heuristic algorithm named SSR in the literature [1]. Finally, we demonstrate in numerical results that our simplified ILP model obtains optimal performance and costs much less time compared to the regular ILP model, and the CLIR-MLTG algorithm outperforms the SSR scheme and it obtains similar performance compared to the ILP model.

In practical multi-granularity WDM optical networks, optical signals will be degraded due to impairments mainly introduced by a number of multi-granularity optical cross-connects (MG-OXC) and fiber links. Even worse, transmission impairments will make the bit-error rate to be unacceptable. We investigate the impact of transmission impairments on optical signal quality in multi-granularity WDM optical networks. A novel dynamic impairment-aware RWA algorithm is proposed based on the presented transmission impairments model. We also evaluate the proposed algorithm in the serial cascaded MG-OXCs network and the interconnecting MG-OXCs network by simulations.

With the successful application of DSL technology, the gradual popularization of FTTH technology, and the evolution from traditional network to NGN, the trend of IP business bearing has been formed. In the future, the optical transmission network of the entire business operations will be primarily responsible for the IP / Ethernet traffic transmission, the optimization of the characteristics of data flow in groups, and will develop in a direction of smart, integrated, broadband, efficient and reliable packet transmission technology (PTN). The PTN based on T-MPLS technology, supporting all types of business (Ethernet, IP / MPLS, FR, ATM, TDM,...), simultaneously supporting general users (wired / wireless) and commercial users in various types of business development, would become a new generation of converged networks of PTN platform. This article analyzes the crucial technologies of PTN in respects of the T-MPLS system of the framework agreement, the hierarchical structure, the network model, operational capacity, protection switching mechanism, etc, and puts forward a PTN-building program facing the entire business operations.

In digital storage system, the aim of channel coding is to improve efficiency and reliability of the channel. In order to improve the capacity and quality for optical storage system, the study on the modulation code is significant. A new RLL (2, 12; 8, 15) run-length limited code is presented. The construction method of the code is discussed and the procedures in encoding and decoding of the code are also given. The major characteristic parameters of different run-length limited codes are compared; the decoder is implemented based on FPGA. The RLL(2, 12; 8, 15) code with high modulation rate, namely high encoding efficient, is fit for high density optical storage systems. The study of modulation code for high-density optical discs with independent intellectual property rights is also of great significance.

Carriers and service providers are rushing to provide Ethernet-based virtual private network services in metro area network (MAN) as the most cost effective way to address the needs of the enterprise network market. To address the fast recovery from any signal failure issue in the Metro Ethernet, we propose a metro Ethernet architecture based on multiple Enhanced Self-protected Spanning Trees (ESST). The recovery mechanism, named Birthday-based Link Replacing Mechanism (BLRM), in this architecture is able to transform a self-protected spanning tree into another spanning tree after any signal link or node failure. Simulation result demonstrates the effectiveness of the BLRM in achieving fast recovery.

Over the past year, there has been a rapid upswing in the number and intensity of standards organizations working on FTTx technology. This document gives a brief outline of each organization, and highlights the main direction of each.

This paper proposes a three-layer service model in optical networks. It utilizes diversity design and service shift mechanisms. Both the analysis and simulation results demonstrate this model can effectively promote the survivability of IP networks.

In this paper, we proposed to integrate WLAN technology with DTV over EPON system to provide wireless service to the DTV set-top-box(STB) for end user. First of all, the whole wireless DTV over EPON system architecture is introduced. There are mainly three parts: the accessing part, EPON system and the wireless receiving part. The working mechanism of each part is also described. Then, the paper focus on the design and implementation of the receiving device: wireless STB. The key point of the design is to integrate WLAN chipset into a normal DTV STB to enable wireless transmission. Hardware design is discussed in detail and software design is also mentioned in the last part of the article. Finally, conclusions are drawn out: WLAN signal could satisfy the transmission bandwidth requirement of DTV over EPON system. The proposed system can work properly.

This work aims to show the possibility of WDM optical network planning using the virtual topology. For this analysis, heuristics methods were applied to take current and future traffic into account to plan a network under limited link capacity. The results suggest that it is feasible to preserve enough open capacity to avoid blocking of future traffic, without network disruption, allocating appropriate link capacities for a given planning period.

We propose and demonstrate a 10Gb/s bidirectional reflective semiconductor optical amplifier (R-SOA) based wavelength-division-multiplexing passive optical network (WDM-PON) utilizing a DPSK signal for down-link and an OOK signal re-modulated for up-link. The eye diagram and the penalty of max eye opening factor (EOP) performance of our simulation results show that the performance of our scheme is as good as those in case of continuous wave(CW) light injected R-SOA. So our scheme is a practical solution to meet the data rate and cost-efficient of the optical links simultaneously in tomorrow's WDM-PON access networks.

