Network slicing in the optical layer enables the creation of multiple virtual network slices on a shared physical infrastructure. When resources are coordinated across multi-domain optical networks, ensuring end-to-end (E2E) performance of the E2E slices is a critical challenge. Due to the autonomy of individual domains and the limited sharing of internal domain information, slice coordinators often face constrained visibility into intra-domain operations. This black-box abstraction of domains makes it difficult to monitor and localize failures from the slice coordinator’s high-level view. In this paper, we address the problem of E2E failure localization under such constraints by focusing on the optimal selection of monitoring paths. We propose a deep reinforcement learning approach to identify a set of E2E monitoring paths that maximizes the capability to localize single-link failures, given a constraint on the number of paths. Our experiments using the GNPy optical network simulator demonstrate that our solution outperforms baseline heuristics and approaches near-optimal performance in localization capability.

Network slicing is a key technological concept for next-generation networking to provide logically dedicated, customized connections for diverse use cases. In many use cases that require access to cloud facilities located far from the network edge, it is crucial to guarantee end-to-end (E2E) performance. However, the composition of E2E network slices demands complex resource coordination among multiple administrative domains that may limit the exposure of network state (e.g., topology, latency) within them for privacy and safety reasons. This paper discusses a resource coordination problem to construct E2E network slices hosted over multiple domains under limited state visibility. The resource coordination task can be formally described as a regret minimization problem with a linear objective function. We present a novel hybrid approach that incorporates partial resource information reported by each domain into the Learning with Linear Rewards (LLR) algorithm. Our experiment results show that the proposed algorithm performs significantly better than other baseline learning algorithms and the LLR algorithm, especially when the traffic patterns of network slices are more dynamic and unstable.

Multicasting is a vital information dissemination technique in Software-Defined Networking (SDN). With SDN, a multicast service can incorporate network functions implemented at different nodes, which is referred to as software-defined multicast. Emerging ubiquitous wireless networks for 5G and Beyond (B5G) inherently support multicast. However, the broadcast nature of wireless channels, especially in dense deployments, leads to neighborhood interference as a primary system degradation factor, which introduces a new challenge for software-defined multicast in wireless mesh networks. To tackle this, this paper introduces an innovative approach, based on the idea of minimizing both the total length cost of the multicast tree and the interference at the same time. Accordingly, a novel bicriteria optimization problem is formulated–Minimum Interference Steiner Tree (MIST), which is the edge-weighted variant of the vertex-weighted secluded Steiner tree problem [1]. To solve the bicriteria problem, instead of resorting to heuristics, this paper employs an innovative approach that is an approximate algorithm for MIST but with guaranteed performance. Specifically, the approach exploits the monotone submodularity property of the interference metric and identifies Pareto optimal solutions for MIST, then converts the problem into the submodular minimization under Steiner tree constraints, and designs a two-stage relaxation algorithm. Simulation results demonstrate and validate the performance of the proposed algorithm.

This paper addresses the virtual network (VN) mapping problems for the network profit optimization in space-division-multiplexing optical data center networks (SDM-ODCNs). We first define both link resource availability (LRA) and node resource availability (NRA) for the VN mapping optimization, by which an integer linear program (ILP) model and two VN mapping approaches are proposed to achieve the high network profit. Simulation results verify that our proposed LRA VN mapping approach achieves greatly close network performance to that by solving solutions of the integer linear program model and significantly outperforms its counterpart approaches. The improved network profits out of the VN mapping is as a result of well suppressed average crosstalk, rejection ratio of VNs, and spectrum fragmentation ratio in SDM-ODCNs.

We address the problem of establishing an end-to-end network slice across multiple domains and propose a Reinforcement Learning-based framework that enables multiple domains to collaborate on end-to-end network slicing admission and allocation. The objective is to maximize the long-term revenue of the network operator. We employ a Graph Neural Network (GNN) to capture the topology features as the encoder. The simulation results show that our framework improves the profit of the network operator by up to 15% compared to a greedy algorithm.

This paper addresses a resource allocation strategy for network slices, where each network slice supports a different federated learning task. A slice is established when a new federated learning model needs to be trained and is released once the training is complete. The goal is to minimize the average network slice holding time while also providing fairness between slice tenants and improving network efficiency. We propose a reinforcement learning-based strategy to periodically reallocate resources according to the current state of each federated learning task. We offer two reinforcement learning models. The first model achieves more stable performance and considers correlations between tasks, while the second model utilizes fewer parameters and is more robust to varying number of tasks. Both approaches have better performance than baseline heuristic methods. We also propose a method to alleviate the effect of various resources scales to make the training stable.

In this work, we consider the network slice composition problem for Service Function Chains (SFCs), which addresses the issue of allocating bandwidth and VNF resources in a way that guarantees the availability of the SFC while minimizing cost. For the purpose of satisfying the availability requirement of the SFC, we adapt a traffic-weighted availability model which ensures that the long-term fraction of traffic supported by the slice topology remains above a desired threshold. We propose a method for composing a single or multi-path slice topology and for properly dimensioning VNF replicas and bandwidth on the slice paths. Through simulations, we show that our proposed algorithm can reduce the total cost of establishment compared to a dedicated protection approach in 5G networks.

This paper addresses a routing selection strategy for elastic network slices that dynamically adjust required resources over time. When admitting elastic initial slice requests, sufficient spare resources on the same path should be reserved to allow existing elastic slices to increase their bandwidth dynamically. We demonstrate a deep Reinforcement Learning (RL) model to intelligently make routing choice decisions for elastic slice requests and inelastic slice requests. This model achieves higher revenue and higher acceptance rates compared to traditional heuristic methods. Due to the lightness of this model, it can be deployed in an embedded system. We can also use a relatively small amount of data to train the model and achieve stable performance. Also, we introduce a Recurrent Neural Network to auto-encode the variable-size environment and train the encoder together with the RL model.

The emergence of diverse network applications demands more flexible and responsive resource allocation for networks. Network slicing is a key enabling technology that provides each network service with a tailored set of network resources to satisfy specific service requirements. The focus of this paper is the network slicing of access networks realized by Passive Optical Networks (PONs). This paper proposes a learning-based Dynamic Bandwidth Allocation (DBA) algorithm for PON access networks, considering slice-awareness, demand-responsiveness, and allocation fairness. Our online convex optimization-based algorithm learns the implicit traffic trend over time and determines the most robust window allocation that reduces the average latency. Our simulation results indicate that the proposed algorithm reduces the average latency by prioritizing delay-sensitive and heavily-loaded ONUs while guaranteeing a minimal window allocation to all ONUs.

This paper addresses the problem of admission control for elastic network slices that may dynamically adjust provisioned bandwidth levels over time. When admitting new slice requests, sufficient spare capacity must be reserved to allow existing elastic slices to dynamically increase their bandwidth allocation when needed. We demonstrate a lightweight deep Reinforcement Learning (RL) model to intelligently make ad-mission control decisions for elastic slice requests and inelastic slice requests. This model achieves higher revenue and higher acceptance rates compared to traditional heuristic methods. Due to the lightness of this model, it can be deployed without GPUs. We can also use a relatively small amount of data to train the model and to achieve stable performance. Also, we introduce a Recurrent Neural Network to encode the variable-size environment and train the encoder with the RL model together.

The increasing demand for diverse network services entails more flexible networks that are realized by virtualized network equipment and functions. When such advanced network systems face a massive failure by natural disasters or attacks, the recovery of the entire system may be conducted progressively due to limited repair resources. The prioritization of network equipment in the recovery phase influences the interim computation and communication capability of systems since the systems are operated under partial functionality. Hence, finding the best recovery order is a critical problem, which is further complicated by virtualization due to the interdependence between virtual network functions and infrastructure elements. This paper deals with a progressive recovery problem under limited resources in networks with VNFs, where some interdependencies exist. We prove the NP-hardness of the progressive recovery problem and approach the optimum solution by introducing DeepPR, a progressive recovery technique based on Deep Reinforcement Learning (Deep RL). Our simulation results indicate that DeepPR can achieve near-optimal solutions in certain networks and is more robust to adversarial failures, compared to a baseline heuristic algorithm.

The increasing reliance upon cloud services entails more flexible networks that are realized by virtualized network equipment and functions. When such advanced network systems face a massive failure by natural disasters or attacks, the recovery of the entire system may be conducted in a progressive way due to limited repair resources. The prioritization of network equipment in the recovery phase influences the interim computation and communication capability of systems, since the systems are operated under partial functionality. Hence, finding the best recovery order is a critical problem, which is further complicated by virtualization due to dependency among network nodes and layers. This paper deals with a progressive recovery problem under limited resources in networks with VNFs, where some dependent network layers exist. We prove the NP-hardness of the progressive recovery problem and approach the optimum solution by introducing DeepPR, a progressive recovery technique based on deep reinforcement learning. Our simulation results indicate that DeepPR can obtain 98.4% of the theoretical optimum in certain networks.

Recent advances in the areas of Internet of Things (IoT), big data, and machine learning have contributed to the rise of a growing number of complex applications. These applications will be data-intensive, delay-sensitive, and real-time as smart devices prevail more in our daily life. Ensuring quality of service (QoS) for delay-sensitive applications is a must, and fog computing is seen as one of the primary enablers for satisfying such tight QoS requirements, as it puts compute, storage, and networking resources closer to the user. In this paper, we first introduce FOGPLAN, a framework for QoS-aware dynamic fog service provisioning (QDFSP). QDFSP concerns the dynamic deployment of application services on fog nodes, or the release of application services that have previously been deployed on fog nodes, in order to meet low latency and QoS requirements of applications while minimizing cost. FOGPLAN framework is practical and operates with no assumptions and minimal information about IoT nodes. Next, we present a possible formulation (as an optimization problem) and two efficient greedy algorithms for addressing the QDFSP at one instance of time. Finally, the FOGPLAN framework is evaluated using a simulation based on real-world traffic traces.

