This PDF file contains the front matter associated with SPIE Proceedings Volume 7633, including the Title Page, Copyright information, Table of Contents, and the Conference Committee listing.

The hierarchical optical path network that utilizes wavebands is recognized as very important in meeting the future explosive growth of traffic demand. The use of backup paths is crucial to realize reliable networks. In order to build survivable hierarchical optical path networks, the two types of protection mechanisms implemented in the optical layer are identified: waveband protection and wavelength path protection. We have already developed a novel network design algorithm that utilizes waveband protection and showed that it can reduce network costs significantly. Another type of protection, wavelength path protection, was also developed and we have demonstrated that further network cost reduction can be attained in the area of small traffic demands. The effectiveness of the wavelength path protection algorithm was confirmed for some network parameter values, however, further clarification is necessary regarding the impact on network cost of network parameters, especially waveband capacity, a major network parameter. This paper investigates network cost variation with waveband capacity for hierarchical optical path networks that utilize waveband and wavelength path protection. Numerical experiments demonstrate the importance of waveband capacity optimization.

A novel multi-domain L1VPN provisioning architecture is proposed based on service plane of the adaptive multiservices provisioning platform. It can provide the inter-domain L1VPN services flexibly, and can establish different L1VPNs by analyzing different service characteristics and constraints. Moreover, the architecture we proposed was experimentally demonstrated in our AMSON testbed.

We propose a novel scheme to overlay multicast service over a wavelength-division-multiplexed (WDM) passive optical network (PON) based on dynamic wavelength reflection scheme. At the optical line terminal (OLT), for each WDM channel, a dynamic wavelength reflector is used to selectively bypass the corresponding optical carrier to avoid multicast data superimposition, while the remaining optical carriers go through the dynamic wavelength reflector and simultaneously carry the multicast data. The proposed scheme is experimentally demonstrated with 5-Gb/s downstream unicast, multicast data and 625-Mb/s upstream re-modulated data.

A reconfigurable protection is extended in ASON survivability testbed SURBED, the performance and extra overhead are presented and also compared with several survivability techniques (1+1, shared mesh, rerouting, etc) in SURBED. Pre configured protection cycles (p-cycle) is a kind of preconfigured protection mechanism that can have mesh-like resource efficiency and ring-link protection speed. But under multi-region survivable network, the topology and resource information is not always available due to policy reasons or other else, so distributed approach with agent negotiation is attractable. As the dimensions of WDM optical network grows larger and larger, it tends to be a mesh structure. Based on SURBED, we take extensive simulation to investigate the performance of different signaling strategies by multiregion agent negotiations. Unlike previous work, we focus on the reconfiguring process rather than the switching over operation of control plane in ASON. The backbone is overlaid, in which most of devices are IP router networks. Routers are interconnected by fibers or SDH or WDM links. It is the same with non-reconfigured p-cycle, the reconfiguration need a criterion, and here we use the network real-time performance as a reconfiguration trigger to take actions on pcycle reconfiguration.

Optical burst switching (OBS) is thought to be the best way to adapt the bursty IP traffic over optical WDM networks. 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 is mainly hampered at network nodes by burst contention. The contention is serious constraints on the capacity of OBS networks. A leading source of contention is the congestion at congested links where a limited number of wavelengths are shared by a lot of bursts. These congested links are stemming from burst routing algorithms. In this paper, we studied on routing path competition phenomenon. We proposed a differentiated QoS aware multipath routing algorithm for OBS networks. This algorithm is evaluated through extensive numerical studies over a typical topology--the 14-node NSFnet. Results show that it outperforms the existing algorithms in terms of burst blocking probability and link utilization.

In transparent Wavelength Division Multiplexing (WDM) networks, the signal is transported from source to destination in the optical domain through all-optical channels, or lightpaths. A lightpath may traverse several fiber segments and optical components that in general degrade the optical signal. This effect introduces the need for considering physical layer impairments during the connection-provisioning phase. Physical layer impairments can be divided into linear and non-linear. Both types of impairments are highly dependent on the fiber characteristics, which in turn are sensitive to length, temperature and age. A close look at the fiber infrastructure of today's network operators reveals a situation where old and newly deployed fibers coexist in the network. This heterogeneous fiber plant presents a challenge. A tradeoff should be found between the QoS requirements of connection requests and the use of the available (old and new) network resources. This calls for a provisioning mechanism able to adapt to the various fiber composition scenarios. In parallel, given the need for service differentiation, the authors recently proposed an Impairment Constraint Based Routing (ICBR) algorithm, referred to as ICBR-Diff, supporting differentiation of services at the BER (Bit Error Rate) level in a network with a homogeneous fiber infrastructure. In this paper the ICBR-Diff algorithm is extended to heterogeneous network; particularly, it is evaluated in WDM networks with fiber links having varying Polarization Mode Dispersion characteristics, i.e., with old and new fiber coexisting. Simulation results show that the ICBR-Diff algorithm exhibits high adaptability in a heterogeneous fiber composition scenario. This translates into improved performance in terms of blocking probability, when compared to traditional impairment aware routing algorithms.

