Dividing a network into clusters is a well-established strategy to address scalability challenges in data collection. With the advent of multi-access edge computing, unmanned aerial vehicles (UAVs) are increasingly being deployed to collect data from these clusters. For efficient and scalable UAVassisted data collection, the cluster formation, beyond being energy efficient, must also satisfy several critical properties, such as maintaining uniformity in cluster shape and size, ensuring appropriate placement of cluster heads, and forming compact, multi-hop structures. When nodes are mobile, clustering also needs to be fast enough to remain consistent despite topology dynamics. Existing approaches struggle to balance these diverse requirements effectively. We propose ReClub, a novel Concurrent-Transmission (CT)-based clustering framework that enables ultra-fast, energy-efficient, and topology-aware cluster formation for UAV-assisted data collection. Extensive evaluations demonstrate that ReClub is up to 92% faster than state-of-theart approaches, while sustaining reliable data collection under mobility. Moreover, ReClub-based data aggregation supports up to 7 × higher information refresh rates, making it highly suitable for next-generation mobile and large-scale edge-assisted IoT applications where 'age of information' is critical.

The integration ofDigital Twin(DT) into Metaverse plays a significant role in providing virtual replicas of the real world.Internet of Things(IoT) enables thePhysical Entities(PE) of the real-world to get appropriately mapped with their respective DTs. In recent works,Multi-Access Edge Computing(MEC)-based architecture shows faster mapping solution through UAV-based real-time data-collection from the PEs. Energy constraint in both the PEs and the UAVs is an important aspect in the whole process. Recent advancements in energy harvesting technologies directly benefit the PEs, the energy constraint in UAVs is still a stringent factor. Unfortunately existing works fail to address both the speed of data-collection as well as energy consumption of the UAVs together. In this work, we propose aConcurrent-Transmission(CT)-assisted mechanism, FETCH which addresses both of these factors together. In contrast to existing approaches, FETCH exploits CT to realize efficient coordination among the PEs so that a significant part of the data-collection can be done in parallel through the PE network. This helps the UAVs to significantly reduce their movements enhancing the utilization of their battery capacity. Evaluation shows that FETCH enables up to 88% faster and 89% energy efficient data-collection compared to the existing best known approaches. Moreover, it provides seamless support for change in topology of PEs. We also show that FETCH accomplishes a significantly higher rate of PE-DT mapping realizing superior quality of service.

Network slicing and Multi-access Edge Computing (MEC) are pivotal elements of 5G communication technology, enabling diverse, low-latency services to distributed users. Unmanned Aerial Vehicles (UAVs) are being increasingly explored in delivering these services temporarily to remote locations, supporting surveillance in regions with restricted ground connectivity, monitoring urban traffic, and disaster relief. However, the resource constraints of UAVs demand efficient optimization strategies. While Artificial Intelligence (AI)-driven methods like Deep Reinforcement Learning (DRL) offer promising potential in optimizing service delays and minimizing power consumption with fewer UAVs, they remain vulnerable to adversarial attacks. This study evaluates two adversarial attacks against DRL baselines: a targeted service disruption attack that impacts the DRL environment to degrade decision-making and service quality, and an action bit-flipping attack that alters UAV selection, resulting in suboptimal provisioning. To address these vulnerabilities, we propose ARCANE, a resilient DRL-based multi-slice MEC framework for UAVs. ARCANE introduces the Exploratory-Thompson Deep-Q Network (ET-DQN), which leverages Thompson Sampling to effectively balance exploration and exploitation under adversarial conditions, optimizing UAV selection for MEC provisioning. Extensive experiments demonstrate that ARCANE outperforms baseline approaches, achieving ~ 4× faster mitigation of the environmental attack and ~ 2× quicker recovery from the attack on the actions. Moreover, we illustrate that ARCANE demonstrates strong resilience by effectively limiting the degradation in hovering time caused by the attacks.

Network intrusion detection systems (NIDSs) face critical challenges from continuously evolving cyber-attacks. Traditional machine learning methods, while requiring extensive labeled training data, still often fail against unknown and out-of-distribution (OOD) attacks. Furthermore, new sophisticated adversaries are exploiting the detection blind spots inherent in traditional feature representation approaches that do not provide adequate comprehensive traffic analysis. In this paper, we propose MDCG-IDS, an NIDS framework that introduces multi-dimensional cross-granularity (MDCG) feature representation for open-set detection, in which network traffic is analyzed thoroughly across three complementary dimensions (traffic statistics, temporal, spatial), each at multiple granularity levels. These dimensions and granularities jointly capture the structures of sophisticated attacks that may be invisible from single analytical perspectives. We design a tensor structure that provides a unified encoding for the MDCG features while supporting the use of optimal transport theory to measure the distance between benign traffic and known or unknown attacks. MDCG-IDS uses a semi-supervised learning model that is trained exclusively on benign traffic and validated on a small set of labeled data, significantly reducing the effort of data labeling. Experiments on various datasets achieve AUC-ROC scores of more than 0.948, exceeding the best competing state-of-the-art methods by up to 7%. Regarding the amount of labeled validating data, MDCG-IDS obtains an AUC-ROC score of over 0.94 with only 3% of entire validating samples, outperforming the baseline models.