With the Internet of Things (IoT) becoming part of our daily life and our environment, we expect rapid growth in the number of connected devices. IoT is expected to connect billions of devices and humans to bring promising advantages for us. With this growth, fog computing, along with its related edge computing paradigms, such as multi-access edge computing (MEC) and cloudlet, are seen as promising solutions for handling the large volume of security-critical and time-sensitive data that is being produced by the IoT. In this paper, we first provide a tutorial on fog computing and its related computing paradigms, including their similarities and differences. Next, we provide a taxonomy of research topics in fog computing, and through a comprehensive survey, we summarize and categorize the efforts on fog computing and its related computing paradigms. Finally, we provide challenges and future directions for research in fog computing.

The interdependency between different network layers is commonly observed in cyber physical systems and communication networks adopting the dissociation of logic and hardware implementation, such as software defined networking and network function virtualization. This paper formulates an optimization problem to improve the survivability of interdependent networks by restructuring the provisioning relations. A characteristic of the proposed algorithm is that the continuous availability of the entire system is guaranteed during the restructuring of dependencies by the preservation of certain structures in the original networks. Our simulation results demonstrate that the proposed restructuring algorithm can substantially enhance the survivability of interdependent networks and provide insights into the ideal allocation of dependencies.

This paper studies a network design problem to improve the survivability of interdependent networks by restructuring the dependencies. As different types of networked systems become more integrated, the relation between distinct kinds of network devices has become more intertwined. In order to guarantee the robustness of such systems, survivability problems of interdependent networks must be addressed. A characteristic of the proposed algorithm is that the continuous availability of the entire system is guaranteed by the preservation of certain structures in the original networks during the restructuring process. Simulation results demonstrate that the restructuring heuristic can substantially enhance the survivability of interdependent networks.

With the Internet of Things (IoT) becoming a major component of our daily life, understanding how to improve the quality of service for IoT applications through fog computing is becoming an important problem. In this paper, we introduce a general framework for IoT-fog-cloud applications, and propose a delay-minimizing collaboration and offloading policy for fog-capable devices that aims to reduce the service delay for IoT applications. We then develop an analytical model to evaluate our policy and show how the proposed framework helps to reduce IoT service delay.

We address the problem of survivable path pair routing in multi- domain optical networks with geographically correlated failures. The objective is to minimize the risk of simultaneous failure of both the primary and backup paths. We develop a probabilistic model to calculate the simultaneous failure probability of both the paths and consider a topology aggregation scheme for domains based on calculating the physical vulnerable overlapping area of two paths within a domain. We develop an inter-domain minimum overlapping area routing algorithm based on the aggregated information from each domain. We compare our algorithm to Suurballe's Algorithm and an optimal approach and we show that our heuristic approach is effective in reducing the total probability of simultaneous failure.

This paper discusses the intractability and heuristics of the Cluster Leader Election Problem (CLEP), which is to assign a controller to one of the switches in each cluster of a network. In order to improve scalability in Software Defined Networking (SDN), a network can be clustered into subnetworks and managed by multiple distributed controllers. CLEP deals with the optimal placement of the distributed controllers in the subnetworks considering metrics between the controllers, such as distance and connectivity. Additionally, it is shown that some metrics within each subnetwork, which are discussed in other literature can be guaranteed by selecting specific clustering methods before the placement.

In this paper, we address the issue of survivability in statistically shared elastic optical networks. Statistical sharing in elastic optical networks is motivated by the recent revolution of software defined optics, where variable data rates, e.g., base rates and peak rates, are supported for a single connection in the network. Our goal is to minimize the blocking of the arriving connection requests, while at the same time maximizing the chance that existing connection requests are able to switch from base rate to peak rate. Furthermore, the base rate of the connection requests should be survivable to any single link failure. In this paper, we introduce an admission control policy and a spectrum management technique that specifies how the spectrum is partitioned and shared for base rate connections, peak rate connections, and back- up capacity. We then propose several survivability schemes in order to provide dedicated protection to the base rate of the connection requests. We evaluate our proposed schemes through simulations, seeking to find an optimum tradeoff between the base rate and the peak rate blocking.

Network function virtualization (NFV) provides an efficient and flexible way to deploy network services in the form of service function chains (SFCs) by adopting generalized equipment. However, software-based virtualized network functions (VNFs) bring new challenge for network operators in providing service availability guarantees. Traditionally, network-level protection mechanisms are considered separately from function-level VNF backup mechanisms. However, a SFC's availability cannot be guaranteed if only network- level protection mechanisms or only function-level VNF backup mechanisms are considered. In this paper, we propose a coordinated protection mechanism that adopts both backup path protection in the network and VNF replicas at nodes to guarantee a SFC's availability. The proposed mechanism determines the number of replicas required for each VNF in the SFC, and allocates the replicas to physical nodes on the working and backup paths while maintaining ordered dependency among VNFs. Simulation results show that the proposed algorithms contribute to reducing the SFC blocking and the cost of computing resources.

We consider virtual optical network (VON) mapping with the objective of minimizing total network link cost while guaranteeing VON availability, where VON availability is supported by providing shared backup path protection for selected VON links. We develop a matrix-based approach for calculating the availability of a VON mapping with shared backup path protection. In order to efficiently evaluate the maximum availability of a VON mapping, we transform the problem to a group node-weighted Steiner tree problem and propose an efficient auxiliary-graph-based availability (AA) algorithm to find a VON mapping with high availability. Based on the availability evaluation, we propose a heuristic algorithm to map the VON, and numerical results show that our algorithms are effective in achieving high availability while reducing the total link cost and the blocking rate.

In this paper, we study the survivable routing, modulation, and spectrum assignment (S-RMSA) problem for bulk data- flow transfer in elastic optical networks. We design a two- dimensional resource model for specifying resource state information that spans both the spectrum and time dimensions, and we develop a dynamic heuristic RMSA algorithm, Max Resource Utilization (MRU), to efficiently transfer data-flows while meeting deadline constraints and surviving single link failures. Furthermore, we provide a protection resource optimization strategy to progressively release the scheduled protection resources so as to increase spectrum utilization. We examine the performance of our proposed algorithm, and numerical results show that our proposed heuristic algorithm can reduce blocking probability and obtain high spectrum utilization while minimizing fragmentation.

We propose a multi-path fragmentation-aware routing, modulation and spectrum assignment algorithm (RMSA) for advance reservation (AR) and immediate reservation (IR) requests in elastic optical networks. To decrease fragmentation, we propose splitting requests into different parts and transferring each of these parts along a single-path or multi-paths utilizing sliceable bandwidth variable transponders. We first introduce a model to solve the problem and propose a two-dimensional fragmentation occurrence measurement in spectrum and time domains. Then we propose a multi-path fragmentation-aware RMSA algorithm (MPFA). Simulation results show that MPFA can achieve better performance than existing algorithms in terms of blocking probability and spectrum utilization.

We consider virtual optical network (VON) provisioning over a flexible-grid multi-domain optical network with the objective of minimizing total network cost, including the cost of transponders, regenerators, and spectrum. We propose a three-step heuristic algorithm that addresses the issues of domain selection, topology aggregation, and routing, modulation format, and spectrum assignment (RMSA) when mapping virtual optical links onto multi-domain physical optical links. We propose a domain selection technique that attempts to minimize the number of inter- domain virtual optical links. We then suggest a topology aggregation (TA) technique to exchange intra and inter-domain information between domains, and propose a method for RMSA over the aggregated topology. Numerical results show that our heuristic approach is effective in reducing total network cost.

In this paper, we study a new optical network paradigm, where statistical sharing is supported in optical networks. This new paradigm is motivated by the recent revolution of Software Defined Optics (SDO). Software defined variable-bandwidth transponders can support variable data rates for a single connection, i.e. base rates and peak rates. Guaranteeing the peak rates for all the connections simultaneously requires the spectrum for all circuits to be provisioned for peak rates, leading to a large amount of bandwidth usage with low utilization. However, if resources are provisioned for the base rate of each connection, with some shared spectrum resources set aside to allow a fraction of these connections to dynamically switch to their peak rates, then spectrum resources can be allocated more efficiently, allowing a greater number of connections to be accommodated. We reserve a fraction of the spectrum resources for provisioning of base rate of the dynamic traffic while the rest of the spectrum resources are reserved for peak rate, allowing statistical sharing. Our goal is to minimize the blocking of the arriving connection requests, while at the same time maximizing the chance that existing connection requests are able to switch from base rate to peak rate. These two goals conflict with each other; therefore, we need to find a trade-off based on the amount of spectrum resources set aside for the peak rate, and based on the routing, modulation format selection, and spectrum allocation (RMSA) scheme. Our evaluation can help network operators to determine the amount of spectrum that requires to be set aside for peak rates in order to maximize revenue.

In this paper, we consider the problem of routing, modulation, and spectrum assignment (RMSA) for data- flow transfers in elastic optical networks. We design a two-dimensional resource model, in which each data transfer with known data size can be assigned a rectangular block of resources that spans both the spectrum and time dimensions. Furthermore, the dynamic spectral resource allocation problem in elastic optical networks is simplified to the two-dimensional rectangle packing problem. We design a three-tuple for each rectangle placement, and develop a dynamic heuristic algorithm, Best Rectangle Fit (BRF), to efficiently schedule requests while minimizing fragmentation in both spectrum and time domains. We simulate the proposed algorithm, and the results show that the proposed RMSA algorithm (BRF) can greatly decrease blocking probability and increase spectrum utilization.

We consider survivable virtual network mapping in a multi-domain optical network with the objective of minimizing total network link cost for a given virtual traffic demand that is embedded over the multi-domain optical network. The survivability constraint guarantees the connectivity of virtual nodes after any single optical link failure. We propose a hierarchical software-defined networking (H-SDN)-based control plane to exchange information between domains, and we propose heuristic approaches for mapping virtual links onto multi-domain optical links using partition and contraction mechanisms (PCM) on the virtual topology. We show that the proposed PCM technique can reduce time complexity compared to traditional cut set graph theory approaches. Numerical results show that our heuristic approach is effective in reducing total network cost and increasing the successful mapping rate.

