Machine learning (ML) models, data and software need to be regularly updated whenever essential version updates are released and feasible for integration. This is a basic but most challenging requirement to satisfy in the edge, due to the various system constraints and the major impact that an update can have on robustness and stability. In this paper, we formulate for the first time the ML model versioning optimization problem, and propose effective solutions, including the update automation with reinforcement learning (RL) based algorithm. We study the edge network environment due to the known constraints in performance, response time, security, and reliability, which make updates especially challenging. The performance study shows that model version updates can be fully and effectively automated with reinforcement learning method. We show that for every range of server load values, the proper versioning can be found that improves security, reliability and/or ML model accuracy, while assuring a comparably lower response time.

Efficient topology management in future 6G networks is a fundamental challenge for dynamic network creation based on location services, where each autonomous sub-network can be tailored to specific application scenarios. This paper studies the performance of a novel topology change management system in a 6G network dynamically organized into autonomous sub-networks. We propose and analyze an algorithm for intelligent prediction of topology changes and compare it with a monitoring-based approach. A case study on edge video distribution, aligned with 3GPP and ETSI MEC (Multi-access Edge Computing) standards, demonstrates the system's practical relevance. The proposed topology change prediction algorithm optimizes and selects the best machine learning models based on the scenario under study. For link change scenario, the results show that ANN demonstrates the best performance in identifying cases with no changes, slightly outperforming random forest and XGBoost. For user mobility scenario, XGBoost is more efficient in learning patterns for topology change prediction. In terms of cost efficiency, our ML-based approach represents a significantly cost-effective alternative to traditional monitoring approaches.

Distributed applications based on micro-services in edge computing are becoming increasingly popular. Kubernetes is the default framework for orchestrating and managing micro-service-based applications. Notwithstanding, the requirement to run applications between multiple sites at cloud and edge poses new challenges since Kubernetes does not natively provide tools to abstract inter-cluster communications at the application level. In this paper, we evaluate for the first time the impact of inter-cluster communication on edge computing performance by using three prominent, open-source inter-cluster communication projects and tools (Submariner, ClusterLink, and Skupper). We develop a fully open-source testbed that integrates these tools modularly and experimentally benchmark sample applications on their performance running in a multi-cluster edge computing system under varying networking conditions. We experimentally analyze two classes of envisioned mobile applications (industrial automation and vehicle decision drive assist). Our results show that ClusterLink performs best out of the three tools in scenarios with increased payloads regardless of the underlying networking conditions or transmission direction. Skupper closely follows it unless requests and replies transport big payloads. Finally, for small payloads, Submariner slightly outperforms the other tools.

We consider the problem of signaling rate and performance for control and management of reconfigurable intelligent surfaces (RISs) in next-generation mobile networks. To this end, we first analytically determine the rates of RIS reconfigurations and handover using a stochastic geometry network model. We derive closed-form expressions of these rates, while taking into account static obstacles (both known and unknown), self-blockage, RIS location density, and variations in the angle and direction of user mobility. Based on the derived rates, we analyze the signaling rates of a sample novel signaling protocol, which we propose as an extension of the current handover signaling protocol. We evaluate the signaling overhead due to RIS reconfigurations and the related energy consumption. We also provide a capacity planning analysis of the related RIS control plane server for its dimensioning in the network management system. The results quantify the impact of known and unknown obstacles on the RIS reconfiguration rate and the handover rate as a function of device density and mobility. We evaluate the scalability of the model, the related signaling overhead, energy efficiency, and server capacity in the control plane. To the best of our knowledge, this is the first analytical model to derive the closed form expressions of RIS reconfiguration rates, along with handover rates, and relate its statistical properties to the signaling rate and performance in next-generation mobile networks.

With the increasing adoption of IoT devices and applications, significant research and development efforts have been centered around engineering novel ecosystems referred to as the IoT-edge-cloud compute continuum. In this article, we implement, analyze and present a case study for performance benchmarking of five well known and select open source MQTT broker implementations in an open-source compute continuum testbed. The proposed MQTT broker implementations are evaluated in terms of response time, different payload sizes and throughput. Measurements and results show that the hardware platform used, the message size, as well as the network parameters (latency, packet loss and jitter) have a significant impact on the resulting performance of various broker implementations and therefore have to be carefully considered in the selection process for the building blocks of the continuum. All implementations and measurements are made to be fully reproducible and free and open source.

The next-generation mobile networks follow the 3GPP disaggregated Radio Access Network (RAN) model, composed of three separate functional units: Centralized Unit (CU), Distributed Unit (DU), and Radio Unit (RU), and its interconnection network infrastructures, front-haul, between RU and DU and mid-haul, between DU and CU. The current focus has been extensively on the front-haul, proposing a myriad of networking technologies and resource allocation solutions. Mid-haul transport infrastructure has received comparably less attention. This paper explores the implementation and performance of optical multi-flow transponders in mobile mid-haul networks. It evaluates the performance of mid-haul mesh topologies with different DU placement methods. Numerical results show the achievable throughput and blocking performance. We also derive the number of Multi-flow optical transponders (MOTP) units deployed, as well as the related trade-offs.

One of the key factors critical to the advancements of IoT systems in remote areas today are energy-efficient IoT deployment and the integration with IoT /edge/continuum. An energy-efficient IoT deployment requires finding adequate solutions for applications that require remote area devices and the related replacement and charging of batteries. On the other hand, an efficient integration of different communication technologies spanning the IoT, edge and cloud continuum that at the same time can integrate energy harvesting devices in remote areas is still an open challenge. In this paper, we integrate energy harvesting with wearable remote IoT devices on freely roaming farm animals within the edge/cloud continuum along its powerful application layer protocols, MQTT and AMQP. We experimentally investigate the performance of kinetic energy harvester used to power a LoRa module to send application layer messages from IoT to cloud. From the functional system testing perspective, we show that these messages can be successfully forwarded for further processing and evaluation in the edge and cloud setting even from the remote areas. We engineered an inexpensive and first open-source multi-protocol MQTT based communication gateway, as an alternative to today's proprietary and expensive gateway solutions, and we built a system that can not only power the capturing of animal movement patterns outdoors, but also the related application-layer protocol messages.

Campus networks are expected to include various new features compared to 4G and 5G networks, such as the exploitation of THz frequency bands and real-time network optimization. Furthermore, to enable future-proof network control, the Digital Twin Network (DTN) reference architecture can be exploited to create an autonomic network system that provides intent-based interfaces and offers a complete adaptive network control tailored for campus network applications. This paper discusses such architecture and presents a case study focused on a THz system that can adapt its coding and modulation parameters in response to channel degradation due to environmental changes. Finally, simulation results evince the advantages of the proposed architecture.

Future 6G networks are expected to heavily utilize machine learning capabilities in a wide variety of applications with features and benefits for both, the end user and the provider. While the options for utilizing these technologies are almost endless, from the perspective of network architecture and standardized service, the deployment decisions on where to execute the AI-tasks are critical, especially when considering the dynamic and heterogeneous nature of processing and connectivity capability of 6G networks. On the other hand, conceptual and standardization work is still in its infancy, as to how to categorize ML applications in 6G landscapes; some of them are part of network management functions, some target the inference itself, while many others emphasize model training. It is likely that future mobile services may all be in the AI domain, or combined with AI. This work makes a case for the serverless computing paradigm to be used to this end. We first provide an overview of different machine learning applications that are expected to be relevant in 6G networks. We then create a set of general requirements for software engineering solutions executing these workloads from them and propose and implement a high-level edge-focused architecture to execute such tasks. We then map the ML-serverless paradigm to the case study of 6G architecture and test the resulting performance experimentally for a machine learning application against a setup created in a more traditional, cloud-based manner. Our results show that, while there is a tradeoff in predictability of the response times and the accuracy, the achieved median accuracy in a 6G setup remains the same, while the median response time decreases by around 25% compared to the cloud setup.

Network Function Virtualization (NFV) is a promising technology that can transform how internet service providers deliver their services. However, recent studies have identified several challenges in adopting NFV. Two key challenges are central to the operation and capacity planning of NFV Data Centers (DCs): (i) Service Function Chain (SFC) realization —determining if a new request with a known profile can be accommodated—and (ii) Network Function Virtualization (NFV) traffic bounds estimation —estimating the total traffic that a data center can handle considering all service requests and their performance constraints. To address these challenges, we propose a model that leverages stochastic network calculus to effectively dimension an NFV DC while ensuring delay and availability bounds for all service requests. Our theoretical model provides a mathematical framework to assess the realization of a single SFC request without delving into the specifics of the realization process. We utilize established availability-aware Virtual Network Function (VNF) placement patterns to obtain traffic bounds essential to planning data center capacity. We analyze NFV data center traffic under various scenarios over a Fat-tree DC topology. The results demonstrate that data center capacity is significantly influenced by the VNF placement strategy. Additionally, for data centers hosting latency-sensitive services, Service Level Objective (SLO) constraints on availability and delay are crucial in determining the number of such requests that can be accommodated.

With rapid advances in containerization techniques, the serverless computing model is becoming a valid candidate execution model in edge networking, similar to the widely used cloud model for applications that are stateless, single purpose and event-driven, and in particular for delay-sensitive applications. One of the cloud serverless processes, i.e., the auto-scaling mechanism, cannot be however directly applied at the edge, due to the distributed nature of edge nodes, the difficulty of optimal resource allocation, and the delay sensitivity of workloads. We propose a solution to the auto-scaling problem by applying reinforcement learning (RL) approach to solving problem of efficient scaling and resource allocation of serverless functions in edge networks. We compare RL and Deep RL algorithms with empirical, monitoring-based heuristics, considering delay-sensitive applications. The simulation results shows that RL al-gorithm outperforms the standard, monitoring-based algorithms in terms of total delay of function requests, while achieving an improvement in delay performance by up to 50%.