Multiple-access interference (MAI) and beat noise (BN) are analyzed in asynchronous coherent time-spreading OCDMA system according to the aperiodic cross-correlation function. Relationship between the mean intensity of aperiodic crosscorrelation and MAI and BN is deduced. Considering the worst system performance, upper bound of BER performance is evaluated by the maximum value of the mean intensity of aperiodic cross-correlation. For 511 length Gold sequence, coherent time-spreading OCDMA system can support less than 5 interfering users in the worst performance.

To address the disadvantages of the EPON+EOC technical scheme used in the bidirectional reformation of HFC network and broadband access application, this paper proposes the concept of Ethernet Passive Electronic Network(EPEN) based on MPCP protocol. All kinds of modulated EOC technologies are based on CSMA/CA half-duplex mode MAC protocol, showing low efficiency in access network application. Learning from EPON point to multi-point (P2MP) control protocol, this paper originally proposes the concept of full-duplex EPEN based on MPCP protocol for upstream. This paper describes the network architecture of EPEN, and discusses the technical implementation scheme of the EPEN system. At present the core technologies and implementation of software and hardware of EPEN are being studied.

VoD is a very attractive service which used for entertainment, education and other purposes. In this paper, we present an evolution method that integrates the EPON and SCM-PON by WDM technology to provide high dedicated bandwidth for the metropolitan VoD services. Using DVB, IPTV protocol, unicasting and broadcasting method to maximize the system throughput and by numerical analysis, the hybrid PON system can implement the metropolitan VoD services.

We proposed and demonstrated an architecture by integrating radio-on-fiber (ROF) and fiber-to-the-home (FTTH) transport systems simultaneously with the help of injection locking scheme. An externally modulated radio frequency (RF) data stream and a directly modulated baseband (BB) data signal are simultaneously transmitted over an-80 km single-mode fiber (SMF) link. Good bit error rate (BER) performances are achieved in our proposed integrated transport systems. Such a proposed system is attractive and suitable for long-haul transmission due to its flexibility and simple architecture with low-cost.

To standardize home networking information and simplify its management, this paper form a universal information module of various devices in home networking by adopting XML technology and BACnet protocol(XML/BACnet). Then, a software architecture of home networking based on this module is designed, having the function like auto management and maintenance, safety, real-time and remote controlling. Consequently, a home networking system based on this architecture is completed. Tested and evaluated, this system is one easy-using, easy-realizing, nice real-time system with strong heterogeneity and stable safety system.

The paper proposes the design of Home Gateway used in FTTH. On the customer's side, the Home Gateway finishes the optical signal receiving, provides three kinds of interfaces for the video, voice data, and attains the integration of the three services. PAS6301 chip achieves the control of voice, video and data services. Especially on the aspect of the video services, the HG system combines the advantages of IPTV and the DVB to offer a platform for receiving two signals. Users can use ordinary TV to receive both digital TV programs encoded by MPEG-2 and IPTV programs encoded by H.264.

A ring based optical virtual private network (OVPN) employing contention sensing and avoidance is proposed to deliver multiple-to-multiple group-multicast traffic. The network architecture is presented and its operation principles as well as performance are investigated. The main contribution of this article is the presentation of an innovative group-multicast capable OVPN architecture with technologies available today.

We propose an output 2-queue optical buffer system to provide multi-class QoS in terms of throughput, loss and delay, that can be realized by implementing the state-of-art optical buffer technology in the near future. The proposed optical buffer system can be used at the output port of an optical router in an optical switching network (OPS). It resembles its electronic equivalent in terms of operational functionality. Analytical modeling derives the performance metrics and the numerical results are demonstrated.

In optical networks, fault detection is an important aspect of network management system, also a prerequisite to the protection/restoration process triggered when abnormal conditions or break-downs of links or nodes happen in the networks. The ever-growing development of optical networks, such as the division of domains and the diversity of the techniques, protocols and administration policies adopted by different network operators, challenges the traditional fault localization methods applied in single domain networks. This paper investigates the fault localization problems for single-domains, and then extends it into multi-domain optical networks. A Multi-Domain Layered Fault Localization (MD-LFL) Model is proposed to address this problem. And then its application scenario is presented to show its effectiveness by sub-dividing the computational complexity among domains and introducing cooperative diagnosis between domains.

The character of 1Gbps bandwidth and tree-based structure make EPON very suitable for broadcast or multicast services such as IPTV. The document proposed a novel scheme, based on the former research for controllable multicast over EPON system, mainly considering system security and maintainability. It can both control the IPTV program source's and the receiver's validity, improving the efficiency and precision. The processing of the two configurations mode is given in detail.