Clock synchronization is an essential problem faced by packet transport network. PTP (precise time protocol) can be used to solve this problem.Its feasibility has been verified by related literatures. In this paper, we use the OPNET simulation platform to further study the key factors affect the clocks' accuracy of packet-based synchronization schemes, e.g., queuing disciplines, network traffic load, whether low-pass filter is used or not, and deployment of the PTP enabled router. We intend to answer the question how much influence can each factor make to the clocks' synchronization performance.

Today networks are too cumbersome due to the incoherent integration of various technologies developed over time. Convergence of data, voice, and video has caused the retro-fitting of a network to support non-native traffic types and this has led to sub-par performance for every traffic type. Fundamental changes are warranted to support high bandwidth traffic. The network must guarantee end-to-end delivery and provide inherent protection within the physical structure. We propose that the main traffic types of today's applications should be carried using the following features: connection-oriented service, traffic grooming to fill those connection oriented pipes, and forward error recovery with protection provided in the physical layer. We present some of our research results.

The recent advances in network control technology such as GMPLS enable us to dynamically manage path connections with different granularities through a unified control interface, which can improve network resource utilization. In this paper, we investigate the effect of path granularity and operation interval on dynamic path network operation with the aim of minimizing unused bandwidth and the frequency of control actions. We develop a simple technique that applies low-pass filters to the varying Internet traffic; we show that Internet traffic streams captured in different environment exhibit a common trend with regard to Internet traffic parameters.

We propose an efficient network architecture to implement optical fast circuit switching. Future bandwidth abundant services such as Ultra High Definition Television (UHDTV), lambda-leased line services, and layer-one optical VPNs will generate less-bursty traffic that will fill wavelength path capacity. To realize effective optical fast switching networks, we introduce a hierarchical structure that combines physical network and optical path levels. A higher physical layer network (transit network) bridges several lower layer networks (local networks). The optical path layer is divided into two layers; a waveband path (a group of wavelength paths), and wavelength path layer. The transit networks employ large granular optical paths, waveband paths. The transit network creates an adaptive virtual topology that can efficiently carry wavelength path connection requests between lower layer network nodes. Numerical experiments show that the proposed hierarchical network greatly reduces the necessary number of optical switch ports at the blocking probability equivalent to that of the single layer network. The effectiveness of the proposed architecture are confirmed for various network sizes.

Fault-tolerant scheduling is an important issue for optical gird applications because of a wide range of grid resource failures. To improve the availability of the DAGs (directed acyclic graphs), a primary-backup approach is considered when making DAG scheduling decision. Experiments demonstrate the effectiveness and the practicability of the proposed scheme.

Different technical challenges for migration from GPON to 10GPON are discussed and also put into context of standardization. The challenges include how to handle an increased upstream capacity through use of burst-mode receivers and DBA algorithms, extension of reach and wavelength planning for future proofing the fiber infrastructure. Finally the roadmap towards 10GPON standardization is discussed.

A scheme for estimating T-CONT bandwidth demand within dynamic bandwidth allocation (DBA) for next-generation passive optical network (NG-PON) is proposed and evaluated. Estimating T-CONT bandwidth demand is considered as a main challenge in gigabit-capable passive optical network (GPON) DBA. The proposed minimum guaranteed T-CONT content (MGTC) scheme allows for a more conservative estimate of bandwidth demand. It is shown that at high load significant improvements in delay, jitter and bandwidth utilization can be achieved with the proposed scheme. For light loads the conventional scheme shows better delay performance. However, this may be overcome by controlled overgranting in the MGTC scheme.

We propose a novel multicast-enable wavelength-division-multiplexed (WDM) passive-optical-network (PON) architecture that uses two sidebands of Optical Carrier Suppressed (OCS)-DPSK signal to carry the unicast and multicast data. The multicast wavelength control is realized through a Wavelength Selective Switch (WSS). At the optical line terminal (OLT), for each WDM channel, a single drive Mach-Zehnder modulator (MZM) is used to generate a subcarrier double-sideband (DSB) DPSK signal. Then, the lower-sideband signals are separated by an interleaver and subsequently modulated to deliver the multicast data, while the remaining upper-sideband DPSK signals carry the downstream unicast data. In the optical network units (ONUs), part of the downstream unicast signal power is re-modulated for upstream transmission, which enables color-free ONUs. This proposed scheme is experimentally demonstrated with a 1.25Gb/s downstream unicast and multicast data.