Explainable AI (XAI) enhances transparency in AI-driven applications but can also be exploited for malicious purposes. This paper investigates adversarial attacks that leverage XAI techniques to undermine Deep Reinforcement Learning (DRL)-based service provisioning in Unmanned Aerial Vehicle (UAV)-enabled Multi-access Edge Computing (MEC). Focusing on latency-sensitive services in dynamic, resource-constrained UAV networks, we propose two XAI-guided attacks: Salient Nodes Overload (SNO) and Subtle Nodes Drift (SND). These attacks exploit DRL explanations to degrade performance and mislead the agent. Experiments on UAV-enabled MEC frameworks using Double-Dueling Deep Q-Network (D3QN) and Deep Q-Network (DQN) show SNO causes immediate degradation, reducing job completion by about 40% in D3QN and 20% in DQN. The stealthier SND causes a delayed decline, as D3QN’s resilience defers deterioration for nearly twice as many episodes, complicating early detection. Both attacks reduce UAV hovering time, with SND causing up to 30% greater loss in D3QN.

A modern IoT-system, to accomplish various sophisticated goals, requires support for a variety of information management services, such as data-collection, gathering, routing and several others. There has been plenty of research on devising dedicated strategies to serve these goals. In recent works, Synchronous-Transmission (ST)-based mechanisms have shown promising solutions. However, in the context of low-power and highly resource-constrained IoT-edge, employing explicit dedicated solutions for multiple frequently encountered issues makes the system sufficiently complex as well as quite inefficient in terms of the memory and energy constraints of the IoT-devices. The availability of a topological structure in the system brings forth a way to accomplish versatile information management goals uniformly under a single hood. It also binds the nodes in a distributed fashion, with a grip over the entire system. However, the existing strategies to accomplish such hierarchical organization in an IoT-system is too slow to serve dynamic requirements. In this work, we propose an ST-based strategy, TurTle to accomplish the goal in real-time. Furthermore, we also design simple applications of TurTle to show its effectiveness. Extensive simulation-based studies demonstrate that TurTle builds the hierarchical structure upto 55 % faster than the existing bestknown approach.

Recent research works have proposed machine learning models for classifying IoT devices connected to a network. However, there is still a practical challenge of not having all devices (and hence their traffic) available during the training of a model. This essentially means, during the operational phase, we need to classify new devices not seen in the training phase. To address this challenge, we propose ZEST—a ZSL (zero-shot learning) framework based on self-attention for classifying both seen and unseen devices. ZEST consists of i) a self-attention based network feature extractor, termed SANE, for extracting latent space representations of IoT traffic, ii) a generative model that trains a decoder using latent features to generate pseudo data, and iii) a supervised model that is trained on the generated pseudo data for classifying devices. We carry out extensive experiments on real IoT traffic data; our experiments demonstrate i) ZEST achieves significant improvement (in terms of accuracy) over the baselines; ii) SANE is able to better extract meaningful representations than LSTM which has been commonly used for modeling network traffic.

Unmanned Aerial Vehicles (UAVs) are like modern weather surveyors, flying through the skies and collecting valuable atmospheric data with their instrumentation. However, collecting accurate timing data involves a delicate balance between energy conservation and high-speed operation across a variety of computing devices. In order to address this issue, we created a unique Federated Learning computer system designed for the development of weather forecasting using data collected from UAVs. This dynamic variation now reduces operational completion instances through outlier detection and a softmax weight allocation. In the long run, the weather forecast’s overall performance and effectiveness have been greatly improved. A splendid innovation in our framework is creating Federated Weighted Average (FedWAvg) set of rules, specifically designed to deal with delays due to outliers or inaccuracies at some stage in the discussion. FedWAvg allows quick convergence without compromising statistical accuracy, increasing the trustworthiness of weather forecasting in real-world conditions. Putting those advancements together, we have made significant improvements to make weather forecasting more reliable and useful, for the people and the businesses that rely on weather forecasting to get better rewards.