In this paper, we study a new paradigm in optical networking in which optical spectrum is allowed to be statistically shared between optical circuits, allowing the oversubscription of optical links. We propose a probabilistic model which estimates the capacity requirements for the optical networks under a certain density of statistical sharing. Simulation results indicate that our proposed model can help with network design decisions, such as admission control and capacity and bandwidth allocation in elastic optical networks.

In this paper, we consider the problem of routing, modulation, and spectrum assignment (RMSA) for immediate and advance reservation requests in elastic optical networks. We design a two-dimensional resource model for maintaining resource state information in both spectrum and time domains. For the purpose of minimizing the blocking probability and increasing spectrum utilization, we develop a two phase RMSA algorithm that attempts to decrease spectrum resource fragmentation by scheduling the requests in a manner that takes into account the holding time of each request. We design a simulation to evaluate the performance of the proposed algorithm, and the experiment results show that our proposed two phase RMSA algorithm can greatly reduce blocking probability and obtain high spectrum utilization.

This paper addresses the minimum network cost problem for survivable virtual optical network mapping in flexible bandwidth optical networks. We develop an ILP model and the LBSD (the largest bandwidth requirement (LB) of virtual links versus the shortest distance (SD)) mapping approach to minimize the network cost for a given set of VONs, and we introduce two baseline mapping approaches, named LCLC (the largest computing resources' requirement versus the largest computing resources' provisioning (LC)) and LCSD (the largest computing resources' requirement versus shortest distance) mapping approaches, for comparison. Simulation results show that LBSD can achieve network cost near the ILP solutions in a 6-node network. Also, LBSD greatly reduces the cost, the spectrum usage, and the number of regenerators compared to LCLC and LCSD in the 6-node and NSFNET networks.

We consider the problem of efficient virtual optical network (VON) mapping in a multi-domain optical network (VON-MD) with the objective of minimizing total network link cost for a given VON demand that is embedded over the multi-domain optical network. Topology aggregation (TA) is used to exchange intra and inter-domain information between domains, and heuristic algorithms are proposed for embedding considering different domain selection techniques and virtual link ordering techniques. We provide an integer linear programming formulation (ILP-VON-MD) to compare with our heuristic approaches. Numerical results show that our heuristic approaches are effective in reducing total network cost.

In this paper, we consider the problem of routing, modulation and spectrum assignment (RMSA) with holding time awareness in elastic optical networks. We consider a two dimensional schedule that maintains link resource usage in both spectrum and time domains, and we develop an RMSA algorithm that attempts to reduce resource fragmentation by scheduling requests in a manner that takes into account the holding time of each request. We examine the performance of the proposed algorithm through simulation experiments, and our results indicates that the proposed RMSA algorithm obtains high spectrum utilization and lower blocking probabilities compared to existing RMSA algorithms that do not make use of holding time information.

This paper addresses the problem of minimizing spectrum usage for shared-path protection with joint failure probability constraint in flexible bandwidth optical networks. To achieve this goal, we propose an integer linear programming (ILP) model for shared-path protection and a heuristic spectrum-aware shared protection (SASP) algorithm that considers joint failure probability. Simulation results show that the ILP model minimizes the total usage of frequency slots and average hops, but leads to high average joint failure probability. Moreover, the SASP algorithm achieves better trade-off between total usage of frequency slots and average joint failure probability compared to the ILP model and a conventional shared-path protection (CSPP) algorithm. As expected, in a 14-nodes network, the SASP algorithm performs better with respect to total spectrum usage and average hops, but results in much larger average joint failure probability compared to the CSPP algorithm.

In this paper, we aim to minimize the total network cost in flexible-grid optical networks with multiple line rates. Besides transponder cost, regenerator cost, and shared infrastructure cost, the cost of regenerator sites is also considered. We first provide the problem definition and formulate the problem as an integer linear program (ILP). We also propose a heuristic algorithm considering both selection and placement of equipment to minimize the total network cost. Simulation results show the heuristic algorithm results in up to 28% cost saving, with no significant increase in spectrum usage.

In a multi-domain network, domains can be defined geographically or administratively. Since only a limited amount of information of each domain is allowed to be broadcasted by domain service providers, Topology Aggregation (TA) is usually adopted. The amount of information provided by each domain may vary based on the service provider. For instance, some domains may provide only a single shortest path between two border nodes, while others may be capable of providing a pair of disjoint paths with minimum total cost. In this case, inter-domain path routing with protection needs to consider and utilize the different levels of information provided by different domains in order to find the best solution. In this paper, we propose two approaches that find a pair of disjoint inter-domain paths with minimum total cost based on a matrix for each domain that includes disjointness information between the aggregated links inside the domain.

Elastic optical networks can increase spectral efficiency by assigning a variable amount of spectral resources to satisfy a traffic request, instead of the fixed amount of spectrum as allocated by current fixed grid optical networks. Waveband networks reduce the number of optical switch ports by aggregating lightpaths into a waveband, and switching them with a single waveband port. We formulate the waveband selection, spectrum allocation, and waveband assignment problem in an elastic optical waveband network that supports a fixed number of waveband sizes. We propose heuristics to select waveband sizes and study the effect of the number of waveband sizes on the number of optical switch ports. We also address the spectrum allocation and the waveband assignment problem in elastic waveband networks with no restrictions on the number of waveband sizes. We propose an auxiliary graph based heuristic and study the trade-offs between spectral usage and the number of optical switch ports. Additionally, we propose traffic ordering strategies and study their efficacy in reducing the number of optical ports and efficiently utilize the optical spectrum in a waveband network.

In this paper, we study the problem of regenerator pool site selection for mixed line rate optical networks (MLR-RPSS), with the objective of minimizing the number of regenerator pool sites for a given set of requests. We first provide the problem definition of MLR-RPSS and show that the MLR-RPSS problem is NP-complete. We then present four algorithms, named Independent algorithm, Sequential algorithm, MLR-combined algorithm, and Weighted MLR-combined algorithm. The performance of the algorithms is compared via simulation and results show that the Weighted MLR-combined algorithm has better performance in most cases. Also, when network load is high, the minimum number of regenerator pool sites will approach a certain limit, and some specific nodes will be more likely to be selected as regenerator pool sites.

In this paper, we study the energy-efficient impairment-constrained regenerator placement (EIRP) problem with the objective of minimizing the total energy consumption in optical networks with mixed line rates. The destination of each path is guaranteed to receive the data correctly from the source based on the regenerator placement. We first provide the problem definition of EIRP and show that the EIRP problem is NP-complete. We then formulate the problem as a mixed integer linear program (MILP) and give results for small scale problems. Two heuristic approaches, named high line rate first (HLRF) and reroute only (RO), are presented. Numerical results show that HLRF achieves good results in both large and small scale problems, and that HLRF achieves higher energy efficiency than RO.

In this paper, we study the problem of reliable collective communication (broadcast or gossip) with the objective of maximizing the reliability of the collective communication. The need for collective communication arises in many problems of parallel and distributed computing, including Grid or cloud computing and database management. We describe the network model, formulate the reliable collective communication problem, prove that the maximum reliable collective communication problem is NP-hard, and provide an integer linear program (ILP) formulation for the problem. We then provide a greedy approximation algorithm to construct collective communication (through a spanning tree) that achieves an approximation ratio of 1 + ln(|V|+α|E|-1) , where α is the average number of shared link risk groups (SRLGs) along links, and |V| and |E| are the total number of vertices and edges of the network, respectively. Simulations demonstrate that our approximation algorithm achieves good performance in both small and large networks and that, in almost 95% of total cases, our algorithm outperforms the modified minimum spanning tree algorithms.

In a multi-domain network, domains may be defined based on geography or administrative boundaries. For domains that are constructed based on geographic locations, multiple links within the same domain may fail simultaneously due to geographically localized failure events. For domains that are defined administratively, a domain administrator may not be able to provide link-disjointness for two paths traversing the domain. Thus, in some cases, it may be necessary to find two domain-disjoint paths to improve survivability. In this paper, we propose an algorithm to find domain-disjoint working and backup paths for multi-domain routing requests. We prove that our approach is optimal in finding domain disjoint paths with minimum total cost.

In order to maintain privacy, domain managers are only willing to broadcast limited information regarding its network for the purpose of necessary multi-domain operations. Topology Aggregation (TA) is now widely adopted to aggregate a domain's detailed topology into an aggregated topology with limited available information. Meanwhile, with the introduction of Shared Risk Link Groups (SRLGs), each of which defines a set of links that fail simultaneously, the TA approach has to consider the presence of SRLGs for generating a more survivable abstracted topology. Exchanging information regarding SRLGs before constructing the TA helps in developing a more survivable aggregated multi-domain network. In this paper, we propose a source domain- specific topology aggregation approach that can help in finding a pair of inter-domain paths with a minimum set of common SRLGs.

In this paper, we introduce an electronic traffic grooming (E-Grooming) technique, and propose a novel bit rate and protocol independent optical-layer traffic grooming (O-Grooming) technique for OFDM-based flexible grid WDM networks. We introduce a novel routing, wavelength assignment, subcarrier assignment, and spectrum allocation (RWSSA) problem for the first time. An effective polynomial-time heuristic algorithms is proposed to address the RWSSA problem in OGrooming capable flexible grid networks. The performance of the proposed E-Grooming and O-Grooming techniques are evaluated with transparent flexible grid networks. Simulation results show that traffic grooming in flexible grid networks improve spectral efficiency compared to transparent networks. In addition to bit rate and protocol transparent, the O-Grooming technique is cost and energy efficient compared to the E-Grooming technique while marginally sacrificing the spectral efficiency.