As the scalability of network grows rapidly, multiple carriers can interoperate under a common control plane and operate separately. A large network is often composed of a lot of network nodes which incurs great burdens for network management and controlling bandwidth spending. Networks comprised of LSRs with different data plane switching layers controlled by a single GMPLS control plane instance is referred as multi-region network (MRN). Most of previous researches are focus on network partition with the principles of shortest path first or simply based on the geographical position. However, the performance of the call block ratio is always not so good. In this paper, we consider a kind of least imbalance flows among all the sub domains, which aims at providing relative average load for each routing area in the network. This work also introduces a method for the address confliction while more than two areas are united, and also the address.

A WDM PON(wavelength-division-multiplexed passive optical network) structure using 40Gb/s FSK signal for downstream transmission and 10Gb/s orthogonal FSK/ASK re-modulated signal for upstream transmission is demonstrated. In the downstream direction, a novel FSK modulation scheme based on phase-to-intensity modulation conversion is adopted to generate a FSK signal at 40Gb/s, while orthogonal modulation format is used in the upstream transmission. The performance of the system is studied both with and without scattering. Simulation results show that the PON system can transmit for more than 20km, which is a promising candidate for future highspeed access network.

Designing efficient bandwidth allocation algorithms is a critical issue in EPON. In this paper, a novel DBA algorithm is presented to efficiently and fairly allocate bandwidth among different users. This algorithm is integrated with non-strict priority scheduling and priority queuing to implement a cost effective EPON network with QoS support.

This paper reports on a FTTH field trial with GPON (Gigabit-capable passive optical network) technology in the network of Deutsche Telekom in the region of the cities of Berlin and Potsdam. Focus of this trial was to gain practical experience regarding GPON technology, fibre installation in existing ducts with micro duct technology, fibre cabling in customer buildings and impact on operational processes. Furthermore it is reported on an initial Deutsche Telekom FTTB deployment based on GPON technology in the city of Dresden with the main targets to obtain practical deployment and operation experiences with fibre-based access networks and to provide broadband access to a part of the city formerly not servable by DSL (digital subscriber line) technology.

We propose and experimentally demonstrate a novel architecture for OFDM-PON system. The down link is based on single side band OFDM modulation while the up link makes use of the carrier information retained in another side band to intensity modulate the data. The downstream transmission will be demonstrated experimentally to exceed 10Gb/s using 4-QAM while the upstream transmission will exceed 2.5-Gb/s using ASK modulation and Time Division Multiple Access (TDMA). This architecture can perfectly overcome the Rayleigh scattering effect. It has been proved the transmission for both directions can reach 10km without bit error. And if the scattering effect is considered, the power penalty will be less than 0.8dB for downlink and 1.8dB for uplink.

A passive double star (PDS) network, which shares one optical fiber with multiple subscribers by using a power splitter, is being deployed as an infrastructure for the passive optical network (PON) systems. One of the main focuses for PON network planning is to determine the optical cable route of a point-to-multipoint network that connects every subscriber to the central offices (COs) through the power splitter(s) within a limited deployment cost under realistic restrictions, such as possible fiber paths, the splitting ratio of optical splitters and locations, when the locations of COs and subscribers are given. In this paper, we propose a novel suboptimal design algorithm for PON outside plant deployment. Under the realistic restrictions, the algorithm can automatically generate a suboptimal PON network that connects every subscriber to the COs in terms of total fiber length.

We propose a novel scheme to realize unicast and multicast in WDM-PON. Unicast data and multicast data are modulated in Amplitude Shift Keying (ASK) format and Differential Phase Shift Keying (DPSK) format respectively. Reflective Semiconductor Optical Amplifier (RSOA) is used to selectively erase the DPSK signal by controlling the optical switch, and multicast is realized.

Backup path reprovisioning and activation planning (RAP) scheme is investigated in survivable WDM mesh networks providing shared path protection under scheduled traffic. Results show that over 70% dual-failure restorability benefits are achieved when the RAP is applied.

We propose an improved postponed lightpath teardown (PLT) strategy by introducing a lightpath splitting mechanism in multi-layer optical networks. Based on the original PLT strategies, this paper introduces a mechanism to split the unoccupied shortest lightpaths into shorter ones when the network operates in a steady state. Results show that the improved strategy achieves higher utilizations of both wavelength-links and transceivers under different scenarios of load and transceiver number, and significantly reduce the blocking probability (CBP).