Spatially multiplexed networks have become one of the prime technology candidates to accommodate the unstoppable network traffic increase. Currently, two core-switching approaches are under consideration in the research community: full and restricted. In full core-switching optical network, the switching flexibility is maximal with the full port capacity and comparably larger equipment cost. In restricted switching, on the other hand, the equipment is simpler and less costly, but the wavelength circuit needs to remain in the same core along the route. We propose the study a combination of the two, which we refer to as sparse core-switching as tradeoff between the core-switching capabilities and the cost of capacity applied to the network. We are especially interested in physical layer aspects in the context of crosstalk and the resulting spectrum fragmentation and in understanding how crosstalk impact the performance of sparse core switching. Numerical results show that with the proper application of a fraction of the equipment budget required for the full-switching scenario, sparse core switching reaches an average performance gain of 31.34% and a cost reduction of 71.2 % and thus can even outperform the full-switching scenario.

Serverless computing has emerged as a new execution model which gained a lot of attention in cloud computing thanks to the latest advances in containerization technologies. Recently, serverless has been adopted at the edge, where it can help overcome heterogeneity issues, constrained nature and dynamicity of edge devices. Due to the distributed nature of edge devices, however, the scaling of serverless functions presents a major challenge. We address this challenge by studying the optimality of serverless function scaling. To this end, we propose Semi-Markov Decision Process-based (SMDP) theoretical model, which yields optimal solutions by solving the serverless function scaling problem as a decision making problem. We compare the SMDP solution with practical, monitoring-based heuristics. We show that SMDP can be effectively used in edge computing networks, and in combination with monitoring-based approaches also in real-world implementations.

Recently, edge computing has demonstrated increasing potential to provide low-latency computing services. Coded edge computing can not only make full use of the resources of heterogeneous edge computing servers, but also significantly reduce the negative effects of slow computing devices on computing time. Nevertheless, since edge servers may be unreliable or untrustworthy, the user will decode and get incorrect computation results even if it uses one incorrect sub-computation result returned by faulty edge servers. In this paper, for the existing coded edge computing schemes, we focus on the distributed matrix-matrix multiplication and design a general and efficientDecode-and-Compare Verification(DCV) scheme to verify the correctness of computation results and identify faulty edge servers by utilizing the properties of coded computing itself. The DCV scheme contains two components: (1) computation result verification,i.e., obtain the computation result and verify its correctness, and (2) faulty edge server identification,i.e., identify the faulty edge servers by verifying the correctness of returned sub-computation results. For both the independent and collusion faulty edge server models, we conduct solid theoretical analyses on the required decoding rounds, the coding redundancy and the successful verification probability to demonstrate that the correct computation result can be efficiently verified. We also conduct a lot of experiments on the DCV scheme from different aspects and the results show that it achieves much less computation time to get the correct computation result compared with other potential schemes, including homomorphic encryption and local computation.

Future smart factories are expected to deploy emerging dynamic Virtual Reality (VR) applications with high bandwidth wireless connections in the THz communication bands, where a factory worker can follow activities through 360° video streams with high quality resolution. THz communications are however known for low fault tolerance and high sensitivity to external factors, resulting in low throughput. To address this challenge, a system is needed that can adaptively react to changing channel states by configuring THz transceivers in terms of coding and modulation. We propose a control plane that can help us configure the THz communication system adaptively. The control plane implements a workflow algorithm to choose between various coding and modulation schemes depending on THz channel states. The results show that an adaptive control plane can improve throughput and signal resolution quality, with theoretically zeroed bit error probability and a maximum achievable throughput in the scenarios analyzed.

Migration and replication of virtual network functions (VNFs) are well-known mechanisms to face dynamic resource requests in Internet Service Provider (ISP) edge networks. They are not only used to reallocate resources in carrier networks, but in case of excessive traffic churns also to offloading VNFs to third party cloud providers. We propose to study how traffic forecasting can help to reduce the number of required migrations and replications when the traffic dynamically changes in the network. We analyze and compare three scenarios for the VNF migrations and replications based on: (i) the current observed traffic demands only, (ii) specific maximum traffic demand value observed in the past, or (iii) predictive traffic values. For the prediction of traffic demand values, we use an LSTM model which is proven to be one of the most accurate methods in time series forecasting problems. Based the traffic prediction model, we then use a Mixed-Integer Linear Programming (MILP) model as well as a greedy algorithm to solve this optimization problem that considers migrations and replications of VNFs. The results show that LSTM-based traffic prediction can reduce the number of migrations up to 45\% when there is enough available resources to allocate replicas, while less cloud-based offloading is required compared to overprovisioning.

In recent years, edge computing has attracted significant attention because it can effectively support many delay-sensitive applications. Despite such a salient feature, edge computing also faces many challenges, especially for efficiency and security, because edge devices are usually heterogeneous and may be untrustworthy. To address these challenges, we propose a unified framework to provide efficiency and confidentiality by coded distributed computing. Within the proposed framework, we use matrix multiplication, a fundamental building block of many distributed machine learning algorithms, as the representative computation task. To minimize resource consumption while achieving information-theoretic security, we investigate two highly-coupled problems, (1) task allocation that assigns data blocks in a computing task to edge devices and (2) linear code design that generates data blocks by encoding the original data with random information. Specifically, we first theoretically analyze the necessary conditions for the optimal solution. Based on the theoretical analysis, we develop an efficienttask allocationalgorithm to obtain a set of selected edge devices and the number of coded vectors allocated to them. Using the task allocation results, we then designsecure coded computingschemes, for two cases, (1) with redundant computation and (2) without redundant computation, all of which satisfy the availability and security conditions. Moreover, we also theoretically analyze the optimization of the proposed scheme. Finally, we conduct extensive simulation experiments to demonstrate the effectiveness of the proposed schemes.

Thanks to the latest advances in containerization, the serverless edge computing model is becoming close to reality. Serverless at the edge is expected to enable low latency applications with fast autoscaling mechanisms, all running on heteroge-neous and resource-constrained devices. In this work, we engineer and experimentally benchmark a serverless edge computing system architecture. We deploy a decentralized edge computing platform for serverless applications providing processing, storage, and communication capabilities using only open-source software, running over heterogeneous resources (e.g., virtual machines, Raspberry Pis, or bare metal servers, etc). To achieve that, we provision an overlay-network based on Nebula network agnostic technology, running over private or public networks, and use K3s to provide hardware abstraction. We benchmark the system in terms of response times, throughput and scalability using different hardware devices connected through the public Internet. The results show that while serverless is feasible on heterogeneous devices showing a good performance on constrained devices, such as Raspberry Pis, the lack of support when determining computational power and network characterization leaves much room for imopovement in edge environments.

The evolution towards Industry 4.0 is driving the need for innovative solutions in the area of network management, considering the complex, dynamic and heterogeneous nature of ICT supply chains. To this end, Intent-Based networking (IBN) which is already proven to evolve how network management is driven today, can be implemented as a solution to facilitate the management of large ICT supply chains. In this paper, we first present a comparison of the main architectural components of typical IBN systems and, then, we study the key engineering requirements when integrating IBN with ICT supply chain network systems while considering AI methods. We also propose a general architecture design that enables intent translation of ICT supply chain specifications into lower level policies, to finally show an example of how the access control is performed in a modeled ICT supply chain system.

The most recent development in the underlying QUIC transport protocol has revived the interest in HTTP over QUIC (HTTP/3) as a communication protocol solution in cloud and edge computing. This development is notable especially given the rise in implementations of IoT edge/cloud continuum, a new computing paradigm that extends the cloud computing IoT via edge computing systems. In IoT cloud/edge continuum, benchmarking performance of the communication protocol of choice is critical. We focus for the first time on experimental benchmarking of HTTP/3 performance in an IoT cloud/edge continuum system. To benchmark the performance in terms of scalability and latency, we first implement HTTP/3 from a set of non-standardized open-source libraries, - which is challenge, and use the implementation in two main IoT scenarios: edge computing only (IoT devices running HTTP/3 clients and edge devices running an HTTP/3 server), and cloud/edge continuum (based on Google Firebase). Experimental results show that latency and scalability remain a challenge for HTTP/3, but its appeal also remains, in its wide adoption, embedded security and compatibility with existing networked systems.

With the rapid advances in IoT, edge and cloud computing solutions, it is critical to educate and train students in computer science and engineering in various aspects of IoT-edge-cloud (IoT-E-C) system architecture implementations. We outline the design and development of an undergraduate laboratory course that sets the goal of implementing various interfaces and communication protocols to connect IoT, edge and cloud computing systems and evaluating their performance. The lab setup is modular and based on open source tools. In the IoT context, it consists of low-cost processing platforms with various sensors and actuators. In the edge and cloud computing context, we implement and deploy single board computers and Firebase cloud solutions, respectively. The modular lab setup allows students to engineer and integrate various communication protocol solutions, including MQTT, COAP and HTTP. In addition to the system implementation, students can evaluate and benchmark the performance of the entire system.

While edge computing is envisioned to superbly serve latency sensitive applications, the implementation-based studies benchmarking its performance are few and far between. To address this gap, we engineer a modular edge cloud computing system architecture that is built on latest advances in containerization techniques, including Kafka, for data streaming, Docker, as application platform, and Firebase Cloud, as realtime database system. We benchmark the performance of the system in terms of scalability, resource utilization and latency by comparing three scenarios: cloud-only, edge-only and combined edge-cloud. The measurements show that edge-only solution outperforms other scenarios only when deployed with data located at one edge only, i.e., without edge computing wide data synchronization. In case of applications requiring data synchronization through the cloud, edge-cloud scales around a factor 10 times better than cloudonly, until certain number of concurrent users in the system, and above this point, cloud-only scales better. In terms of resource utilization, we observe that whereas the mean utilization increases linearly with the number of user requests, the maximum values for the memory and the network I/O heavily increase when with an increasing amount of data.