We formulate the regenerator site selection problem as an integer linear program (ILP) and propose heuristics to solve the design problem in the context of waveband optical networks. We study the effect of the number of regenerator sites on the number of optical ports in waveband networks that support mixed line rates. Results demonstrate that increasing the number of regenerator sites reduces the electronic cost, however the optical switch port cost decreases initially and then remains the same with an increase in the number of regenerator sites.

In this paper, we address the problem of survivable traffic grooming and regenerator placement in optical WDM networks with impairment constraints. The working connections are protected end to end by provisioning bandwidth along a sequence of lightpaths through a dedicated connection-level protection scheme. An auxiliary-graph-based approach is proposed to address the placement of regenerators and grooming equipment for both working and dedicated backup connections in the network with the goal of minimizing the total equipment cost. Simulation results show that the proposed algorithm outperforms a lightpath-level protection algorithm, in which each lightpath is protected separately. We also show the effect of different cost models on equipment placement and evaluate the performance for networks with different line rates.

In multi-domain optical networks, due to confidentiality and management considerations, it may be infeasible to broadcast complete domain state information among domains. Nevertheless, sufficient information on each domain is required in order to keep network activities operating successfully. Topology Aggregation (TA) may be adopted to abstract a domain's detailed topology into an aggregated topology which only shows limited information about the network. With the introduction of Shared Risk Link Groups (SRLGs), each of which includes a set of links affected by the same failure, the TA approach needs to consider the presence of SRLGs to generate a more survivable aggregated topology. In this paper, we propose SRLG-aware topology aggregation approaches that can help in finding a pair of inter-domain paths with a minimum set of common SRLGs.

Waveband networks allow multiple lightpaths to be aggregated and routed as wavebands, resulting in fewer switch ports. In an impairment-aware waveband network, a lightpath needs to be electronically regenerated before its signal to noise ratio reaches an unacceptable level. However, when a lightpath in a waveband needs to be regenerated the entire waveband needs to be demultiplexed into wavelengths resulting in additional wavelength ports at a node. Regenerator placement in a waveband network should be done so as to facilitate banding of lightpaths and to reduce the total cost of switch ports in the network. We formulate the regenerator placement problem in an impairment constrained waveband network as an integer linear program (ILP). We compare the performance of the heuristics with an ILP for a small size network. Results demonstrate that the proposed band-aware reachability graph heuristic yields solution that are very close to the ILP. We study the performance of the heuristics on larger networks and study their effectiveness in reducing the combined cost of optical switch ports and electronic regenerators.

In this paper, we study the problem of multicast flow aggregation with the objective of minimizing the total number of trees. Each established tree can choose a line rate from multiple available line rates. The destinations of each tree are guaranteed to receive data correctly from the source through the chosen line rate. We first use two auxiliary graphs to describe the problem and show that the problem is NP-hard. An integer linear program (ILP) is given to solve small scale problems. We then provide a heuristic approach, named highest rate first (HRF). Numerical results show that HRF achieves good results in both large and small scale problems.

In this paper we evaluate the performance of various power-efficient grooming strategies in WDM optical networks using a tap-or-pass (TOP) node architecture. We examine two grooming concepts, namely lightpath extension and lightpath dropping. Through a mathematical model, we compare the performance of lightpath dropping and lightpath extension in terms of optical power budget. We also propose several grooming strategies, which are applied to lightpath dropping and lightpath extension, as well as traditional grooming. Through extensive simulation, we demonstrate that, given a network with dynamic traffic requests, the proposed grooming strategies lead to considerable energy saving and comparable request blocking, in particular when the network load is moderate.

Abstract Delay‐tolerant networking (DTN) is an architecture to enable data communications between isolated or remote regions, where long delays and intermittent connectivity can be tolerated. An emerging class of DTN, called Vehicular DTNs (VDTN), exploits transportation systems as the transport layer to transfer data. In these networks, vehicles (e.g., buses, boats, and trains) act as mobile nodes and carry data messages around. Mobile nodes can exchange data messages using devices called relay nodes. Relay nodes, placed in strategic positions along vehicle routes, have the capability to download, store, and upload the data messages from/to the mobile nodes. An important issue in VDTN is the optimal placement of the relay nodes such that delay‐tolerant connectivity in VDTN is ensured at minimum cost. In this paper we show that the problem of optimal relay node placement is an NP‐hard problem. Other contributions of this paper are the formulation of the relay node placement problem using integer linear programming (ILP) formulation and the proposal of heuristic algorithms solving the optimization problem. Using simulation results, we compare the performance of each algorithm under different network constraints, such as node storage capability and network topology. Copyright © 2009 John Wiley & Sons, Ltd.

This paper investigates the benefits of enabling services with differentiated energy-saving (DiffEnS) in WDM optical networks with traffic grooming capabilities. DiffEnS can reduce the operating cost (OPEX) as more customers gradually join green programs and subscribe to the energy efficient traffic services referred to as the Green. Based on a modular node design with grooming capabilities, two heuristic algorithms are introduced for minimizing energy consumption during provisioning of energy-aware services. Using our previously developed auxiliary graph, we report on performance of these heuristic algorithms. Our results indicate that, thanks to DiffEnS moderate energy saving can be achieved, even when a small percentage of customers switch to Green services. A major trade-off is the possibility of an increased average end-to-end delay experienced by Green services.

Partitioning a large network into a connected hierarchy is a proven technique for scalability, manageability, and reliability. Although a significant body of works is devoted to this area of interest, a reference to partitioning schemes where each partition remains 2-connected is missing. In optical networks, a failure of a physical component disrupts multiple logical links that share the component. Such a shared risk link group (SRLG) failure is more common than a single link or node failure in the logical topology. In this paper, we introduce the k-SRLG-connected partitioning problem for optical networks, where k ≥ 1. A logical topology is said to be k-SRLG-connected if its nodes remain connected upon any k-1 SRLG failures. We particularly focus on k = 2. Without 2-SRLG-connectivity, a partition can neither provide protection by allocating two SRLG-disjoint paths, nor provide restoration after a failure. Moreover, a single SRLG failure can split the partition into multiple disconnected components, resulting in the failure of the hierarchical network organization itself. We show that finding a 2-SRLG-connected partitioning for a given network is NP-complete and provide a distributed heuristic approach. Simulation results verify that the proposed 2-SRLG-connected partitioning scheme performs better than traditional 1-connected partitioning approaches with regard to survivability and system stability.

In this paper, we consider the problem of dynamic inter-domain traffic routing between a VDTN and a non-DTN (e.g., Internet). The inter-domain traffic can be classified as inbound and outbound traffic. Our main contribution in this work is the introduction of several fault-tolerant routing algorithms for inbound and outbound traffic. Using simulations, we compare the performance of the proposed algorithms in terms of required resources, packet delivery time, and blocking probability.

Compared to energy-awareness research in wireless and copper networks, less attention has been given to energy awareness in optical counterparts. In this paper, we introduce energy-aware traffic grooming problems for optical networks. We show that by looking further into the modular physical architecture of a node during request allocations, we can significantly reduce the number of active components and, hence, total energy consumption in the network, especially when traffic load is low. Since energy usage is an important element of operational expenditure (OPEX), this approach provides the financial motivation for service providers along with the desired environmental motivation. We present auxiliary graph based heuristics and justify our cases compared to traditional approaches with simulation results.

We investigate the sequential task anycast scheduling problem (STAS) for grid applications in an optical burst switched network. In STAS, the subtasks of a given request have sequential dependency and must be processed in order. Each node in the network provides a set of services and each subtask may demand a different type of service. The problem is to find an ordered set of nodes that can minimize the request completion time and request blocking probability. In this work, we consider the delay caused by buffering at service nodes and the blocking caused by burst contention. We propose an auxiliary graph based heuristic with cost assignment policies to obtain ordered set of sequential nodes which can minimize task completion time. We compare the heuristics with a random task assignment policy, and the results show that the proposed heuristics can improve request blocking probability and end-to-end delay.

In this paper, we address the problem of survivable traffic grooming in optical WDM networks in which lightpaths are hop constrained. Survivability is provisioned at the wavelength granularity through either dedicated or shared path protection schemes. We propose an auxiliary-graph-based algorithm that addresses grooming, protection, and impairment constraints in a combined manner and that determines the placement of regenerators and grooming equipment in the network with the goal of minimizing equipment cost. Numerical results illustrate that the proposed algorithm outperforms an algorithm in which grooming, protection, and impairments are handled separately. We also evaluate effects of different equipment placement policies on the network cost and evaluate the cost-performance trade-offs for different network line rates.

This paper investigates the subject of reliability via two link-disjoint paths in mesh networks. We address the issues of how reliable two-path protection can be and how to achieve the maximum reliability. This work differs from traditional studies, such as MIN-SUM, MIN-MAX, and MIN-MIN, in that the objective in this paper is to maximize the reliability of the two-path connection given the link reliability, or equivalently, to minimize the end-to-end failure probability. We refer to this problem as MAX-REL. Solving MAX-REL provides 100% protection against a single failure while maximizing the reliability regardless of how many link failures occur in the network. We prove that this problem is NP-complete and derive a corresponding upper bound, which is the theoretical maximum reliability for a source-destination pair, and a lower bound, which is the worst case of the proposed algorithm. The time efficiency of the algorithms is analyzed, and the performance of the algorithms is evaluated through simulation. We demonstrate that our heuristic algorithms not only achieve a low computing complexity, but also achieve nearly equivalent performance to the upper bound.

Variable bandwidth circuit switching is an approach in which a variable amount of bandwidth is reserved over the duration of circuit, in contrast to the fixed amount of bandwidth reserved in traditional circuit switching and advance reservation. In this paper, we focus on the routing and scheduling problem in variable bandwidth circuit switching. We develop three heuristics and compare the performances of the proposed heuristics to an optimal algorithm, which minimizes the net file transfer time. While the optimal algorithm minimizes the net file transfer time for a single given request, it may end up allocating excessive resources in the network, leading to higher blocking and end-to-end delay for subsequent requests. The results demonstrate that in a dynamic traffic scenario, the proposed heuristics outperform the optimal algorithm in terms of request blocking, end-to-end delay, and throughput with significantly lower time complexity.