A post path calculation process considering the flexibility of rerouting backup paths is proposed in this paper to achieve blocking-differentiated path provisioning for two classes of connection requests in semi-dynamic sharedpath- protected optical networks. Numerical results show that with the proposed post routing process, the blocking gap between two request classes is over 90% under all network load cases without increasing the overall blocking probability.

Silica waveguide planar lightwave circuit (PLC) technology is very useful for fabricating compact and high performance optical devices for optical communication. Wavelength multiplexers and optical switches for ROADM and OXC are still being developed to improve performance further. New devices for an advanced modulation format can also be fabricated with PLC technology.

Ultralow-latency and less power consumption have become necessary in multi-processor interconnection network on chip, photonic interconnection as a solution to meet above requirement, provides high performance interconnection on chip. But the photonic network on chip architecture design and performance is limited because photonic interconnection hasn't buffer, photonic network architecture must be designed to relieve this limitation. In this paper, we present a multi-channel photonic network on chip architecture employing deflection routing, optical data packets can inject/eject from processor core by four channels at the same time. Simulation result shows this network architecture has 60% latency decrease compared to generic photonic network on chip, and the photonic network architecture is only consume 7% power of the electronic interconnection network on chip with the same scale.

A large number of IP cores will be included in the future systems-on-chip (SoC). Traditional bus-based architectures are no longer suitable for modern chip design, since it is difficult to expand, consumes much power and takes much area. Network-on-chip (NoC), which employs networks to replace buses as a scalable global communication platform, has been proposed to cope with these problems. However, limited bandwidth, long delay and high power consumption will become bottlenecks as NoC scales to large sizes. Based on silicon optical interconnect, optical network-on-chip (ONoC) can offer significant bandwidth and power advantages, which provides a promising solution to overcome these limitations. In this paper, we simulated and compared several ONoCs based on the topologies including 2D Mesh, 3D Mesh, 2D Fat Tree(FT) and 2D Butterfly Fat Tree(BFT) in terms of the end-to-end delay and network throughput. The results showed that 3D Mesh has the best performance among the listed topologies.

In this paper, a novel and simple all-optical scheme for generating UWB monocycle pulses based on the nonlinear polarization rotation in an SOA is proposed and experimentally demonstrated. UWB spectrum generated from 20Gb/s RZ signal has a centre frequency at about 5.0 GHz and a -10 dB bandwidth of about 10 GHz with a fractional bandwidth close to 200% which is wide enough for UWB communication. The UWB signals have been transmitted in a 100km SMF transmission link. The quality of UWB signal deteriorates with the increase in delay time between RZ and dark RZ, while the attenuation increase for RZ needs a consideration trade-off between restraint for CW and stationarity of UWB signal.

A cost-effective capacity planning method of WDM networks is proposed for increase of traffic load. An improved costbased ant colony algorithm is presented from pheromone increase and update mechanisms to solve the planning problem. Simulation results show the applicability of our planning method.

Effective management of large amount of heterogeneous resources in multi-domain optical grid networks is critical to the performance of the grid application. This article proposes the design of a distributed system for an integrated management of both application and network resources in Distributed Hash Table (DHT) mode, called Distributed Resource Manager (DRM), and the core component Quord in DRM is also implemented and evaluated.

A dynamic domain-sequencing scheme is proposed for Backward Recursive Path Computation Element (PCE) -Based Computation (BRPC) in multi-domain WDM networks. Simulation evaluation shows that the proposed scheme is effective in inter-domain path computation with more efficient resource utilization and lower blocking probabilities.

Recent studies have shown that path diversity provided by overlay networks brings about possibilities to choose alternate optimal path against default one. In the present paper, we present a novel routing mechanism which chooses path based on a composite metric, whose comprising individual metrics are determined by active probing measurements including delay, jitter and bandwidth on the overlay network in a real-time fashion. The applicability of this mechanism is examined under overlay emulation environment, which is supported by the performance data collected from a real enterprise overlay network.

By synergistically capitalizing emerging communication technologies from both the RF and the optical domains, this paper proposes a hybrid system architecture by leveraging the emerging programmable/cognitive radio technology and free space optical communication. We identify some key research issues to enable such a system with high data rate, mobility and ubiquitous coverage, and heterogeneous QoS support.