REST HTTP is the communication protocol of choice for software developers today. In IoT systems with unreliable connectivity, however, a stateless protocol like REST HTTP needs to send a request message multiple times, and it only stops the retransmissions when an acknowledgement arrives at the sender. In our previous work, we studied the usage of random linear network coding (RLNC) for REST HTTP protocol to reducing the amount of unnecessarily retransmissions. In this paper, we experimentally validate the study and analyze REST HTTP with and without RLNC in a simple testbed in dynamic IoT systems. The measurements show notable improvements in bandwidth utilization in terms of reducing the retransmissions and delay when using network-coded REST HTTP.

Edge computing is a promising technology that can fulfill the requirements of latency-critical and computation-intensive applications. To further enhance the performance, coded edge computing has emerged because it can optimally utilize edge devices to speed up the computation. In this paper, we tackle a major security issue in coded edge computing: how to verify the correctness of results and identify attackers. Specifically, we propose an efficient verification scheme, namely Decode-and-Compare (DC), by leveraging both the coding redundancy of edge devices and the properties of linear coding itself. To design the DC scheme, we conduct a solid theoretical analysis to show the required coding redundancy, the expected number of decoding operations, and the tradeoff between them. To evaluate the performance of DC, we conduct extensive simulation experiments and the results confirm that the DC scheme can outperform existing solutions, such as homomorphic encryption and computing locally at the user device.

Optical networks are prone to power jamming attacks intending service disruption. This paper presents a Machine Learning (ML) framework for detection and prevention of jamming attacks in optical networks. We evaluate various ML classifiers for detecting out-of-band jamming attacks with varying intensities. Numerical results show that artificial neural network is the fastest (106 detection per second) for inference and most accurate (≈ 100%) in detecting power jamming attacks as well as identifying the optical channels attacked. We also discuss and study a novel prevention mechanism when the system is under active jamming attacks. For this scenario, we propose a novel resource reallocation scheme that utilizes the statistical information of attack detection accuracy to lower the probability of successful jamming of lightpaths while minimizing lightpaths' reallocations. Simulation results show that the likelihood of jamming a lightpath reduces with increasing detection accuracy, and localization reduces the number of reallocations required.

With the development of new system solutions that integrate traditional cloud computing with the edge/fog computing paradigm, dynamic optimization of service execution has become a challenge due to the edge computing resources being more distributed and dynamic. How to optimize the execution to provide Quality of Service (QoS) in edge computing depends on both the system architecture and the resource allocation algorithms in place. We design and develop a QoS provider mechanism, as an integral component of a fog-to-cloud system, to work in dynamic scenarios by using deep reinforcement learning. We choose reinforcement learning since it is particularly well suited for solving problems in dynamic and adaptive environments where the decision process needs to be frequently updated. We specifically use a Deep Q-learning algorithm that optimizes QoS by identifying and blocking devices that potentially cause service disruption due to dynamicity. We compare the reinforcement learning based solution with state-of-the-art heuristics that use telemetry data, and analyze pros and cons.

In recent years, edge computing has attracted increasing attention for its capability of facilitating delay-sensitive applications. In the implementation of edge computing, however, data confidentiality has been raised as a major concern because edge devices may be untrustable. In this paper, we propose a design of secure and efficient edge computing by linear coding. In general, linear coding can achieve data confidentiality by adding random information to the original data before they are distributed to edge devices. To this end, it is important to carefully design code such that the user can successfully decode the final result while achieving security requirements. Meanwhile, task allocation, which selects a set of edge devices to participate in a computation task, affects not only the total resource consumption, including computation, storage, and communication, but also coding design. In this paper, we study task allocation and coding design, two highly-coupled problems in secure coded edge computing, in a unified framework. In particular, we take matrix multiplication, a fundamental building block of many distributed machine learning algorithms, as the representative computation task, and study optimal task allocation and coding design to minimize resource consumption while achieving information-theoretic security.

Spectrum defragmentation (DF) as a connection reconfiguration method is essential in elastic optical networks (EONs) in order to minimize connection blocking probability. This paper proposes the first exact Markov model to computing exact blocking probabilities in EONs with DF by taking into account the occupancy status of spectrum slices of all network links, and waiting and serving connections during a DF process. Since the complexity of the exact Markov model increases exponentially with the network capacity and size, we propose a reduced state model where a link occupancy state is defined by the total occupied slices on a fiber link. Furthermore, using a spectrum fragmentation factor in each occupancy state we calculate state-and class-dependent connection setup rates, which is used to compute approximate blocking in EONs with DF. Notably, we show individually the distinct blocking values, one due to resource unavailability and the other due to fragmentation, both under a random-fit and a first-fit spectrum allocation policies. Our numerical results show that the DF process is very useful in reducing overall connection blocking in EONs. We also observe that blocking due to spectrum fragmentation can be reduced, but not eliminated in a mesh network topology even when an optimal DF scheme is deployed.

We study whether the layered encryption and, thus, anonymity of communication can be implemented all-optically, granting ultra-high speed communication and processing. To this end, we evaluate the feasibility of the state-of-the-art all-optical components, including optical linear feedback shift register (oLFSR) and optical XOR (oXOR), for implementing an all-optical stream cipher (oSC). Just like the function of key expansion in today's electronic ciphers, we use pRNGs and oLFSRs to expand short bit sequences of the secret key, thus enabling the transmission and en-/decryption at line rate; oXOR is used for the actual all-optical encryption. Using the principles of information-theoretical entropic security, we evaluate the key uncertainty (entropy and equivocation) and, thus, the computational complexity of various guessing attacks. We also analyze the maximum length of optical key stream and show that oSC can be designed as secure as in today's systems while operating all-optically at higher line rates.

In this paper, we study end-to-end service reliability in Data Center Networks (DCN) with flow and Service Function Chains (SFCs) parallelism. In our approach, we consider large flows to i) be split into multiple parallel smaller sub-flows; ii) SFC along with their VNFs are replicated into at least as many VNF instances as there are sub-flows, resulting in parallel sub-SFCs; and iii) all sub-flows are distributed over multiple shortest paths and processed in parallel by parallel sub-SFCs. We study service reliability as a function of flow and SFC parallelism and placement of parallel active and backup sub-SFCs within DCN. Based on the probability theory and by considering both server and VNF failures, we analytically derive for each studied VNF placement method the probability that all sub-flows can be successfully processed by the parallelized SFC without service interruption. We evaluate the amount of backup VNFs required to protect the parallelized SFC with a certain level of service reliability. The results show that the proposed flow and SFC parallelism in DCN can significantly increase end-to-end service reliability, while reducing the amount of backup VNFs required, as compared to traditional SFCs with serial traffic flows.

The Network Function Virtualization (NFV) paradigm is enabling flexibility, programmability and implementation of traditional network functions into generic hardware, in form of the so-called Virtual Network Functions (VNFs). Today, cloud service providers use Virtual Machines (VMs) for the instantiation of VNFs in the data center (DC) networks. To instantiate multiple VNFs in a typical scenario of Service Function Chains (SFCs), many important objectives need to be met simultaneously, such as server load balancing, energy efficiency and service execution time. The well-known VNF placement problem requires solutions that often consider migration of virtual machines (VMs) to meet this objectives. Ongoing efforts, for instance, are making a strong case for migrations to minimize energy consumption, while showing that attention needs to be paid to the Quality of Service (QoS) due to service interruptions caused by migrations. To balance the server allocation strategies and QoS, we propose using replications of VNFs to reduce migrations in DC networks. We propose a Linear Programming (LP) model to study a trade-off between replications, which while beneficial to QoS require additional server resources, and migrations, which while beneficial to server load management can adversely impact the QoS. The results show that, for a given objective, the replications can reduce the number of migrations and can also enable a better server and data center network load balancing.

The wide spread deployment of smart edge devices and applications that require real-time data processing, have with no doubt created the need to extend the reach of cloud computing to the edge, recently also referred to as Fog or Edge Computing. Fog computing implements the idea of extending the cloud where the

Parallelism has become one of the key architectural features in 40-/100-/400-Gb Ethernet multi lane distribution (MLD) standards. The MLD packetizes and distributes traffic adaptively over parallel lanes and maps them to parallel network interfaces for wide area transmission, typically over optical networks. As such, the MLD creates not only new network topology abstractions but also enables modular implementations of various new features to improve the system performance. In this paper, we study the performance of the parallelized Ethernet in combination with erasure coding and more specifically random linear network coding (RLNC). We present a novel theoretical modeling framework, including the derivation of upper and lower bounds of differential delay and the resulting receiver queue size-a critical performance measure in the high-speed Ethernet. The results show benefits of a combined usage of parallelism and RLNC: with a proper set of design parameters, the differential delay and the receiver buffer size can be reduced significantly, while cross-layer design and path computation greatly simplified.

Network Function Virtualization (NFV) is a new paradigm, enabling service innovation through virtualization of traditional network functions located flexibly in the network in form of Virtual Network Functions (VNFs). Since VNFs can only be placed onto servers located in networked data centers, which is the NFV's salient feature, the traffic directed to these data center areas has significant impact on network load balancing. Network load balancing can be even more critical for an ordered sequence of VNFs, also known as Service Function Chains (SFCs), a common cloud and network service approach today. To balance the network load, VNF's can be placed in a smaller cluster of servers in the network thus minimizing the distance to the data center. The optimization of the placement of these clusters is a challenge as also other factors need to be considered, such as the resource utilization. To address this issue, we study the problem of VNF placement with replications, and especially the potential of VNFs replications to help load balance the network. We design and compare three optimization methods, including Linear Programing (LP) model, Genetic Algorithm (GA) and Random Fit Placement Algorithm (RFPA) for the allocation and replication of VNFs. Our results show that the optimum placement and replication can significantly improve load balancing, for which we also propose a GA heuristics applicable to larger networks.