This paper studies the coexistence of several independent and dynamic wireless networks using the frequency hopping technique in the unlicensed radio band. We propose a new hopping scheme that allows more networks to collocate effectively, but does not violate federal restrictions regarding frequency constraint (related to the minimum number of frequencies in a hopping set) and time constraint (related to the maximum duration of using a particular frequency). The coexisting networks follow the Gaussian distribution in choosing transmission frequencies from parallel hopping spaces without the overhead of extra message exchange. Simulation results include comparison with other contemporary approaches and establish the viability of the proposed scheme.

In this paper, we address the dynamic multicast flow aggregation problem and propose a novel graph model with different aggregation policies. We first give the network model with an auxiliary graph for the aggregation. We then discuss an aggregation algorithm using the auxiliary graph and propose three policies for aggregating multicast flows. Numerical results show that aggregation can tremendously reduce the total trees in the network while the three policies can achieve good performance for various scenarios.

Standard advance reservation (SAR) mechanisms have been proposed as a way to provision resources for high-bandwidth applications in order to provide QoS guarantees to users; however, variations in the available bandwidth on different links over time can lead to situations in which standard advance reservations must be pushed further into the future, leading to higher end-to-end delays. We introduce the concept of time-shift advance reservation (TSAR), which improves bandwidth utilization by making use of delay elements at intermediate nodes within the network. We describe the TSAR architecture and discuss various design and implementation issues. The routing and scheduling problem in TSAR networks is defined, and a routing and scheduling algorithm is proposed. We find that TSAR with the proposed algorithm minimizes latency for each request and reduces request blocking compared to standard advance reservation and immediate reservation approaches.

Over the past several years, a significant amount of research has been conducted in the areas of optical packet switching (OPS) and optical burst switching (OBS). This research has been motivated by the need for techniques that are capable of supporting the demanding requirements of emerging dynamic high-bandwidth network applications in a flexible and efficient manner. Although optical packet and burst switching have yet to be widely deployed in commercial settings, recent research progress indicates that such deployments are not infeasible in the near future. The authors review the literature on OPS and OBS. Basic concepts are discussed and an overview of current and emerging research issues and challenges for optical packet and burst switched networks is presented. The authors attempt to cover issues that are essential for the practical deployment of such networks and highlight promising research directions that are likely to facilitate such deployments.

Delay-tolerant networking (DTN) is an architecture to enable data communications between isolated or remote regions, where long delays and intermittent connectivity can be tolerated. An emerging class of DTN, called vehicular DTNs (VDTN), exploits transportation systems as the transport layer to transfer data. In these networks, vehicles (e.g., busses, boats, trains) act as mobile nodes and carry data messages around. Mobile nodes can exchange data messages using devices called relay nodes. Relay nodes, placed in strategic positions along vehicle routes, have the capability to download, store, and upload the data messages from/to the mobile nodes. An important issue in VDTN is the optimal placement of the relay nodes such that delay-tolerant connectivity in VDTN is ensured at minimum cost. In this paper we show that the problem of optimal relay node placement is an NP-hard problem. Other contributions of this paper are the formulation of the relay node placement problem using ILP and the proposal of heuristic algorithms solving the optimization problem. Using simulation results, we compare the performance of each algorithm under different network constraints, such as node storage capability and network topology.

In this paper, we study the content protection problem through multicast flow aggregation (CPMFA) in which programs with protection requirement are aggregated into different light trees at the access routers and delivered from the head end to one or more destinations. We first use a tri-partite graph to describe the problem and show that the problem is NP-complete. We then formulate the problem as an interger linear program (ILP) and give results for small scale problems. Two hueristic appoaches, named two-step separated multicast flow aggregation (S-MFA) and mixed multicast flow aggregation (M-MFA), are presented. Numerical results show that S-MFA and M-MFA achieve good results in both large and small scale problems.

For a large-scale mesh network with dynamic traffic, maintaining the global state information in a centralized fashion is impractical. Hence, distributed schemes are needed to organize nodes and to manage state information in a more localized manner. One such effective scheme for organizing nodes is to cluster the nodes into a hierarchical structure. In this paper, we address the problem of determining the appropriate clustering of nodes for providing scalability in wavelength division multiplexed (WDM) optical networks with dynamic traffic. We present an on-line load-based (or bandwidth availability based) clustering technique that determines clusters adaptively in response to current network conditions. We also consider the problem of dynamic multicast on clustered networks with wavelength conversion capability. We introduce a heuristic using an auxiliary graph model to address routing, wavelength assignment, and traffic grooming jointly. Simulation results demonstrate the feasibility of our approach.

In this paper, we study the problem of multicast flow aggregation combined with light-tree establishment in optical networks. We first give the network model with a tri-partite aggregation graph to describe the problem and show that the problem is NP-complete. We then formulate the problem as an integer linear program (ILP) and give results for small scale problems. A heuristic approach, named fixed tree aggregation (FTA), is presented and results show that FTA achieves good results in both large and small scale problems.

In this paper, we focus on the problem of effectively supporting a large number of small group multicasts in optical burst switched (OBS) networks. We first propose a multicast scheme for small group multicast in OBS networks. To reduce burst loss due to potential burst contentions, a deflection routing scheme for multicast is proposed. We then develop an analytical model to evaluate the performance of the multicast scheme and the deflection routing scheme. To the best of our knowledge, this is the first analytical model proposed for the problem of multicast with deflection routing. The model is very general in the sense that it can handle unicast traffic, multicast traffic, and the mixture of unicast and multicast traffic, with or without deflection routing. The analytical model is verified through simulations. Numerical results show that the analytical model is accurate, and that, under low or medium network load, deflection routing for multicasting can significantly reduce burst loss while slightly increasing burst delay.

In this paper, we address both lifetime and survivability issues in wireless sensor networks (WSNs). We present an easy-to-implement clustering method called DED (distributed, energy-efficient, and dual-homed clustering) which promises robustness for WSNs without relying on the redundancy of sensors, i.e., without depending on node density. DED uses the information already gathered during clustering to determine additional alternate routes from sources to observers, thus incurring low message overhead. It does not make any assumptions about network dimension, node capacity, or location-awareness; and terminates in a constant number of iterations. The correctness of the algorithm is proved analytically. Simulation results demonstrate that our approach is effective both in prolonging the network lifetime and in providing survivability.

In this paper, we investigate the problem of survivable grooming for anycasting over wavelength division multiplexing (WDM) networks. Anycasting is a communication paradigm in which a source node sends data to any one out of multiple candidate destinations. In WDM networks, the standard anycasting problem involves the establishment of a connection from the source to any one destination with the objective of minimizing the resources used. In order to provide survivability, the problem can be extended to find both a connection to a primary destination and a connection to a backup destination. The connections, which may require an amount of bandwidth that may be less than the bandwidth of an entire wavelength, can be groomed with other connections in the WDM network. We propose an auxiliary graph approach to solving the survivable grooming problem for anycasting in WDM networks, and we propose three different protection approaches for providing different degrees of survivability against link failures and destination failures. The effectiveness of the proposed schemes is verified through simulation.

In this paper, we study the problem of time-constrained multicast flow aggregation in which programs with specified starting times and specified ending times are aggregated together and delivered from a head end to one or more destinations. We first use a four-partite graph to describe the problem and show that the problem is NP-complete. We then formulate the problem as an integer linear program (ILP) and give results for small scale problems. A heuristic approach, named time compatible first (TCF), is presented, and results show that TCF achieves good results in both large and small scale problems.

In this paper, we discuss for the first time the issue of supporting manycasting service over optical burst- switched (OBS) networks. One of the primary challenges in providing manycasting service over OBS networks is to reduce data loss due to burst contentions. We propose two new schemes, static over-provisioning (SOP) and dynamic membership (DM), to alleviate this data loss problem. The proposed schemes take into consideration the specific properties of manycasting, and the schemes may complement existing contention resolution schemes. The effectiveness of the proposed schemes is verified through simulation.

This paper investigates the path protection problem in mesh networks under multiple generic risks. Disjoint logical links may fail simultaneously if they share the same components in the physical layer. Each component is associated with one or more risks, and each risk may involve more than one logical link. This property introduces failure dependence for different logical links in layered networks. In this paper, the network failure scenario is under multiple risks. Each risk is a generic failure event, which may result in single-link failure, multiple-link failure, or all-link failure associated with a node (node failure). The objective aims at minimizing the end-to-end failure probability, or equivalently, maximizing the survivability by using two disjoint paths under multiple generic risks. The problem is shown to be NP-complete, and a novel graph model and corresponding heuristic algorithm are proposed. The time efficiency of the algorithms is analyzed, and the simulation results show that, compared to some existing diverse routing algorithms, the proposed heuristic algorithm can significantly improve the survivability at the cost of higher hop distance.

We define and investigate the problem of multi-resource manycast over optical burst switched (OBS) networks for supporting distributed computing applications. In multi-resource manycast, each destination has multiple computing resources, and each source generates requests that require multiple resources. For a given request, the problem is to select a set of destinations that have the required available resources and to And routes to these destinations. This problem differs from the traditional manycast problem in that different destinations have different resource availability. The objective is to minimize the resource blocking rate, which results from burst contention in the OBS network and from resource unavailability at the destinations. We investigate various approaches to implement multi-resources manycast over OBS networks, and verify the effectiveness of the proposed schemes through simulation.

In this paper, we focus on the problem of effectively supporting a large number of small group multicasts in optical burst switched (OBS) networks. We first propose a multicast scheme for small group multicast in OBS networks. To reduce burst loss due to potential burst contentions, a deflection routing scheme for multicast is proposed. We then develop a generic analytical model to evaluate the performance of the multicast scheme and the deflection routing scheme. To the best of our knowledge, this is the first analytical model proposed for the problem of multicast with deflection routing. The analytical model is verified through simulations. Numerical results show that the analytical model is accurate and that deflection routing can significantly reduce burst loss probability while slightly increasing burst delay.