In this article the author presents the RoF technology based broadband access network architecture for the metro and intercity trains, analyzes the relating issues in realizing the network, and addresses challenges towards the direction. The author first identifies insufficiency of the present wireless communication infrastructure, reviews the RoF technology and its characteristics, presents the network architecture meeting the performance and reliability, describes multi-diversities mechanisms for the two-channel transmission, discusses the RoF system structures and their realization, and finally analyzes the issues of the MMW propagation.

We propose an energy-efficient communication protocol, called optical- LEACH (O-LEACH), for hybrid sensor networks that consist of distributed optical fiber sensor links located at the center and two isolated wireless sensor networks (WSNs) with randomly scattered nodes. Network performances in terms of lifetime are simulated with ~30% improvement over LEACH protocol.

The path protection with reliability guarantee is investigated under multiple constraints. The key challenge of providing ample multi-constrained loop-less paths to the detouring as candidates is addressed by a modified relaxation principle of dominated path. The loops in the new extended sub-paths are explored with an elastic depth, and additive criterions derived from the path information are adopted to obtain more accurate identification of feasible paths. Our hybrid detouring gives priority to the end to end disjoint backup paths and uses the residual space of inter-media nodes to store the potential local disjoint subpaths. The obtained candidate backup paths are selected by appropriate cost functions according to the combined reliability and the consumed resources. Simulation results demonstrate its superiority in restoration path-finding capability leads to a significant improvement in the success rate of the QoS path protection.

A routing architecture based on PCE has been designed for the large, multi-layer and multi-domain optical networks, and a PCE-based fast reroute algorithm has been proposed for multi-failures in multi-domain optical networks.

Current trends in communications indicate that multicasting is becoming increasingly popular and important in networking applications. Since multicasting can be supported more efficiently in optical domain by utilizing the inherent light-splitting capacity of optical devices, such as optical splitters, than by copying data in electronic domain, issues concerning running multicast sessions in the all-optical networks have received much attention in recent years. In this paper, different multicasting schemes and their performance in the Optical Packet Switched networks are investigated, including the parallel mode, serial mode, and hybrid mode multicasting schemes. Computer simulation results show that compared with the parallel-mode and serial-mode multicasting schemes, hybrid-mode multicasting scheme is the best way to deliver multicast sessions in the Optical Packet Switched networks due to its highest performance.

Requirement of packet transport network based on MPLS-TP are analyzed including in transport plane, OAM, survivability, QoS, control plane and management plane. MPLS-TP standard status is also introduced.

In this paper, a dynamic impairment-aware routing and wavelength assignment (RWA) algorithm is proposed to efficiently provide signal-quality-guaranteed connections and achieve lower blocking probability in all-optical wavelength division multiplexing (WDM) network with high speed optical channel. In the all-optical network, optical connection is set up to carry data signal from source node to destination node without optical-electrical-optical (OEO) conversion via all-optical lightpath. During the travel, the signal transmitted along the lightpath must pass through a number of optical components, such as optical amplifiers, optical crossconnect switches (OXCs) and fiber segments. While the optical signal propagates to the destination, the impairment would accumulate along the all-optical lightpath with these components, which makes the quality of signal degrade continuously. When the impairment accumulation comes to a serious degree, the bit-error rate (BER) would be too high to be acceptable in the destination receiver, the good quality of service for a connection request could not be provided in physical layer and the connect request would be rejected in control plane because of impairment accumulation. Therefore new techniques in both physical layer and network layer are necessary for decrease the influence of impairment accumulation. Here we investigate a new dynamic RWA algorithm to lower blocking probability of connection. By jointly consideration of both available wavelength and wavelength dependent impairments, the routing weight functions can be dynamically updated to accommodate between wavelength blocking and physical layer blocking. Simulation results indicate that the proposed algorithm can achieve lower blocking probability.

A novel multi-granularity optical switching node with wavelength management poll resource is presented, which is much more cost effective and multifunctional than the conventional architectures.

Proposed in this paper is a dual routing engine-based routing architecture for multi-domain multi-layer optical networks. An interpreter-based modeling of the group engine is discussed and a preliminary routing strategy is designed with consideration of scalability and flexibility.

To improve robustness for the complex network, this paper propose a novel probabilistic approach to allocate redundancy links based on nodal degree, i.e. preferential allocation, inverse preferential allocation, and hiding redundancy links (HRL) strategy. We find that allocating redundancy links can efficiently improve robustness of networks to tolerate the coordinated attack. Especially, I-I case, which both source node and destination node are chosen by inverse preferential allocation, performs the best. Furthermore, we investigate the effects of HRL strategy under the coordinated attack, and the simulation shows that HRL strategy can achieve the better performance than never. More interestingly, the relationship between threshold of crash and the number of redundancy links appears to achieve nonlinear by the HRL strategy. Therefore, a small fraction of redundancy links is highly contributed to improve the tolerance attack dramatically. This paper provides an insight on improving network robustness to tolerant the coordinated attack by allocating redundancy links.