Privacy in form of anonymous communication could be comparably both faster and harder to break in optical routers than in today's anonymous IP networks based on The Onion Routing (Tor). Implementing the practical privacy all-optically, however, is not straightforward, as it requires key generation in each anonymization node to avoid distribution of long keys, and layered encryption, both at the optical line rate. Due to the unavailability of cryptographically strong optical key generation and encryption components, not only a layered encryption is a challenge, but an optical encryption in general. In this paper, we address the challenges of optical anonymous networking for the first time from the system's perspective, and discuss options for practical implementation of all-optical layered encryption. To this end, we propose an optical anonymization component realized with the state-of-the-art optical XOR logic and optical Linear Feedback Shift Registers (oLFSRs). Given that LFSR alone is known for its weak cryptographic security due to its linear properties, we propose an implementation with parallel oLFSRs and analyze the resulting computational security. The results show that proposed optical anonymization component is promising as it can be practically realized to provide a high computational security against deanonymization (privacy) attack.

Scrambling or information randomization has been effectively used in various applications to prevent the information falling into the hands of rogue attackers. In this paper, we use the idea of spectrum scrambling to proactively randomize the spectrum across multiple fiber cores to improve connections' security, while at the same time effectively defragmenting the spectrum to improve connections' blocking performance. To this end, we propose a scheme called random spectrum defragmentation (RSD), and model the occupancy pattern of a multi-core fiber link using a multi-class continuous-time Markov chain (CTMC) under the two different spectrum allocation methods, first-fit and random-fit. As the efforts have been on preventing unauthorized access to optical channels, we show that scrambling can be effectively used to help dealing with the physical layer attacks. At the same time, numerical results show that also the blocking can be improved for a particular so-called randomization process (RP) arrival rate, and lower connection reconfiguration time.

Link capacity dimensioning is the periodic task where ISPs have to make provisions for sudden traffic bursts and network failures to assure uninterrupted operations. This provision comes in the form of link working capacities with noticeable amounts of headroom, i.e., spare capacities that are used in case of congestions or network failures. Distributed routing protocols like OSPF provide convergence after network failures and have proven their reliable operation over decades, but require over-provisioning and headroom of over 50\%. However, SDN has recently been proposed to either replace or work together with OSPF in routing Internet traffic. This paper addresses the question of how to robustly dimension the link capacities in emerging hybrid SDN/OSPF networks. We analyze the networks with various implementations of hybrid SDN/OSPF control planes, and show that our idea of SDN Partitioning requires less amounts of spare capacity compared to legacy or other hybrid SDN/OSPF schemes, outperformed only by a full SDN deployment.

In current Data Center Networks (DCNs), Equal- Cost MultiPath (ECMP) is used as the de-facto routing protocol. However, ECMP does not differentiate between short and long flows, the two main categories of flows depending on their duration (lifetime). This issue causes hot-spots in the network, affecting negatively the Flow Completion Time (FCT) and the throughput, the two key performance metrics in data center networks. Previous work on load balancing proposed solutions such as splitting long flows into short flows, using per-packet forwarding approaches, and isolating the paths of short and long flows. We propose DiffFlow, a new load balancing solution which detects long flows and forwards packets using Random Packet Spraying (RPS) with help of SDN, whereas the flows with small duration are forwarded with ECMP by default. The use of ECMP for short flows is reasonable, as it does not create the out-of-order problem; at the same time, RPS for long flows can efficiently help to load balancing the entire network, given that long flows represent most of the traffic in DCNs. The results show that our DiffFlow solution outperforms both the individual usage of either RPS or ECMP, while the overall throughput achieved is maintained at the level comparable to RPS.

Data randomization or scrambling has been effectively used in various applications to improve the data security. In this paper, we use the idea of data randomization to proactively randomize the spectrum (re)allocation to improve connections' security. As it is well-known that random (re)allocation fragments the spectrum and thus increases blocking in elastic optical networks, we analyze the tradeoff between system performance and security. To this end, in addition to spectrum randomization, we utilize an on-demand defragmentation scheme every time a request is blocked due to the spectrum fragmentation. We model the occupancy pattern of an elastic optical link (EOL) using a multi-class continuous-time Markov chain (CTMC) under the random-fit spectrum allocation method. Numerical results show that although both the blocking and security can be improved for a particular so-called randomization process (RP) arrival rate, while with the increase in RP arrival rate the connections' security improves at the cost of the increase in overall blocking.

Recently, physical layer security in the optical layer has gained significant traction. Security threats in optical networks generally impact the reliability of optical transmission. Linear Network Coding (LNC) can protect from both the security threats in form of eavesdropping and faulty transmission due to jamming. LNC can mix original data to become incomprehensible for an attacker and also extend original data by coding redundancy, thus protecting a data from errors injected via jamming attacks. In this paper, we study the effectiveness of LNC to balance reliable transmission and security in optical networks. To this end, we combine the coding process with data flow parallelization of the source and propose and compare optimal and randomized path selection methods for parallel transmission. The study shows that a combination of flow parallelization, LNC and randomization of path selection increases security and reliability of the transmission. We analyze the so-called catastrophic security threat of the network and show that in case of conventional transmission scheme and in absence of LNC, an attacker could eavesdrop or disrupt a whole secret data by accessing only one edge in a network.

Hybrid IP networks that use both control paradigms - distributed and centralized - promise the best of two worlds: 1) programmability of software-defined networking (SDN) and 2) reliability and fault tolerance of distributed routing protocols, like open shortest path first (OSPF). Typically, a hybrid network deploys SDN to control prioritized traffic and OSPF to assure care-free operation of best effort traffic. We propose a new hybrid network architecture, called SDN partitioning (SDNp), which establishes centralized control over the distributed routing protocol by partitioning the topology into sub-domains with SDN-enabled border nodes. In our approach, OSPF's routing updates have to traverse SDN border nodes to reach neighboring sub-domains. This allows the central controller to modify how sub-domains view one another, which in turn allows to steer inter-sub-domain traffic. The degree of dynamic control against simplicity of OSPF can be traded off by adjusting the size of the sub-domains. This paper studies the technical requirements, presents a novel scheme for balanced topology partitioning, and provides the corresponding network management models for SDNp. Our performance evaluation shows that - already in its minimum configuration with two sub-domains - SDNp provides significant improvements in all respects compared to legacy routing protocols, whereas smaller sub-domains provide network control capabilities comparable to full SDN deployment.

We focus on the problem on security, and to an extent, privacy in optical networks, under the assumption that the wiretap attacker either gains knowledge about secret key and encryption algorithm, or is able to decipher the identity of source and destination, and thus infer the user's privacy of communication. We propose two combined features to address this: (1) to parallelize the transmission between source and destination, i.e., split serial optical flows into multiple flows, and (2) to randomize the traffic forwarding along the parallel paths. To this end, we compare two forwarding methods for parallel transmission: i) explicit forwarding (PAR-EXP), akin to WDM networks with mulitpath routing, ii) random forwarding (PAR-RND), akin to either an optical packet switched network, or a multilayer optical network with Ethernet-based forwarding. The study shows that parallel transmission always increases security compared with serial transmission, whereby random forwarding is potentially more secure than explicit forwarding. We define the \emph{security degree} of the network and show that it depends on various factors, such as the number of wiretap edges, network topology and number of disjoint paths available, and link capacity.

A lot of research, development, and standardization efforts are going on to extend the latest buzz in packet networks called Software-Defined Networking (SDN) to the optical domain. We argue though that a reasonable combination of packet layer SDN with dynamic optical transport (in whatever flavor) is already possible with some architectural adaptations. We therefore propose to combine a recently proposed networking scheme called SDN Partitioning with dynamic optical circuits used in the form of Optical Bypasses. The method uses a few SDN nodes in a hybrid SDN/OSPF network, such that they partition the OSPF domain into sub-domains, thereby already achieving traffic engineering capabilities comparable to full SDN operation only by manipulating OSPF routing protocol updates when they cross sub-domain borders. The method provides that local routing inside sub-domains remains stable at all times, while inter-sub-domain routes can be optimized. The Optical Bypasses among border nodes come into play to dynamically offload transit traffic from frequently traversed (transit) sub-domains to the optical layer in case of increased traffic demands. Our simulation results show that the combination of a few SDN nodes with a few Optical Bypasses allows to cope with traffic surges up to a degree that renders excessive over-provisioning of link capacities or full SDN migration completely unnecessary.

As optical networks evolve towards more dynamicity and an ever more efficient and elastic spectrum utilization, a more integrated, fault tolerant and system efficient design is becoming critical. To increase efficiency of spectral resource in bit rate per Hz (bit/s/Hz), high-level modulation formats are used, challenged by the accompanying optical impairments and the resulting limitation of optical reach. Previous work has addressed the issue of optical reach and transmission fault tolerance in the physical layer by deploying various FEC schemes and by a careful design of optical transceivers and links. This paper uses a different approach, applicable to link and networking layers. We propose a novel theoretical framework, whereby a randomized linear network coding (LNC) is applied to the main optical path, and in parallel, an auxiliary optical path is used at much lower transmission speeds, i.e., in addition to the main path. With the reception of the auxiliary path, as we analytically show, the system is highly tolerant to bit errors and packet loss caused by optical impairments in the main path, whereby alleviating the constraints on optical transmission quality and indirectly achieving better optical reach and spectral efficiency. The results are shown for a case study of high-speed Ethernet end-system transmitted over optical OFDM networks, which due to the inherent system-level parallelism in both networks, present one of the most interesting candidate technologies for the proposed method to yield best performance.