With the growth of mobile users and the increasing deployment of wireless access network infrastructures, the issue of quality of service is becoming an important component of efficient wireless access network design. In this paper, we study the survivability problem for users that are connected to the core network by fully or partially dual homed paths, or by a single path. Given a hierarchical wireless access network with the available capacity and reliability at each level, the problem is to minimize overall connection cost for multiple requests such that the capacity and minimum survivability requirements are not violated. We formulate the problem using mixed integer linear programming and propose a genetic-algorithm-based heuristic.

This paper investigates the problem of finding two link-disjoint paths in mesh networks. Different from traditional studies, such as MIN-SUM, MIN-MAX, and MIN-MIN, the objective in this paper is to minimize the end-to-end failure probability, or equivalently, maximize the survivability under multiple link failures. We refer to this problem as MAX-SURV. Solving MAX-SURV provides 100% protection against single failure while maximizing the survivability under multiple failures. We prove that this problem is NP-complete, and we propose a heuristic algorithm for finding a solution. The corresponding upper bound and lower bound for given networks are also found. The time efficiency of the algorithms is analyzed, and the performance of the algorithms is evaluated through simulation.

Abstract — In this paper, we study the survivability problem in hierarchical wireless access networks with dual-homed end users, who are connected to two base stations (BSs), a primary BS and a backup BS. The dual homing mechanism is resilient to a single failure of a BS. However, if a failure occurs at the base station controller (BSC) layer or at the mobile switching center (MSC) layer, dual-homing may not prevent connection loss. We address the problem of routing from BSs to BSCs and from BSCs to MSCs, with the objective of minimizing the maximum number of connections lost due to a single failure of BS, BSC, or MSC. We first formulate the problem using Integer Linear Programming (ILP). We then prove that this optimization problem is NP-hard by showing some of its subproblems with relaxed constraints are still NP-hard. A Tabu Search (TS) based heuristic is then proposed for the problem, which provides near optimal results in most cases. I.

Optical burst switching (OBS) has been proposed as a promising switching technology to support the next-generation optical Internet. Two critical issues in OBS networks are designing effective contention resolution algorithms and implementing such algorithms in the core switch node. The first part of this paper deals with the problem of contention in OBS networks. We introduce a new contention resolution algorithm called Lookahead window Contention Resolution (LCR) that can support service differentiation. Through complexity analysis, a number of important properties of LCR and its variations are examined and verified. We also investigate the performance of LCR in a single core switch node. We show through simulation that the performance of LCR is competitive to existing contention resolution mechanisms in terms of reducing burst loss. The second part of this paper presents a practical design approach for a generic control packet processor architecture. Based on the proposed architecture, we implement the look-ahead contention resolution algorithm using a high-density field-programmable gate arrays (FPGA). Through hardware test and verification, we evaluate the performance characteristics of the look-ahead contention resolution algorithm in terms of hardware speed, scalability, and cost.

Packet contention is a major issue in asynchronous optical packet switching networks. Optical buffering, which is implemented by fiber delay lines (FDLs), is fundamental to many optical switch implementations for resolving contention. Most existing optical buffering implementations are output-based and require a huge amount of FDLs as well as larger switch sizes, which impose extra cost on the overall system. In this paper, we consider a shared optical buffering architecture which can reduce the buffer size at a switch. We propose an analytical model to evaluate the packet loss probability and the average delay For shared buffers at a single switch. We then compare the performance of output buffers to shared buffers under different granularities of FDLs. We observe that, by choosing an appropriate granularity, the shared buffering scheme can significantly reduce packet loss with much smaller switch sizes and fewer FDLs than the output buffering architecture. The accuracy of the analytical model is also confirmed by extensive simulation.

In this paper we address the problem of data burst grooming in optical burst-switched (OBS) networks. In OBS networks IP packets with the same edge node destination are assembled into larger packets called data bursts. Depending on the core node's switching technology, data bursts are required to have a minimum length. On the other hand, each IP packet in a burst has a time delay constraint, called maximum end-to-end delay, which determines the upper time limit before which the packet must reach its destination. Thus, a data burst cannot wait indefinitely until sufficient number of IP packets are assembled and the minimum burst length requirement is met. In order to satisfy the packet maximum end-to-end delay requirement, many bursts will be timed out and released before they reach the minimum length requirement. Under such circumstances, padding overhead must be added to these short bursts, called sub-bursts. Excessive padding results in high overhead and high data burst blocking probability. One approach to minimize the amount of padding overhead, while maintaining the end-to-end delay requirement of IP packets, is to groom multiple sub-bursts together. That is, sub-bursts with different destinations are aggregated together at the edge node and transmitted as a single burst until they are separated at some downstream node. In this paper we present an edge node architecture enabling burst grooming capability. We also develop two basic grooming approaches, namely no-routing-overhead (NoRO) and minimum-total-overhead (MinTO). Through a comprehensive simulation study we show that, in general, our proposed grooming algorithms can significantly improve the performance compared to the case of no grooming. However, careful considerations must be given to network loading condition and the number of sub-bursts allowed to be groomed together. We show that although simple greedy algorithms can reduce network overhead, they may alter the traffic characteristics and increase its burstiness, resulting in high packet blocking probability.

Dual homing is a fault-tolerance mechanism generally used in IP-based access networks to increase the survivability of the network. In a dual-homing architecture, a host is connected to two different access routers; therefore, it is unlikely that the host will be denied access to the network as the result of a failure in the access network, a failure of the access router, or congestion at the access router. However, dual homing cannot provide survivability with respect to possible failures in the optical core network. To provide survivability in the core network, optical protection and restoration techniques must be used. In the past, dual homing architectures and optical protection schemes have been studied independently of one another. This paper studies coordinated multi-layer survivability techniques that use both dual-homing schemes and optical protection schemes in an IP-based access network over a WDM-based optical core network. Specifically, we investigate the protection design problem in the WDM core network, given that a dual-homing infrastructure is implemented in the access network. Several solutions are proposed, and it is shown that the proposed coordinated survivability schemes can reduce cost compared to the case in which the survivability mechanisms arc not coordinated between the IP layer and the optical layer.

In this paper, we evaluate the performance of a burst retransmission scheme in which the bursts lost due to contentions in an OBS network are retransmitted at the OBS layer. The retransmission scheme aims to reduce burst loss probability in OBS networks. We develop an analytical model for obtaining the burst loss probability over an OBS network that uses the retransmission scheme. We also compare the performance of the burst retransmission scheme with the deflection scheme. Simulation results also show that at a moderate traffic load, the retransmission scheme provides an improvement of up to four times the burst loss probability with the deflection scheme. Results also show that the retransmission scheme significantly improves the burst loss probability compared to an OBS network without the retransmission scheme.

In this paper we describe a rate-based OBS network architecture in which core switch nodes send explicit messages to edge nodes requesting them to reduce their transmission rate on congested links. Within this framework, we introduce a new contention avoidance mechanism called proportional control algorithm with explicit reduction request (PCwER). Through source rate control, PCwER proactively attempts to prevent the network from entering the congestion state. Basic building blocks and performance trade-offs of PCwER are the main focus of this paper. In addition, through a simple fluid model we analyze the characteristics of the algorithm. Our simulation results show that the proposed contention avoidance techniques improve the network utilization and reduce the packet loss probability

Due to the bufferless nature of OBS networks, random burst losses may occur, even at low traffic loads. For optical burst-switched (OBS) networks in which TCP is implemented at a higher layer, these random burst losses may be mistakenly interpreted by the TCP layer as congestion in the network, leading to serious degradation of the TCP performance. In this paper, we reduce random burst losses by a burst retransmission scheme in which the bursts lost due to contention in the OBS network are retransmitted at the OBS layer. The OBS retransmission scheme can then reduce the probability that the TCP layer falsely detects congestion, thereby improving the TCP throughput. We analyze the TCP throughput when OBS networks employ the burst retransmission scheme and develop a simulation model to validate the analytical results. Based on our simulation results, we show that an OBS layer with burst retransmission provides an improvement of up to ten times the TCP throughput over an OBS layer without burst retransmission. This significant improvement is primarily because the TCP layer triggers fewer time-out based retransmissions when the OBS retransmission scheme is used

In this paper, we propose a novel proactive scheme, burst cloning, to reduce data loss due to burst contention in optical burst-switched (OBS) networks. The idea is to replicate a burst and send duplicated copies of the burst through the network simultaneously. If the original burst is lost, the cloned burst way still be able to reach the destination. Primary design issues in burst cloning are to select the optimal nodes at which to do cloning and to prevent cloned bursts from contending for resources with original bursts. An analytical model is developed to evaluate the proposed scheme. The model is verified through extensive simulations. We observe that burst cloning could significantly reduce data loss in OBS networks.

Packet contention is a major issue in asynchronous optical packet and burst switching networks. Optical buffering, which is implemented by fiber delay lines (FDLs), is fundamental to many optical switch implementations for resolving contention. Most existing optical buffering implementations are output-based and require a huge amount of FDLs as well as larger switch sizes, which impose extra cost on the overall system. In this paper, we consider shared optical buffering which can reduce the buffer size at a switch. Since no previous study is available to analyze the performance of asynchronous architectures with shared buffers, we propose an analytical model to evaluate the packet loss probability and the average delay for shared buffers at a single switch. We then compare the performance of output buffers to shared buffers under different granularities of FDLs. We observe that, by choosing an appropriate granularity, the shared buffering scheme can significantly reduce packet loss with much smaller switch sizes and fewer FDLs than the output buffering architecture. The accuracy of the analytical model is also confirmed by extensive simulation.