We propose a Mixed Sharing Auxiliary Graph (MSAG) for dynamic traffic grooming in heterogeneous WDM mesh networks. Based on MSAG model, a novel heuristic named BLSW-ILMS (Backup LSP Shared Working Lightpath with Inter-layer Mixed Sharing) is proposed. Simulation results show that the proposed algorithm can efficiently decrease the blocking probability.

The protection scheme is a key issue in survivable wavelength division multiplexing (WDM) optical networks. In this paper, we investigate the problem of multicast requests in survivable networks against any single-link failure and propose a novel scheme called segment protection with segment route (SPSR). Through dynamically adjusting link-cost according to the current network state, SPSR establishes a primary light-tree and corresponding link-disjoint backup segments for each multicast connection request. An analytical model was proposed for blocking probability analysis and adaptive routing algorithm to solve the potential shortcomings in the real network. We make the use of the idea of segmented routes according to the branch nodes. We then combine the each segment to get the whole blocking probability. Simulation results show that with the consideration of load balancing SPSR exhibits better blocking probability and resource utilization performance relative to other schemes.

Wavelength division multiplexing (WDM) technology is considered to be the ultimate answer to the rapidly growing capacity demand of next generation networks. Many routing and wavelength assignment (RWA) algorithms, proposed for lightpath provisioning, base their routing decisions on the availability of network resources, and assume that optical fibers and components are ideal. In reality, physical impairments degrade the quality of the optical signal propagating through fiber segments and optical components. To cope with this problem, Impairment Constraint Based Routing (ICBR) algorithms, that consider physical impairments during connection provisioning, are currently proposed to prevent selecting lightpaths with poor signal quality. However, these algorithms support only a single (highest) quality of transmission threshold, the same one for all connection requests. This does not fit well with the variety of services, with potentially disparate QoS requirements, that the next generation networks are expected to support. Consequently, the efficiency of network resource utilization is reduced. This paper demonstrates that a significant improvement, in terms of blocking probability, can be achieved when using an ICBR algorithm with differentiated physical layer constraints. Performance is compared with conventional impairment aware routing approaches when unprotected, shared path protected (SPP), and dedicated path protected (DPP) connection requests are considered. The achieved improvement is a result of more efficient resource utilization as unnecessary connection blocking can be avoided by selecting network resources offering optical signal quality that is "good enough" to satisfy a specific connection request.

Zero Touch photonic solutions require a carefully designed control plane to cope with physical impairments and the need for 3R regeneration. GMPLS providing the flexibility to add regeneration where needed, extends the reach of WDM network domains beyond the physical limits of pure wavelength switching.

We propose an inter-domain path routing algorithm for multi-domain photonic networks. The proposed algorithm introduces a step-wise weighting technique and utilizes averaged link load information of each domain and that of a link set between each adjacent domain pair. Numerical experiments demonstrate that the proposed algorithm outperforms a conventional method in terms of blocking probability and matches that achieved without domain segmentation.

In this paper, a p-cycle generation algorithm is proposed that can protect service of static traffic model in WDM network. As the basic p-cycle sets calculated by algorithm such as the straddle link algorithm (SLA) always contain a straddle link in it. That is why the protection efficiency is higher than bidirectional line switch ring (BLSR). Any idle resource can be used for traffic restoration. But if the resource is not configured carefully then not all the resource of idle state can be utilized efficiently. Some extensions of algorithms are brought into effect to improve the protection capability. Based on these extensions, a novel iterative expanding p-cycle is introduced. A platform is constructed in OPNET modeler to verify the performance of this algorithm. Simulation results show that after several expansions for the basic p-cycle sets, the efficiency is improved. The pre-configured resource is deployed before any requests come. Comparison between different network topologies is studied. Finally, the results are analyzed and some directivity is given for further research.

It is essential for us to do research on collaboration mechanism in management-control plane in intelligent optical networks. We propose an interaction schemes using mobile agent between the control plane and management plane to solve information collection delay and management messages congestion. We construct the integrated platform by mobile agent.