As the Software Defined Networking (SDN) paradigm gains momentum, network operators wonder about how to go about replacing their legacy IP routers with SDN-compliant ones. While forklift network upgrades are impractical in operational networks, the promise of SDN is too compelling. Thus, a viable solution is to gradually migrate over time, leading to hybrid OSPF/SDN networks. Although recent studies proposed device architectures, protocols and algorithms required in such hybrid networks, a single routing domain's SDN migration trajectory - as a whole - has hardly been studied. Significance of the sequence in which IP routers are replaced with SDN routers and optimal sequence of router replacements need to be investigated. In this paper, we address these questions based on SDN's traffic engineering (TE) gains and an operator's SDN investment constraints. We propose optimization techniques and heuristics to plan an effective migration schedule in a single routing domain and demonstrate the significance of such a schedule using relevant network management metrics. Our results suggest: (a) the sequence of IP routers migrating to SDN largely determine the resulting TE gains, and, (b) to determine the best migration sequence, novel greedy algorithms perform almost as good as optimization techniques.

Free Space Optics (FSO) is an attractive high-speed communication technique, with advantages of low power consumption, low cost, easy setup and secure data transmission. On the other hand, the FSO channels are sensitive to atmospheric effects, which can decrease the reliability and capacity of the FSO system. This paper proposes a novel method to achieve a reliable FSO transmission, that combines redundancy of parallel transmission supported by additional radio frequency (RF) or FSO links with the redundancy achievable through the adaptive channel coding. We provide a comparative analysis of three scenarios: serial transmission over one FSO path, parallel transmission over multiple FSO paths, and hybrid transmission with one FSO path and one RF path. For each scenario, we derive analytical expressions of expected code rate as well as the upper bound to evaluate the expected information bit rate. Finally, we show that a reliable transmission is possible in each scenario. The hybrid RF/FSO system shows the most promising performance in terms of reliability and system capacity. The best performance can be achieved in systems where the RF link can offer a variable transmission rate, i.e., able to adapting to the atmospheric effects of the FSO channel.

Software Defined Networking (SDN) is an emerging network control paradigm focused on logical centralization and programmability. At the same time, distributed routing protocols, most notably OSPF and IS-IS, are still prevalent in IP networks, as they provide shortest path routing, fast topological convergence after network failures, and, perhaps most importantly, the confidence based on decades of reliable operation. Therefore, a hybrid SDN/OSPF operation remains a desirable proposition. In this paper, we propose a new method of hybrid SDN/OSPF operation. Our method is different from other hybrid approaches, as it uses SDN nodes to partition an OSPF domain into sub-domains thereby achieving the traffic engineering capabilities comparable to full SDN operation. We place SDN-enabled routers as subdomain border nodes, while the operation of the OSPF protocol continues unaffected. In this way, the SDN controller can tune routing protocol updates for traffic engineering purposes before they are flooded into sub-domains. While local routing inside sub-domains remains stable at all times, inter-sub-domain routes can be optimized by determining the routes in each traversed sub-domain. As the majority of traffic in non-trivial topologies has to traverse multiple subdomains, our simulation results confirm that a few SDN nodes allow traffic engineering up to a degree that renders full SDN deployment unnecessary.

Given the assumption that network equipment vendors will soon deploy new energy saving mechanisms in their routers, the question we are aiming to answer is this: Where in the network to start the migration to energy efficient routers, and how large is the impact of the migration scheduling on the total energy saving? To this end, the paper studies a scenario where network nodes can reduce the power consumption proportional to the router traffic load, and elaborates on the impact of the migration node sequence that determines at which point in time which node migrates. While every router can gradually increase and decrease the consumption based on load, at the same time, the network as a whole is subject to a gradual migration on its own, where energy efficient network routers are installed over multiple planning periods. Our results show that an increase of saved energy up to 12 percentage points is possible only by changing the migration sequence of energy efficient routers. Additionally, we show that deploying mechanisms of Traffic Engineering can further significantly improve the overall energy savings especially during the early stages of migration.

We analyze the effects of dynamic optical circuit setup on IP routing in general and on two routing mechanisms in particular, i.e., explicit routing and shortest-path-first routing. We present analytical models for computing the size and placement of optical circuits and propose model adaptations driven by the IP router system design. The results show that without careful consideration of intrinsic capabilities of IP routing protocol and forwarding, the size and location of optical circuits used can be vastly underestimated, also leading to significant disruptions in real networks. We present the Optical Bypass mechanisms and show that these methods, unlike traditional IP routing-based solutions, affect a comparatively lower number of IP routes and can be computed near-optimally, even under unknown traffic matrix conditions, making them effective and feasible.

In this paper, we analyze the question of network migration to Software Defined Networking (SDN) from the perspective of Traffic Engineering (TE). For a given network topology and migration planning horizon, we ask the question of which routers in the IP network should migrate to SDN-enabled operation to reduce the need for network capacity upgrades over a given time horizon. We propose an algorithm to determine the optimum schedule for node substitution within a network, which shows that already a few SDN routers, when strategically located, provide a stably large number of path alternatives to be used in TE, thus substantially reducing the need for large network capacity upgrades.

Parallel transmission, as defined in high-speed Ethernet standards, enables to use less expensive optoelectronics and offers backwards compatibility with legacy Optical Transport Network (OTN) infrastructure. However, optimal parallel transmission does not scale to large networks, as it requires computationally expensive optimal multipath routing algorithms to minimize differential delay, and thus the required buffer size to ensure frame synchronization. In this paper, we propose a novel parallel transmission framework for high-speed Ethernet, which we refer to as network coded parallel transmission, capable of effective buffer management and frame synchronization without the need for complex multipath algorithms. We show that using network coding can reduce the delay caused by packet reordering at the receiver, thus requiring a smaller overall buffer size, while improving the network throughput. We design the framework in full compliance with high-speed Ethernet standards specified in IEEE802.3ba and present detailed schemes including encoding, data structure of coded parallel transmission, buffer management and decoding at the receiver side. The proposed network coded parallel transmission framework is simple to implement and presents a potential major breakthrough in the system design of future high-speed Ethernet.

Elastic optical path networks have been proposed as a promising approach to flexibly use the optical spectrum (hence the name flexi-grid) thus overcoming the rigidness of traditional WDM networks where optical channel spacing is equidistant irrespectively of the type of signal transmitted. Past research effort in spectrum allocation focused on single transmission paths with a uniform modulation format in the network. However, by using optical OFDM technologies, the flexi-grid networks can benefit from multipath routing with parallel transmission that uses flexible modulation format assignment. In this paper, we propose novel multipath routing and spectrum assignment (RSA) algorithms with distance-adaptive modulation format assignment and analyze the effectiveness in dynamic traffic scenarios. We show that a combined usage of multipath routing and flexible distance-adaptive modulation format assignment can reduce blocking ratio while consuming portions of spectrum comparable to single path routing. Our results also show that a careful dimensioning of guard-band size is a critical factor, and that the full benefits of parallel transmission can be better achieved in networks requiring a small guard-band size.

Conventionally, network migration models study competition between emerging and incumbent technologies by considering the revenue increase and migration cost. We propose to extend the existing network migration models with new critical factors, including (i) synergistic relationships across multiple technologies, (ii) reduction in operation expenditures (OpEx), and, (iii) effect of social factors on human decisions. To this end, we propose a novel agent-based migration model considering these factors. Based on the model, we analyze the case study of optimal path computation with joint migration to two emerging networking paradigms, i.e., IETF Path Computation Element (PCE) and Software-Defined Networking (SDN). Our results demonstrate the synergistic effects of migration to multiple complementary technologies, and shows that a technology migration may be eased by the joint migration to multiple technologies.

In elastic optical networks, the spectrum consecutive and continuous constraints may cause the so-called spectrum fragmentation issue, degrading spectrum utilization, which is especially critical under dynamic traffic scenarios. In this paper, we propose a novel multipath de-fragmentation method which aggregates spectrum fragments instead of reconfiguring existing spectrum paths. We propose an optimization model based on Integer Linear Programming (ILP) and heuristic algorithms and discuss the practical feasibility of the proposed method. We show that multipath routing is an effective de-fragmentation method, as it improves spectral efficiency and reduces blocking under dynamic traffic conditions. We also show that the differential delay issue does not present an obstacle to the application of multipath de-fragmentation in elastic optical networks.

Load adaptive energy saving (LAES) schemes have been widely considered as a solution to reduce the energy consumption in networks. In LEAS, the time is sliced into time periods according to the daily pattern of traffic loads, such as peak and off-peak traffic, and energy saving operations are typically activated during the off-peak traffic. The effectiveness of LAES depends on the number of time slices precisely. A larger number of small time slices is desirable due to a better adaptiveness to traffic fluctuations. At the same time, energy saving operations should not be too frequent, as they have impact on routing stability and operational procedures. In this paper, we study the effect of nonuniform time quantization on the efficiency of energy saving. We propose a novel time quantization algorithm and show that optimizing time quantization can considerably reduce the number of energy saving operations, while comparably saving the same energy as previously reported.