In this work, we consider the problem of maximizing the reliability of connections in mesh networks against failure scenarios in which multiple links may fail simultaneously. We consider the single-path connection problem as well as multiple-path (protected) connection problems. The problems are formulated as minimum-color path problems, where each link is associated with one or more colors, and each color corresponds to a given failure event Thus, when a certain color fails, all links which include that color will fail. In a single-path problem, by minimizing the number of colors on the path, the failure probability of the path can be minimized if all colors have the same probability of causing failures. In the case of two paths, where one path is a protection path, if all colors have the same probability of causing failures, the problem becomes that of finding two link-disjoint paths which either have a minimum total number of colors, or which have a minimum number of overlapping colors. By minimizing the total number of colors, the probability that a failure will occur on either of the paths is minimized. On the other hand, by minimizing the number of overlapping colors, the probability that a single failure event will cause both paths to fail simultaneously is minimized. The problems are proved to be NP-complete, and ILP formulations are developed. Heuristic algorithms are proposed for larger instances of the problems, and the heuristics are evaluated through simulation.

As high capacity all-optical networks and WDM technologies advance and merge together, aggregating low-speed traffic streams onto high-speed wavelengths becomes more critical. Efficient aggregation techniques, known as traffic grooming, allow higher bandwidth utilization and can reduce request blocking probability. These algorithms can also result in lower network cost in terms of electronic switching. In this paper we focus on traffic grooming in WDM mesh networks with dynamic traffic patterns. We offer two new grooming concepts called lightpath dropping and lightpath extension. These concepts are based on an alternative node architecture in which incoming optical signals can be dropped at a node, while optically continuing to the next node. Based on these concepts, we develop several grooming algorithms and study them under various network objectives. We also compare their performance with previously proposed lightpath-based grooming algorithms. Through extensive simulation results we show that our proposed approaches lead to lower request blocking probability and lower average number of logical hops when the number of transceivers per node is limited.

We address the traffic grooming problem in WDM mesh networks with dynamic unicast traffic. We develop a dynamic tree grooming algorithm (DTGA) that can support multi-hop traffic grooming by taking advantage of light-trees. In this algorithm, a light-tree can be dropped, branched, and extended when a route is to be established for a new request. In order to implement the DTGA, we develop a layered graph model which can support different routing policies. Extensive simulation shows that DTGA has better performance than lightpath-based algorithms when transceivers are limited.

In this paper, we provide a new analytical model for evaluating the blocking performance of dynamic lightpath establishment in multifiber wavelength-routed networks. By adopting the simple link-independent model together with the wavelength correlation assumptions, we manage to achieve a good balance between analytical accuracy and computational complexity. Extensive numerical results show that the proposed model can quickly produce accurate analytical results under different traffic loads and in different networks.

As the number wavelengths in WDM systems continue to increase, the switching fabric of optical cross connects (OXCs.) becomes increasingly complex. This complexity can be reduced by introducing multi-granular OXCs (MG-OXC) into the network. An MG-OXCs is capable of switching an entire group of wavelengths through a single switch port, thereby reducing the number of switching ports and reducing the complexity of the switch fabric. In this paper, we consider the problem of establishing active and backup paths for connections in networks with MG-OXCs. The problem of establishing protected connections for a given traffic demand in MG networks differs from the protection problem in standard wavelength-routed networks in that the primary objective in MG networks is to minimize the number of switch ports in the network. The mode of protection (dedicated or shared) affects the ability to aggregate and route traffic together. We propose a graph-based heuristic that attempts to solve the problem of routing and waveband assignment in an integrated manner. The heuristic reduces the total number of ports in the MG network. We also study the effect of shared protection on the port count.

Path protection requires finding a working path and a protection path that are link disjoint. We consider the dynamic lightpath protection problem in WDM mesh networks where a single risk factor may cause multiple links to fail simultaneously. The objective is to find link-disjoint lightpaths that are also risk disjoint. A similar problem has recently been proven to be NP-complete. We give an alternative proof of the NP-completeness, and formulate the problem as an integer linear program. We then develop heuristic algorithms and evaluate the performance of the algorithms through computer simulation. We show that we can achieve satisfactory performance using the heuristic techniques.

Path protection requires finding a working path and a protection path that are link disjoint. In this paper, we consider the dynamic lightpath protection problem in WDM mesh networks under the wavelength continuity constraint. Existing polynomial time algorithms can be applied to find a pair of link-disjoint lightpaths on a single wavelength; however, such algorithms fail if the working and protection lightpaths are on two different wavelengths. We prove the problem is NP-complete for both dedicated protection and shared protection. We develop an integer linear programming (ILP) formulation and heuristic solutions for the problem. Computer simulations are conducted to evaluate the performance of the heuristic algorithms.

Several schemes have been proposed recently in the literature for providing absolute QoS differentiation in OBS networks, such as early drop and wavelength grouping schemes. However, these schemes only provide loss guarantees at a per-hop level. In this paper, we propose a path clustering technique to implement these per-hop schemes over an entire network. The path clustering technique provides a solution to prioritizing the traffic based on hop-distances between source and destination pairs. We develop an analytical model for obtaining the optimal path clustering for a given network. By using the path clustering technique, we can improve the end-to-end loss performance of nonguaranteed traffic, and also provide absolute end-to-end loss probability of guaranteed traffic.

We propose an architecture and an algorithm for the realization of a pragmatic framework for optical packet transport. The architecture enables the transport of IP packets over optical frames in a network. While doing so, it relaxes the need for address recognition as well as for high speed switching, the two key hindering factors that have prevented contemporary optical packet transport solutions from being deployed. We propose the light-frames architecture for sub-lambda level provisioning of optical paths. The idea is to create a logical topology that allows N/sup 2/ connectivity using optical paths that are by themselves sublambda provisionable yielding in packet transport. We show that by using substantially fewer resources, we can provide N/sup 2/ connectivity in arbitrary graphs to support optical packet transport.

A fault-tolerant scheme, called dual homing, is generally used in IP-based access networks to increase the survivability of the network. However, dual homing itself cannot provide survivability with respect to possible failures in the wavelength division multiplexed (WDM) core network. To provide survivability in the core network, protection and restoration techniques must be used. In the past, dual homing architecture and protection are studied separately. This paper observes that the dual homing architecture introduces new issues for protection and restoration design, especially when providing survivability against two independent failures, one in the access network and the other in the core network. This paper provides an integrated solution and studies the protection design problem in the WDM core network, given a dual-homing infrastructure in the access network. Several algorithmic solutions are proposed, and performance of the solutions is compared.

In this paper, we propose a framework for providing differentiated contention resolution in photonic packet-switched networks by exploiting recirculation buffering and deflection routing. We develop an analytical model to evaluate the packet loss probability and the end-to-end delay for different buffering and deflection routing schemes, and we investigate the effectiveness of the control schemes in providing differentiated loss and delay. The accuracy of the analytical model is confirmed by simulation.

A number of schemes have been proposed for providing quality-of-service (QoS) differentiation in optical burst-switched (OBS) networks. Most existing schemes are based on a relative QoS model in which the service requirements for a given class of traffic are defined relative to the service requirements of another class of traffic. In this paper, we propose an absolute QoS model in OBS networks which ensures that the loss probability of the guaranteed traffic does not exceed a certain value. We describe two mechanisms for providing loss guarantees at OBS core nodes: an early dropping mechanism, which probabilistically drops the nonguaranteed traffic, and a wavelength grouping mechanism, which provisions necessary wavelengths for the guaranteed traffic. It is shown that integrating these two mechanisms outperforms the stand-alone schemes in providing loss guarantees, as well as reducing the loss experienced by the nonguaranteed traffic. We also discuss admission control and resource provisioning for OBS networks, and propose a path clustering technique to further improve the network-wide loss performance. We develop analytical loss models for the proposed schemes and verify the results by simulation.

Optical burst switching (OBS) has been proposed as a competitive hybrid switching technology to support the next-generation optical Internet. This paper addresses the problem of contention in OBS networks and introduces a new contention resolution algorithm called look-ahead window contention resolution (LCR) that can also support service differentiation. Simulation results show that the performance of LCR is competitive to existing contention resolution mechanisms in terms of reducing burst loss.

In optical burst-switched networks, data loss may occur when bursts contend for network resources. There have been several proposed solutions to resolve contentions in order to minimize loss. These localized contention resolution techniques react to contention, but do not address the more fundamental problem of congestion. Hence, there is a need for network level contention avoidance using load balanced routing techniques in order to minimize the loss. In this paper, we propose two dynamic congestion-based load balanced routing techniques to avoid congestion. Our simulation results show that the proposed contention avoidance techniques improve the network utilization and reduce the packet loss probability.

A number of schemes have been proposed to support QoS in optical burst-switched (OBS) networks. Most schemes only support relative QoS differentiation instead of absolute QoS guarantee. However, absolute QoS differentiation is a basic requirement for many delay and loss sensitive applications. In this paper, we propose two mechanisms for providing absolute QoS differentiation in OBS networks, an early drop mechanism, which selectively drops non-guaranteed traffic, and a wavelength grouping mechanism, which manages wavelengths for guaranteed traffic. We show that the combination of these two mechanisms not only outperforms other schemes in providing loss guarantees, but also improves the loss performance of non-guaranteed traffic.

In this work, we propose a new distributed signaling scheme, within the GMPLS framework for establishing lightpaths in wavelength-routed networks with sparse wavelength conversion. Analytical models are developed to evaluate the performance of the proposed scheme. Theoretical and simulation results show that compared to the classic schemed designed primarily for networks with no wavelength conversion, the proposed signaling scheme can achieve much lower blocking probability.