Hierarchical (multi-core) Wavelength Division Multiplexing (WDM) networks present a challenging design problem to the network designer who wishes to establish all-optical circuits end-to-end and across multiple network cores. Due to the nature of the hierarchical structure and its traffic distribution, it is likely that the inner core requires more capacity when compared to the capacity required by the metro cores, which are individually connected to the inner core. This capacity mismatch cannot be addressed by assigning distinct transmission rates to each core, as this solution would result in using electronic time division add-drop multiplexer to interconnect the traffic across cores with distinct rates. An alternative solution to addressing the capacity mismatch betweenWDM metro and inner core is explored in this paper, which is based on a limited number of wavelengths (a subset of the full set) being used in the metro core, when compared to the full set of wavelengths being used in the inner core. Two available architectures are presented in the paper, discussing their respective advantages and disadvantages.

With ever increasing requirements posed by significant advances in networking, service providers may use traffic engineering (TE) techniques to efficiently manage resources and provide consistent quality of service (QoS). Briefly, constraint-based path computation involves the pruning of links that do not satisfy constraints and subsequently using the shortest path algorithm on the resulting sub-graph. This process is simple and efficient when the path involves only one domain, but can potentially become severely resource heavy, complex, and inefficient when multiple domains are involved. To address this problem, the Path Computation Element (PCE) architecture has been proposed to allow multilayer path computation to be simple and efficient. The PCE architecture introduces a special computational entity that will cooperate with similar entities to compute the best possible path through multiple layers. This paper analyses the PCE architecture and the traditional method to computing the routing through multiple layers, and then an improved scheme of path selection for multi-layer network based on PCE and VNTM is present and evaluated on the simulation platform of NSFNET, which is used to avoid traffic redundancy caused by low layer's invisibility to the high layer.

This paper studies generalized k-ary Data Vortex networks both with and without node buffering. The comparison study focuses on effects of buffering in 4-ary networks and in original binary networks. The results show that with buffers the 4-ary networks gain more performance benefit than that in the binary networks. This is due to significantly reduced traffic backpressure that is present in 4-ary networks in bufferless operations. Detailed comparisons between 4-ary and binary networks with two different buffering schemes are presented. The routing performance in throughput and latency are studied for different network and load conditions. For practical implementation, additional cost due to buffering should be justified for the desirable performance enhancement.

Recently, several distributed attacks based on local information have been emerged threating network security. The coordinated attack is one of the most destructive attacks, with an important property to identify the next-step target through coordinating all crashed nodes. It has been ignored in most of existing researches that the crashed nodes network size will become larger and larger along with the coordinated attack, which makes the coordinated attack crash each node in the same time virtually impossible. In this work, we propose R.C.H strategy, which removes a fraction of crashed hub nodes to improve network robustness against the coordinated attack. The simulation shows that R.C.H strategy is highly contribute to degrade efficiency of the coordinated attack by means of the larger cluster diameter of the crashed node network, which provides some insights for the future research on enhancing robustness of complex communication networks.

The implementation of Cloud Computing over optical networks faces many challenges and opportunities. A cloud computing architecture over optical networks is proposed based on the service plane. And the validity of the architecture we proposed was experimentally demonstrated in our Adaptive Multi-Service Optical Network testbed.

This paper measures and optimizes the network survivability based on average distance. A method is proposed to design a network with constrained minimum average distance and to reduce the computation complexity based on adjacency matrix. This paper studies the optimization of simple connected graph with arbitrary nodes and links under degree limit. There may have several networks with minimum average distance in our method. So we choose the optimal network from them by the network performance analysis under random failure.

We propose a novel scheme of orthogonal wavelength-division-multiplexing (WDM) using superstructure fiber Bragg gratings (SSFBGs) as a square pulse shaper in passive optical networks (PONs). The system supports large numbers of subscribers by the orthogonal interleaving of the sinc-shaped spectrum from each channel and expends the capacity.

In TCP over OBS networks, the parameters from both of TCP and OBS layers will affect the network performance when supporting the upper layer applications, such as Grid application. According to our previous work, we found that TCP window is the key limitation for such a network to support Grid application. The embedded AIMD model of TCP is too conservative. Especially in the network with high bandwidth-delay product scenario, it will take TCP a long time to increase the TCP window according to the slow start and the congestion avoidance rules. In this paper, according to the established analytical model, the TCP window size is optimized in Grid over OBS network. The analytical results show that the optimization of TCP window can improve the TCP throughput significantly.

In this paper, the problem of integrated routing and wavelength assignment in wavelength division multiplexing (WDM) networks is studied. We propose a colored multigraph model for the temporarily available wavelengths. Based on this colored multigraph model, a polynomial time algorithm with complexity O(N²) is also proposed to develop an integrated dynamic routing and wavelength assignment, where N is the number of nodes in a WDM network. Simulation results show that the proposed algorithm achieves lower the connecting blocking probability than the existing ones.