The Path Computation Element (PCE) is positioned nowadays as one of the solutions that almost every carrier will eventually deploy. The PCE architecture as well as a number of components, including the PCE Communication Protocol (PCEP), have been standardized by the IETF. However, a number of challenges remain to be solved on its way from standards to deployment. In particular, the existing proposals for multilayer path computation within the PCE framework need to be further developed and tested, before considering their possible integration into operational networks. This is especially true for the interoperability of the various PCE implementations and the extensions such as the Virtual Network Topology Manager (VNTM) which cannot be taken for granted. This paper presents a functional implementation of coordinated computation of multilayer paths supported through inter-layer PCE communication, where one PCE is developed by industry and the other as an open-source effort. To this end, we consider an IP/MPLS network deployed over a Wavelength Switched Optical Network (WSON), each of which deploying its own PCE, in an attempt to create an inter-operable multilayer solution. We discuss the key challenges that the research community will face in this area, which in turn will drive a considerable part of the upcoming efforts in terms of standardization.

Despite the research advances in intra-domain light-path provisioning in WDM networks, efficient and practical schemes for path computation and resource advertisement in multi-domain mesh networks still need to be developed. Most of the proposed solutions in the literature lacks the ability to convey optical network-specific Traffic Engineering information and are based on periodic message flooding technique. The proposed solution provides a novel and economical way to disseminate information while performing the domain chain computation. It also provides a policy-based wavelength selection scheme that allows a load balanced provisioning of end-to-end lightpaths.

Optical networks can benefit from multipath routing by routing traffic over diverse fiber links to fulfill bandwidth requirements, balance network load and improve resource utilization. This paper focuses on the effectiveness of multipath routing on aggregating residual bandwidth of the established lightpaths using dynamic traffic grooming. An optimization model based on Integer Linear Programing (ILP) is formulated to leverage multipath routing and grooming to serve connection requests with known duration, in a scenario where the dynamic traffic can demand bandwidth either larger or smaller than a single wavelength capacity. The impact of a balancing policy, referred to as Holding Time Balancing (HTBalancing), on multipath routing with traffic grooming in optical networks is also investigated. Numerical results show that the proposed optimization model can achieve a lower bandwidth blocking and a better load balancing with HTBalancing policy, when compared with single path routing. The proposed relaxation algorithm can effectively find near optimal solutions, and be applied in realistic scenarios.

The rapid growth of network services is forcing operators to use of inter-layer coordination between the IP and the transport network management systems in order to optimize resource utilization within the carriers network ecosystem. However, the diversity of technology and the related management systems within the IP and transport networks poses significant challenges in developing a single unified management ecosystem for both IP and transport networks. In this work, we present a basic prototype of a Service Oriented Architecture (SOA) based framework proposed in the EU project ONE for programmable orchestration of IP and transport management processes. The proof-of-concept prototype uses web services and BPEL to facilitate programmable inter-layer coordination, and was tested on a number of use cases. We measured the performance penalty of using the SOA framework against the traditional JAVA based provisioning approaches used in integrated multi-layer management, and show that the penalty incurred in execution times were sufficiently small while bringing several significant advantages during the software design, development, integration and maintenance, which is simple and highly effective.

This paper discusses the evolution of optical networks from foundational data transport utilities and infrastructure to true enablers of next-generation services and facilities, specifically diverse rich digital media services, which are being produced, processed, and consumed by an increasingly sophisticated variety of end users. This evolution has been enabled by several major recent advances in optical networking, including the deployment of high-speed optical connections from core networks to enterprise and residential users; the dramatic increase in optical signal speed and reach; and the improvement in control and management planes to support increasingly diverse types of traffic and services. We describe two distinctive areas of broad impact of optical network technologies related to rich media services: high-resolution scientific visualization and high-quality, real-time consumer-driven media production and distribution. Finally, we discuss key trends that may develop next as optical networks further evolve shaped by both rich digital media service requirement and technology innovation.

In this paper, we present a simple case study on energy efficiency in IP-over-WDM networks with dynamic circuit capability and compare two different load adaptive schemes, referred to as switch-on and switch-off. While the switch-off technique was already proposed for energy conservation, the switch-on approach is a new paradigm which is based purely on one-time implementations of dynamic circuit capability. Our results show that both approaches can significantly reduce the power consumption and decrease the necessary totally installed capacity, but unlike the switch-off scheme, the switch-on scheme does not affect the path redundancy in the network. Furthermore, switch-on can reduce the number of routing reconfigurations required in the network. While this seems ideal, our results also show that the switch-on scheme uses a large number of small capacity interfaces which may not be suitable from a network planning perspective as it might require frequent capacity upgrades, which alone is an interesting avenue for future research.

We present an novel analytical framework for modeling current IP offloading schemes and show their comparative performance. We analyze relevant parameters, including IP routing re-convergence time, the number of affected IP routes, and the cost of circuit capacity used for offloading. Our results show that emerging offloading solutions based on invisible bypasses can better maintain network stability and reduce the routing reconfiguration effort compared to the traditional traffic engineering approaches that modify the IP routing.

Significant research and standardization efforts are underway to enable automated computation and reservation of connection-oriented paths (circuits) across multiple domains. In the absence of a secure authentication and authorization mechanism, however, carriers continue to provision connections manually, which leads to large setup delays and increases possibility of configuration errors. Carriers also lack mechanisms to meter connection quality during the service lifetime and typically do not exchange accounting information for established connections for auditing and billing purposes. In this paper, we address the challenge for automatic multi-domain path provisioning with authentication, authorization and accounting (AAA) capabilities in carrier-grade transport networks. The designed solution secures computation and reservation for path provisioning and also leverages a standard accounting model which incorporates the accounting signaling for an inter-domain connection. In order to evaluate the impact of the proposed framework on signaling performance, we also provide an analytical framework scalable to large inter-domain network scenarios. We verify the analysis using event-driven simulations and then use this analytical model to quantify the feasibility of our model in terms of signaling load and signaling delay for a wide range of network scenarios.

In this paper, a new algorithm for dynamic traffic grooming is introduced. It considers the holding-time of the connections and it aims at balancing the load among existing lightpaths to avoid the formation of bottlenecks and, consequently, high blocking probability values. Results indicate that it produces significantly lower blocking probabilities when compared to other holding-time-aware algorithm. Moreover, it promotes a fair distribution of blocking among source-destination pairs.

The Internet is susceptible to congestion due to progressively increasing traffic as well as short-lived traffic surges. Traditional mechanisms to counter these effects over-provision the IP network to a significant degree. At the same time, the advent of dynamic circuits in L2/L1 networks has fueled the research in the area of network engineering which deals with the ability to add capacity in the higher layer (IP) by establishing dynamic circuits in the lower layers. However, effective usage of dynamic circuits is a challenge, as they can lead to IP routing instabilities. We present an approach wherein IP bypass links are established using dynamic circuits to alleviate congestion while keeping the routing stable in the IP layer. Proposed is an ILP based approach which computes the optimal set of circuits with and without the knowledge of the traffic matrix in the IP layer. The results show that even without the knowledge of the traffic matrix, the proposed method computes only a marginally higher number of bypasses, albeit at a higher capacity.

Connection-oriented path setup is becoming one of the key features of the next-generation Internet with the Multi-Protocol Label Switched (MPLS) framework and its generalized version GMPLS addressing the automated label switched path (LSP) setup. Despite the significant technological progress in various LSP implementations, the integration of robust security features for authentication and authorization for global path setup remains an open issue. However, security is becoming essential for carrier-grade operations as it directly translates to inter-carrier's service level agreements and user's satisfaction with quality of the services purchased. In this paper, we propose to study the applicability of NSIS (Next Step Signaling Protocol) for LSP setup signaling with security features; NSIS is a generic protocol for configuring network nodes that supports multiple existing transport and security protocols. We design an NSIS application called NSIS-LSP which takes advantage of the NSIS transport security features and has own features for application layer authentication. Unlike the existing path setup protocols that refer to security mechanisms between neighboring domains for resource provisioning, NSIS-LSP also allows mutual authentication between source and remote provisioning domains. We use an open-source NSIS testbed and simulations to obtain the performance results, which show that the NSIS-LSP application carries significant potential for future implementations.

Recently, cellular networks have witnessed major developments pertaining to user mobility, rich multimedia service offering and a melange of network access options including 3G/4G, WiFi, WiMAX, etc. As a result, sessions are expected to last longer and users are more likely to roam between access technologies and to other networks. As architectural design is advancing in all-IP cellular systems such as the Long Term Evolution (LTE), the question we propose to address is whether the established results on handoff and roaming statistics used in cellular theory still apply. To this end, we revisit the theory for handoff statistics and take up the challenge on extending the current model under general assumptions for session distributions, user mobility, network coverage and access technology. We show that the derived model yields estimates of handoff frequency and roaming statistics which cannot be obtained otherwise. The key strength of the proposed analysis is offering closed form results, which are easy to use and can lead to more accurate conclusions about the signaling load and the observed QoS, as a direct function of handoff statistics.

Authentication, Authorization, and Accounting (AAA) deployments are expected to grow significantly in emerging mobile systems as they offer a plethora of services and mobile applications. In current systems, network access servers (NAS) periodically report the service usage of mobile users located within their coverage areas. The periodic reports are used by the billing systems to minimize the incurred capital losses if the serving NAS fails. While shorter reporting intervals are desired for lower losses, they can potentially result in undesirably high signaling load. Because it is prohibitively difficult to obtain optimal reporting intervals in mobile systems due to multitudes of services with different mobility profiles, current accounting standards offer no quantitative measures for selecting a proper reporting interval and AAA systems are typically designed via over provisioning. To address this issue, we propose an adaptive optimization mechanism in multiservice AAA systems which limits the potential loss without excessively generating unnecessary usage reports. Our optimization mechanism embraces the current AAA IETF standards RADIUS and its successor Diameter and does not require any modifications to the AAA protocols nor to the network access servers' implementation, and its implementation scope is limited to the AAA systems. The results demonstrate that our mechanism is robust under various operational conditions, easy to implement, and offers considerable potential for loss control compared to the current static approaches.