When establishing lightpaths in an all-optical DWDM network, it is possible that concurrent lightpaths requests will block one another if the lightpaths attempt to reserve the same wavelength of the same link. In this paper, we propose a novel signaling mechanism, referred to as label prioritization, which attempts to reduce the backward-link blocking in GMPLS-centric all-optical networks by assigning different priorities to the suggested wavelengths of each connection request. The prioritization of the wavelength encourages concurrent lightpath requests to choose different wavelengths, thereby reducing the possibility that the requests will be blocked. The label prioritization mechanism consists of a signaling extension to GMPLS to support the label prioritization and a modification in the optical switch controller to support the signaling extension. Simulation results show that the label prioritization method can effectively reduce wavelength conflicts.

Optical burst switching is a promising solution for terabit transmission of IP data bursts over WDM networks. One of the key components in the design of optical burst-switched nodes is the development of channel scheduling algorithms that can efficiently handle data burst contentions. Currently, traditional scheduling techniques use wavelength conversion and buffering to resolve burst contention. In this paper, we reduce packet losses by proposing a number of data channel scheduling algorithms that use burst segmentation and fiber delay lines (FDLs). The proposed scheduling algorithms are classified based on the placement of the FDL buffers in the optical burst-switched node and are referred to as delay-first or segment-first schemes. Simulation results show that these algorithms can effectively reduce the packet loss probability compared to existing scheduling techniques.

A problem of many distributed lightpath provisioning schemes is wavelength contention, which occurs when a connection request attempts to reserve a wavelength channel that is no longer available. This situation results from the lack of updated global link-state information at every node. In networks with highly dynamic traffic loads, wavelength contention may seriously degrade the network performance. To overcome this problem, we propose a new framework for distributed signaling and introduce a class of schemes referred to as intermediate-node initiated reservation. In the new scheme, reservations may be initiated at any set of nodes along the route; in contrast, reservations can only be initiated by the destination node in the classic destination initiated reservation (DIR) scheme. As a result, the possibility of having outdated information due to propagation delay is significantly lowered. Specifically, we consider two schemes within this framework, for networks with no wavelength conversion and for networks with sparse wavelength conversion, respectively. Theoretical and simulation results show that, compared with the classic DIR scheme, the new schemes can significantly improve the network blocking performance. The accuracy of the analytical models is also confirmed by extensive numerical simulations.

We address the issue of providing quality-of-service (QoS) in an optical burst-switched network. QoS is provided by introducing prioritized contention resolution policies in the network core and a composite burst-assembly technique at the network edge. In the core, contention is resolved through prioritized burst segmentation and prioritized deflection. The burst segmentation scheme allows high-priority bursts to preempt low-priority bursts and enables full class isolation between bursts of different priorities. At the edge of the network, a composite burst-assembly technique combines packets of different classes into the same burst, placing lower class packets toward the tail of the burst. By implementing burst segmentation in the core, packets that are placed at the tail of the burst are more likely to be dropped than packets that are placed at the head of the burst. The proposed schemes are evaluated through analysis and simulation, and it is shown that significant differentiation with regard to packet loss and delay can be achieved.

We address the issue of providing differentiated services to IP packets over an optical burst switched core network, and we introduce a new approach for assembling packets into a burst. In this technique, a composite burst is created by combining packets of different classes into the same burst. The packets are placed from the head of the burst to the tail of the burst in order of decreasing class. The performance of this approach is enhanced by using a burst segmentation technique in which, during burst contention, only the packets in the tail of a burst are dropped. We describe a generalized model for burst assembly and burst scheduling, and we propose several composite burst assembly methods. We observe that having multiple classes of packets in a burst performs better than having a single class of packets in a burst.

We present a label-based approach for implementing deflection in a photonic packet network, and we introduce an algorithm for determining deflection options in a manner which eliminates looping. A general analytical model is developed to evaluate packet loss probabilities in networks with deflection. The analysis may be applied to a wide class of deflection schemes and may be applied to networks with any arbitrary topology. The analysis is verified through simulation.

We address the issue of contention resolution in optical burst switched networks, and we introduce an approach for reducing packet losses which is based on the concept of burst segmentation. In burst segmentation, rather than dropping the entire burst during contention, the burst may be broken into multiple segments, and only the overlapping segments are dropped. The segmentation scheme is investigated by simulation in conjunction with a deflection scheme, and it is shown that segmentation with deflection can achieve a significantly reduced packet loss rate.

We develop a model for evaluating the blocking probability of various connection management protocols for wavelength-routed optical networks with dynamic lightpath establishment. The model characterizes both blocking due to insufficient resources and blocking due to multiple interfering connection requests. We then use the analytical model to compare two connection management schemes, one which utilizes source-initiated reservation, and another which utilizes destination-initiated reservation.

For a wavelength-routed network in which connection requests are arriving and departing at high rates, an appropriate control scheme must be implemented to set up light paths for each request in a fast and efficient manner. The control scheme, which includes routing and wavelength assignment algorithms, must also be scalable, and should attempt to minimize the number of blocked connections. In this paper, we consider a distributed control scheme which utilizes a new adaptive routing approach called alternate-link routing. In the proposed approach, routing decisions for a light path are made adaptively on a hop-by-hop basis by individual nodes in a distributed manner. The scheme does not require the maintenance of any global information. A simulation is developed to analyze blocking performance, and it is shown that the proposed approach outperforms fixed routing and, under certain conditions, also outperforms fixed alternate-path routing.

Wavelength-division multiplexing (WDM) local-area networks based on the optical passive-star coupler have traditionally been classified as being either single-hop or multihop. A single-hop network provides a direct connection between the source and the destination of a packet during the packet transfer duration, but may require some amount of coordination between the nodes which may involve tuning of the transmitters or receivers at each node. Since the time required to tune a tunable optical transmitter or receiver may be high, a single-hop network may incur significant overhead. On the other hand, a typical multihop network requires little or no tuning, but a packet may traverse a number of intermediate nodes between the source and destination nodes. Each hop incurs additional queueing delays at each node and also increases the overall load on each link and on the network. In this paper, we propose a new class of multiconfiguration multihop protocols (MMPs) which use tunable transmitters and receivers to cycle through a number of configurations which together make up a multihop logical topology. This class of protocols offers a trade-off between the tuning required in a single-hop network and the number of hops required in a multihop network. We present a generalized framework for comparing the proposed protocols with existing single-hop and multihop protocols, and we show that these protocols may offer significant performance gains for systems with high tuning delays and a limited number of transmitters and receivers at each node.

Node architectures which are being considered for wavelength-routed WDM optical networks do not provide a high degree of scalability with respect to the number of wavelengths in the system. In this investigation, we propose a new node architecture which allows new wavelengths to be added to the network without significant additional costs. We study the performance of such nodes in a network environment, and we develop a simple cost model for the network. We show that the proposed architecture provides a higher degree of scalability as well as significant cost benefits when compared to a traditional node architecture, such as the wavelength-routing switch (WRS) or a wavelength-selective cross-connect (WSXC), while also maintaining comparable performance in terms of blocking probability.

Wavelength-division multiplexing (WDM) local-area networks based on the optical passive-star coupler have traditionally been classified as being either single-hop or multihop. A single-hop network provides a direct connection between the source and the destination of a packet during the packet transfer duration, but may require some amount of coordination between the nodes which may involve tuning of the transmitters or receivers at each node. Since the time required to tune a tunable optical transmitter or receiver may be high, a single-hop network may incur significant overhead. On the other hand, a typical multihop network requires little or no tuning, but a packet may traverse a number of intermediate nodes between the source and destination nodes. Each hop incurs additional queuing delays at each node and also increases the overall load on each link and on the network. We propose a new class of multiconfiguration multihop protocols (MMPs) which use tunable transmitters and receivers to cycle through a number of configurations which together make up. A multihop logical topology. This class of protocols offers a trade-off between the tuning required in a single-hop network and the number of hops required in a multihop network. We present a generalized framework for comparing the proposed protocols with existing single-hop and multihop protocols, and we show that these protocols may offer significant performance gains for systems with high tuning delays and a limited number of transmitters and receivers at each node.

Mobility support in IP networks requires the use of servers to forward packets to mobile hosts and to maintain information pertaining to a mobile host's location in the network. In one proposed protocol, the mobile IP protocol, location and packet forwarding functions are provided by servers referred to as home agents. These home agents may become the bottleneck when there are a large number of mobile hosts in the network. In this paper, we consider the design and analysis of load balancing mechanisms for multiple home agents in the mobile IP protocol. We propose a load balancing scheme in which a home agent may periodically transfer the control of a mobile host to another home agent in the same network through the use of functions supported in mobile IP. The periodicity with which this transfer is performed affects the load balancing gain as well as the associated overhead. We analyze our load balancing mechanism under bursty traffic arrival conditions using a Markov modulated Poisson process. The results show that the proposed load balancing scheme can yield modest gains over alternative load balancing strategies.

In a single-hop WDM optical network, a straightforward approach to implementing multicasting is to schedule a single transmission to multiple destinations so that all of the destinations' receivers must tune to the same channel at the same time. Although scheduling a single transmission in this manner reduces the amount of transmitter and channel resources being used, it may also place a burden on the receivers in the network. If all receivers do not become available at the same time, then some receivers may have to wait (and be idle) for significantly long periods of time before receiving the message. In this paper, we investigate methods for partitioning a multicast group into a number of smaller subgroups and for scheduling a separate transmission for each of these subgroups. We show that this approach more effectively conserves and balances the usage of transmitter and receiver resources in the network and may lead to significantly improved system performance over the conventional single-transmission multicast approach.

Rainbow is a prototype optical metropolitan area network (MAN) developed at IBM. It employs wavelength-division multiplexing (WDM) on a fiber-optic, passive-star network topology, with each station equipped with a laser, which is fixed tuned to its own unique wavelength, and a Fabry-Perot filter, which is tunable across all wavelengths. This paper presents a model and analysis of the protocol used in the Rainbow prototype using the equilibrium point analysis (EPA) technique. We examine the system's throughput and how it is affected by various system parameters such as message arrival rate, message length, and timeout duration. We show that, for a given arrival rate, there is a timeout duration that will yield the optimal throughput. The analytical results are verified by simulation.