In large dynamic networks it is extremely difficult to maintain accurate routing information on all network nodes. The existing studies have illustrated the impact of imprecise state information on the performance of dynamic routing and wavelength assignment (RWA) algorithms. An algorithm called Bypass Based Optical Routing (BBOR) proposed by Xavier Masip-Bruin et al can reduce the effects of having inaccurate routing information in networks operating under the wavelength-continuity constraint. Then they extended the BBOR mechanism (for convenience it's called EBBOR mechanism below) to be applied to the networks with sparse and limited wavelength conversion. But it only considers the characteristic of wavelength conversion in the step of computing the bypass-paths so that its performance may decline with increasing the degree of wavelength translation (this concept will be explained in the section of introduction again). We will demonstrate the issue through theoretical analysis and introduce a novel algorithm which modifies both the lightpath selection and the bypass-paths computation in comparison to EBBOR algorithm. Simulations show that the Modified EBBOR (MEBBOR) algorithm improves the blocking performance significantly in optical networks with Conversion Capability.

The constructing of multicast light-trees in WDM network with sparse splitting has been proven to be an NP-complete problem. The Member-Only (MO) algorithm provides the best links stress and wavelength usage, and the Distance Priority (DP) algorithm reduces the diameter of tree and average delay, which are sensitive in multicast applications. In order to diminish the diameter of tree and average delay further while maintaining the same link stress and average cost, we present an algorithm called Nearest Connector First Heuristic (NCFH) in this paper. Simulation results show that the proposed algorithm achieves a good performance.

ASON (Automatic Switch Optical Network), which provides fast and dynamic connection from point to point, is the typical representation of the next generation intelligent optical network. There are two main kinds of resource reservation protocols in RSVP-TE (Resource Reservation Protocol-Traffic Engineering), the most widely applied signaling in ASON, which are FRP (forward reservation protocol) and BRP (backward reservation protocol). Both of them are used to apply in single-domain optical network, however, when they are used in multi-domain optical network, there are apparent drawbacks. On base of FRP and BRP, we propose a reservation protocol-IDRPBP, which is based on priority in multi-domain optical network with wavelength converter at border nodes. Through analysis and simulation, we find it can achieve the trade-off between blocking probability and connection setup time.

In order to guarantee the quantity of data traffic delivered in the network, congestion control strategy is adopted. According to the study of many active queue management (AQM) algorithms, this paper proposes a novel active queue management algorithm named JFED. JFED can stabilize queue length at a desirable level by adjusting output traffic rate and adopting a reasonable calculation of packet drop probability based on buffer queue length and traffic jitter; and it support burst packet traffic through the packet delay jitter, so that it can traffic flow medium data. JFED impose effective punishment upon non-responsible flow with a full stateless method. To verify the performance of JFED, it is implemented in NS2 and is compared with RED and CHOKe with respect to different performance metrics. Simulation results show that the proposed JFED algorithm outperforms RED and CHOKe in stabilizing instantaneous queue length and in fairness. It is also shown that JFED enables the link capacity to be fully utilized by stabilizing the queue length at a desirable level, while not incurring excessive packet loss ratio.

We propose and experimentally investigate a highly reliable WDM-PON system. The proposed system adopts the colorless solution. It can detect the failure of the feeder fiber by the polling and restoring mechanism of the system management module, and then trigger the OSW switching to the preplanned path for restoring the disrupted traffic, and produce a broken fiber alarm if needed. The system management module can send polling frames to ONUs periodically and judge whether it is necessary to send trigger signals to the OSW according to the number of ONUs registered in line cards. The preplanned path is the feeder fiber of the neighboring line card. This system can automatically protect against the feeder fiber failures without duplicating the feeder fiber link. The OTDR or optical power meter can be saved which is used for on-line monitoring. Its restoration time is less than 16ms in experiments.

We developed a framework for constraint routing in multi-domain optical network which combines PCE with service plane. With the distributed routing computation characteristic of PCE and the adaptability of service plane for service attributes, the framework provides an optimal SLA-based constraint routing strategy. Experiment results verify our framework and strategy.

In WDM networks, when a session request is received by the network, the problem of establishing light-paths by routing and assigning a wavelength to this connection session is called the Routing and Wavelength Assignment (RWA) problem. The selection of routing algorithm and wavelength assignment method is the key point to solve this problem. And the flooding technique is always used for those routing algorithms to get the resource status and topology information. In this paper, we propose a novel routing algorithm, Rank Total wavelengths and Available wavelengths (RTAW), which dramatically reduces the flooding frequency while guaranteeing a low blocking rate. Compared with other algorithms, the simulations results prove the benefits of this algorithm.