In this paper, we propose to apply multipath routing in optical networks for the emerging high-performance applications with extremely high bandwidth requirements, typically larger than the capacity of one wavelength. To this end, we present a novel Multipath lightpath Provisioning mechanism and derive an optimal solution by an ILP (Integer Linear Programming) approach, with differential delay and bandwidth as constraints to multipath finding. Our mechanism can set up multiple lightpaths over multiple fiber-level paths not only to satisfy the extremely high bandwidth requirements, but also to reduce the minimum bandwidth required for backup paths as it reduces the amount of traffic affected by single fiber breaks. For comparison, we also present an ILP-based Single Path Lightpath Provisioning mechanism and show that its multipath counterpart performs better independently of the mesh topology under study. The performance results demonstrate that the proposed multipath lightpath provisioning mechanism outperforms the traditional single path routing by decreased bandwidth request blocking ratio, while reducing the amount of traffic that may be affected by single link failures.

The mean number of handoffs is a fundamental performance measure in any mobile system, as it directly relates to the signaling load in the network as well as to the delivered QoS. As the mobile services are evolving from simple cellular voice calls towards media and data sessions, and as cellular providers are reinventing their businesses by incorporating third party services, the handoff rate will even play a more pivotal role. This is because future sessions are expected to last longer than voice calls and users are more likely to roam into other networks. Existing results provide an estimate of the mean number of handoffs in networks composed of an infinite number of access gateways and hence consider neither the topological arrangement of the gateways nor the mobility patterns between them. In this paper, we obtain a closed form solution for the mean number of handoffs under generic assumptions of two- dimensional Markovian mobility patterns, spatial arrangement of the access gateways, as well as generic session times and gateway residence times. Our solution unveils a new insight into mobility foundations as it shows that the consideration of the mobility pattern and the access gateways' layout simply transforms the known equation for the mean number of handoffs of an infinite network size from scalar to vector representation. We demonstrate that the mean number of handoffs is a non-linear function of the gateway spatial arrangement and user mobility.

While the introduction of new Ethernet-based wide-area solutions, such as provider-backbone bridging traffic engineering- PBB-TE, paves a way for Ethernet to become a carrier class service, it is restricted to the metro area and hence unable to provision global end-to-end communication. In this paper, we propose a new scheme for all-Ethernet wide area networking that involves a unique addressing and routing mechanism, and leads to a scalable, hierarchical and service-oriented transport network architecture. The scalability is achieved by means of abstraction of any irregular physical topology into a regular logical topology, based on the concept of binary trees. The regular logical topology is represented with logical 1 x 2 Ethernet switches as the fundamental building blocks which allow switching using a unique binary addresses in a simple and automated fashion. We propose an evolutionary architecture to provide end-to-end Ethernet routes using binary addresses embedded in stacked VLAN tags on native Ethernet frames, in line with the emerging standards. The results show that a significant simplification of wide-area internetworking can be achieved, while supporting carrier-grade network performance with all-Ethernet features.

IETF is working on the design of new architectures and signaling solutions to support inter-AS (autonomous system) GMPLS-TE (generalized multi protocol label switching with traffic engineering) for multi-domain, multi-carrier connection setup with guaranteed quality of service (QoS). In addition, the path computation element (PCE) working group is developing the framework for inter-domain path computation. However, many issues are still open regarding the joint path computation and path setup signaling solutions for inter-carrier authentication and authorization (AA). In this paper, we propose the first security solution which integrates inter-domain AA features in the PCE path computation framework. Specifically, we define a new architecture for inter-domain QoS path provisioning based on an extension of the PCE framework to include features that allow domains interested in inter-domain resources to get AA for end-to-end path provisioning over multiple domains belonging to different carriers. In addition, we introduce a mechanism to tie policies controlling path setup with the AA mechanisms introduced in the PCE framework. While at present provisioning of inter-domain paths is based on rather static settlements between neighboring domains that make end-to-end QoS provisioning a challenge, we propose an AA framework that allows domains interested in setting an inter-domain QoS path to have guarantees about resource provided by each domain along the path from source to destination. Simulation results show the performance of the model proposed in networks having different size and connectivity.

Recently, all-IP wireless systems have been standardized under the IP multimedia subsystem (IMS) framework to support next generation value added mobile services. In all-IP networks, multiple base stations connect to IMS policy servers through IP-based access gateways. As users move from one access gateway area into another, the corresponding access gateways initiate signaling flows for QoS authorization toward the IMS policy server. Depending on the service, the policy function may contact one or more application servers resulting in variable and prolonged signaling delays upon each handoff. Such delays can be of the order of seconds, depending on the specific system design and whether the user is roaming, which maybe critical to the quality of real time services. In this paper, we propose a novel proactive signaling method in the application layer that conveys authorization delay constraints from the IMS to the radio layer and thus mitigates the effects of variable signaling delay. Our method is practically relevant as it uses the already established mechanisms of authentication and authorization signaling via standardized interfaces and protocols. Using the OPNET simulator, we demonstrate that our scheme is also scalable, as its corresponding signaling overhead is upper bounded to approximately double the handoff rate.

Authentication, authorization, and accounting (AAA) systems are one of the most significant architectural components of the current cellular networks and within the emerging IP multimedia subsystem (IMS) standard. Despite the operators' experience in AAA operations, very little is fundamentally known about the expected signaling rate towards the AAA system. In this paper, using stochastic and renewal theoretic techniques, we develop the first analytical model for the AAA signaling rate as a function of protocol parameters, users' access rates, session durations, and mobility. We provide model approximations and evaluate their accuracy under various operational conditions. Our results show that the AAA signaling rate is a monotonic non-linear function of the mobility rate and asymptotically converges to the AAA signaling rate in fixed networks. We also show that by adjusting the accounting interim and the authorization-lifetime intervals from half to full mean session duration, it is possible to define an AAA operational range for accounting messages that minimizes the signaling rate fluctuations due to likely perturbations in session and mobility statistics. The results also include the effect of the session dropping during handoffs and shows that is marginal in operational networks.

Recent advances in wireless all-IP telecommunications made it possible for the operators to offer more sophisticated services with mixtures of high quality audio, video, data, and interactive applications. However, mobility still poses a challenge for quality of service (QoS) maintenance in such networks as it typically causes large signaling delays which not only result in poor service quality but also in service interruptions. In this paper, we propose a standard-friendly proactive signaling mechanism that minimizes handover delay between access gateways for intra-and inter-operator mobility scenarios. Based on the current standard framework, which was proposed to allow services in the Internet multimedia subystem (IMS) to authorize and control QoS in the Internet and radio tiers, we provide detailed signaling flows and evaluate our mechanism using delay budgets and signaling loads. We show that our method offers measurable delay savings relative to the current systems. We also analyze the signaling overhead of our mechanism using a Markov chain model. Our results related to the signaling delay and load give us confidence that our method can lead to higher perceived quality, more intelligent network selections, and seamless network convergence.

The backhaul network, which connects base stations to the radio network controllers, poses a major challenge in cellular networks today due to the expensive maintenance, operation and poor scalability in response to traffic dynamicity. The majority of the present backhaul networks is composed of either leased lines, i.e., T1/E1 and/or microwave links. Recently, wireless mesh networks have been proposed as a possible backhaul transport media, not only for their significant cost savings but also for their scale, flexibility and resilience. At the same time, however, wireless mesh backhauls pose many technical challenges including timing synchronization for GSM networks, bandwidth reservation techniques, dynamic bandwidth control, and billing. To address some of these challenges, we propose the first ever billing architecture for wireless mesh networks and analyze its performance and scalability in response to a novel, threshold based bandwidth management algorithm. Using analysis based on a simple Markov process, we evaluate a highly unstable reservation mechanism and establish the corresponding upper bounds on the billing signaling performance over mesh backhauls. The analysis and simulation results show that our billing scheme scales well even for poor implementations of bandwidth reservations.

This paper analyzes the performance of a system architecture for the authentication, authorization, and accounting (AAA), in a scenario where 3G users generate traffic and invoke signaling messages towards the AAA system. The AAA system performance, in terms of input arrival rate, retransmissions, timeouts, and number of proxy servers, is particularly critical in today's 3G networks due to the traditional separation of radio and IP engineering. Whereas the traffic generated by the wireless users is a known parameter to radio engineers, 3G operators lack practical methods for obtaining the corresponding parameters at the AAA side. This problem is even more complex for roaming scenarios where the AAA functions are configured in proxy server chains. To address these issues, we analyze an AAA system architecture based on the RADIUS protocol, which is currently the main AAA protocol adopted by the telecommunications industry. We present an analytical model for obtaining the RADIUS signaling rate, and verify it by simulations. We also analyze two typical application scenarios: 1xEV-DO and roaming. Based on the achieved results, we propose new and practical approaches to optimize the AAA system performance both in terms of signaling and normalized server load.

Dynamic sharing of the common physical network is envisioned as a key enabler for the emerging Internet technologies. This paper addresses challenges related to resource sharing in the physical layer and analyzes the performance of infrastructure service provision with control plane mechanisms based on generalized multi protocol label switching (GMPLS). In our approach, the provisioning of infrastructure services is supported by two novel concepts for GMPLS traffic engineering (TE): resource visibility and inter-domain exchange. Resource visibility is a new network control plane concept, which defines the usage polices for transmission, multiplexing, and switching resources in multiple GMPLS layers. In our architecture, every network resource may exhibit different visibility to different services at different layers. The inter-domain exchange, here referred to as GMPLS exchange point (GXP), is the physical layer equivalent of the Internet exchange point (IXP). Just as how the IXP manages interconnections of autonomous systems (AS) in the Internet, the GXP manages dynamic interconnections of multiple provider domains and enables them to advertise their physical resources to other domains. We model the dynamic provisioning of infrastructure services using graph theory and deploy GMPLS traffic engineering (TE) to optimize the routing and resource yields. The results obtained demonstrate that traffic engineering with resource visibility and GXP brings significant performance benefits in resource utilization and infrastructure extensibility, especially when network providers set up LSPs as a result of collaborative and carrier-neutral traffic engineering where they share information about resource capabilities and utilization

This paper analyzes the session setup delay in the IP multimedia subsystem (IMS) with the CDMA2000 evolution data only rev. A (EV-DO rev. A) standard for wireless transmission. Session setup delay is particularly critical for interactive multimedia applications, such as gaming, push-to-X and voice over IP (VoIP), as it directly translates in user perception of service quality. Keeping signaling delay low, however, is a challenge in IMS due to the text-based nature of the session initiation protocol (SIP) for signaling, and, more significantly, due to the lossy and capacity constrained wireless links. To address this challenge, we analyze the session setup delay end-to-end, by taking into account key system properties across all layers, ranging from radio links to IMS signaling architecture. We present a model for cross-layer performance analysis and simulation, which includes the statistical properties of the EV-DO (rev. A) wireless channel, and also takes into consideration the properties of transport protocols (TCP, UDP) and SIP signaling (message size and compression). By means of analysis and simulations, we study the setup delay performance of a generic, multi-operator IMS communication scenario between two mobile users. We describe how session setup delay can be estimated and reduced in realistic IMS settings and we propose architecture alternatives to the basic IMS scenario. The results derived from this study show that the proposed methods can incrementally lead to a lower setup delay and less sensitivity to the radio transmission quality and frame error rate compared to the base IMS scenario

All-IP wireless networking infrastructures, such as those based on Internet Multimedia Subsystem (IMS), are becoming increasingly important due to the variety of applications they support, which include Voice over IP (VoIP) and real-time streaming multimedia. These applications are imposing specific requirements on network and wireless systems in terms of session (call) setup time. Call setup time is particularly important for interactive applications, such as gaming and push-totalk communication, where the user experience of time that takes to start or join a session directly translates in user perception of QoS. In this paper, we present a comprehensive approach to analyze the call set up time in IMS systems deploying Session Initiation Protocol (SIP) for signaling and CDMA2000 Evolution Data Only Rev. A (EV-DO-A) standard for wireless link transmission. Our model is based on practical assumptions for the system design parameters, including radio link delays, IMS processing and signaling properties. We derive important numerical results and present methods to improve the call setup time. Finally, we discuss directions for further research.

This paper analyzes the signaling efficiency for multimedia call (session) establishment in all-IP wireless networking infrastructures based on the Internet Multimedia Subsystem (IMS) standard framework and CDMA2000 Evolution Data Only Rev. A (EV-DOrA) standard for wireless link transmission. We present a comprehensive, bottom up analysis of signaling delay for setting up multimedia sessions and we evaluate system architecture alternatives to reduce it. Call setup time is particularly critical for interactive applications, such as gaming, push-to-X and Voice over IP (VoIP), where user experience of time that takes to start or join a session directly translates in user perception of service quality. Our analysis takes into account a large set of system design parameters across all layers, ranging from radio link properties to IMS processing and specific characteristics of the Session Initiation Protocol (SIP). For an example signaling flow between two mobile users, we derive an important set of numerical results for call set up time under various IMS design scenarios that consider effects of SIP compression efficiency, choice of transport protocols (TCP, UDP) and radio link quality. Based on the results obtained, we propose and discuss new methods to improve the signaling efficiency and to reduce the call setup time.

In this paper, we propose a multisegment optical network framework as a tool to solve a generalized category of problems related to end-to-end provisioning over interconnected optical networks. Based on the multisegment framework, we developed routing schemes for multigranularity multidomain optical networks. For examples of regional all-optical networks interconnected over an all-optical WDM backbone under a variety of traffic conditions, we present and compare numerical results. The performance results demonstrated the ability of our schemes to handle various network conditions under different control plane architectures.

We propose a new approach to constraint-based path selection for dynamic routing and wavelength allocation in optical networks based on WDM. Our approach considers service-specific path quality attributes, such as physical layer impairments, reliability, policy, and traffic conditions, and uses a flooding-based transfer of path information messages from source to destination to find multiple feasible paths. It is fully decentralized, as it uses local network state information. To better understand how multiple constraints impact the efficiency of wavelength routing, and consequently provision the service guarantees, we specifically focus on electronic regenerators that, while being widely considered as the basic building blocks for optical switching nodes, are likely to impose conflicting constraints on routing. For example, electronic regenerators extend the optical reach and could perform wavelength shifting, but also induce impairments, such as delays and operational costs. The question for constraint-based routing is how to account for these conflicting effects. To validate the network modeling, a wide range of networking scenarios are simulated, such as ring, mesh and interconnections of all-optical networks with electronic gateways. For all these scenarios, our approach is shown to efficiently accommodate multiple, conflicting routing metrics related to different services and network architectures.

In this paper we present three wavelength routing algorithms, "end-to-end", concatenated shortest path" and "hierarchical routing" for large-scale optical networks where optical paths traverse a number of networking segments interconnected by gateways. We analyze them with respect to the type of traffic, i.e. global vs. local and consider different requirements on wavelength routing based on gateway adaptation capabilities (wavelength merging, conversion or waveband interchange). Finally, for examples of regional all-optical networks interconnected over an all-optical WDM backbone under a variety of traffic conditions, we present and compare numerical results of blocking performances of different routing algorithms and gateway selection strategies.

This paper focuses on the issue of end-to-end service accommodation on the interfaces between user networks and multi-provider, multi-strategy optical transport network domains. We propose a new, multipoint-to-multipoint artifact called multi-provider federation interface (MPFI), which we believe is a valid architectural alternative to the UNI/NNI concept, if inter-domain administration, multiple service level agreements (SLA), as well as heterogeneity in network infrastructure characterize service provisioning. Two evolutionary scenarios for MPFI implementation are considered. In the first scenario, multiple networks create a federation, for which the collocation points are dictated by the physical topology of each single domain. For that scenario, we study the policy-based provisioning, given a choice of candidate domains to perform the provision. In the second scenario, we study the case where one large network domain splits into smaller domains along the same physical topology, while keeping the federated operation. In such scenario, we study the optimal placement of multi-provider interfacing points. Based on the solution for optimal MPFI installation, we finally analyze the improvement in blocking performance of the so-called end-to-end path provisioning crossing multiple provider domains based on WDM, with particular policy-based requirements on service accommodation.

We propose a framework for decentralized path selection and on-demand wavelength channel provisioning in WDM networks with routing constraints. Within this framework, the path information of choice, such as transmission quality, reliability, policy and traffic conditions, is updated following a connection request, based on a local, autonomous and service-differentiated characterization of optical network elements. It is this local and autonomous network state information that makes our approach particularly suitable for distributed implementation, and applicable to optical network architectures and control protocols that support service level guarantees. To illustrate this, we study mesh networks consisting of transparent, short-reach networking segments interconnected by long-reach WDM links with electronically regenerative gateways, where different link types show different SNR degradation, reliability and delay. On the example of electronically regenerative gateways, we address a new class of constraint-based path selection problems, where a certain a type of network element can at the same time deteriorate and improve network performance; e.g. electronic regenerators improve optical path quality and can adapt to wavelengths, but induce delays or operational costs. The performance study has shown the capability of our methods to accommodate arbitrary number and type of optical path properties, related to different network architectures and services.

We propose a routing methodology for multi-hop low Earth orbit (LEO) satellite networks aiming at minimising the number of link handovers per connection, while at the same time satisfying QoS requirements. For the representation of the network topology, an approach with a graph method is used, where a new metric for inter-satellite links, called lifetime, is introduced. This metric is then considered for minimisation of connection re-routing attempts due to the connectivity changes. In addition to this, we propose a QoS-based routing strategy for multi-service LEO satellite networks, for which the so-called per service performance is studied by numerical examples.

The need to establish wavelength-routed connections in a service-differentiated fashion is becoming increasingly important due to a variety of candidate client networks (e.g., IP, SDH, ATM) and the requirements for QoS-delivery within transport layers. The multiservice operation changes the way we deal with wavelength-routed paths, as they are now being characterized by manifold properties, such as transmission quality, restoration, network management, and policies. We propose a generic approach to service-differentiated connection accommodation in wavelength-routed networks where, for the network state representation, the supplementary network graphs are defined and referred to as service-specific wavelength-resource graphs. These graphs are used for the appropriate allocation of wavelengths on concatenated physical resources building a wavelength route, along which the necessary transmission quality is achieved and the required management and surveillance functions are provided. By considering twofold wavelength routing metrics, i.e., QoS metrics (service requirements) and resource metrics (quality constraints), these graphs can yield the solution to the QoS-routing problem, i.e., the provision of service-specific guarantees under quality constraints-a feature that is still missing from the existing architectures. The numerical analysis of dynamically reconfigurable multiservice WDM networks is presented for regular network operation as well as for optical network service restoration.

In this paper, resource allocation strategies in QoS-routed optical networks are proposed. We particularly focus on one QoS-routing approach, where various quality attributes, like transmission performance or blocking probability, directly influence the cost functions of the so-called transformed network graph. In this graph, the new nodes are derived from the network elements relevant to the decisions on quality for a certain user. The costs for the graph edges are compound functions of that quality and the particular resource allocation strategy, like least-quality resource allocation, wavelength shifting minimisation, or alternate routing for multi-wavelength network resources. Numerical results for multi-service optical networks have shown that the QoS-based resource allocation can simultaneously accommodate particular user requirements, blocking probability and traffic dynamics.