The advent of 5G has brought network slicing and Multi-access Edge Computing (MEC) to the fore, which has enabled provisioning of various services at the network edge. With Unmanned Aerial Vehicles (UAVs), these edge services can be offered in areas with limited accessibility to the traditional network, without compromising service requirements. Due to cost, power and resource constraints, it is desirable to let only a few UAVs provide a service. UAVs flying around to cover different zones may result in longer service delays to the ground User Equipments (UEs). On the other hand, hovering the UAVs over fixed zones and letting a UAV offload jobs to any other UAV would require line-of-sight communication among all UAVs. In our work, we address the above challenges and develop an efficient solution. In the proposed framework, MEC-Hopper, the UAVs hover over zones iteratively defined by clusters of ground UEs, with only a subset of the UAVs instantiating the services while others act as relays for offloading jobs, facilitating multi-hop communications. We develop a centralized Deep Reinforcement Learning (DRL) algorithm to select a subset of UAVs for hosting the services, and a multi-hop latency-aware job distribution algorithm to fairly distribute the jobs, providing acceptable low latency. Experimental results demonstrate MEC-Hopper’s ability to outperform a baseline static edge service provisioning framework in terms of the latencies incurred. Moreover, we exemplify the resilience of the proposed framework against a Denial-of-Service (DoS) attack, without necessitating re-training or implementing additional mitigation strategies. Additionally, MEC-Hopper’s longer hovering time demonstrates the advantages of choosing a subset of UAVs for service provisioning.

Unmanned Aerial Vehicles (UAVs) are integral in various sectors like agriculture, surveillance, and logistics, driven by advancements in 5G. However, existing research lacks a comprehensive approach addressing both data freshness and security concerns. In this paper, we address the intricate challenges of data freshness, and security, especially in the context of eavesdrop-ping and jamming in modern UAV networks. Our framework incorporates exponential AoI metrics and emphasizes secrecy rate to tackle eavesdropping and jamming threats. We introduce a transformer-enhanced Deep Reinforcement Learning (DRL) approach to optimize task offloading processes. Comparative analysis with existing algorithms showcases the superiority of our scheme, indicating its promising advancements in UAV network management.

Unmanned Aerial Vehicles (UAVs) play a vital role in modern Internet of Things (IoT) ecosystems by providing services like task offloading. Although simultaneous execution can be done through resource optimization and offloading, it is crucial to choose which task to be executed where considering a trade-off between energy consumption and execution delay. Considering these constraints, in this work, we propose a Federated Learning (FL) aided framework for multi-device task offloading in UAV-enabled edge networks while reducing energy consumption and task execution delay. The proposed approach involves a two-step process to execute tasks on various computing devices. In the first step, a task's priority is determined, considering factors like delay deadlines (maximum allowed delay) and resource requirements which include memory, storage and CPU instructions. In the second step, we leverage FL to dynamically calculate energy and delay of the task for different paradigms. This approach aims to maintain a balance between minimizing energy consumption by the UAV and reducing task execution delay. The effectiveness of the proposed framework is evaluated through experiments in terms of energy efficiency and end-to-end execution delay of the UAVs.

The technological advancements in Unmanned Aerial Vehicles (UAVs) have brought significant changes in various domains including surveillance, agriculture and disaster rescue. The convergence of Machine Learning (ML) and UAV networks contributes significantly to their automation and decision-making capabilities. Traditional ML techniques are centralized, i.e., they face many issues such as privacy due to data sharing, scalability and single-point failure. In this work, we propose a hierarchical decentralized framework for Federated Learning (FL) that addresses all the aforementioned issues. The proposed framework, Cluster Based Pseudo Hierarchical Decentralized Federated Learning (PHDFL), is tailored to UAV networks for learning where the learning and aggregation tasks are distributed among different UAVs in the network. This also introduces the concept of pseudo-hierarchy as all the UAVs are at the same level due to Decentralized Federated Learning (DFL) but the learning happens in a hierarchical manner where the network is divided into clusters and each cluster has a cluster head which in then communicates with other cluster heads. The effectiveness of the proposed framework is evaluated through experiments in terms of learning time, energy consumed and convergence of the model.

In recent times, Unmanned Aerial Vehicles (UAVs) have played a significant role in various fields like agriculture, defense, environmental monitoring, and many more. By minimizing the latency, energy consumption and improving the quality of services (QoS) while offloading the tasks, UAVs have played an outstanding role in the Internet of Things (IoT). Despite having a huge advantage, as the UAVs have limited computation, they cannot handle all the tasks that require intensive computation. To tackle the above problem, we have adopted the Deep Reinforcement Learning (DRL) technique in the UAV network, which helps in handling computationally expensive tasks by sharing the workload among the UAVs. The DRL-based strategy helps in reducing latency and energy consumption while maximizing the resource utilization of UAVs. Experiments have shown the significance of the adopted DRL strategy in reducing energy consumption by at least 16% compared to traditional algorithms.

Over the last few decades, Internet of Things (IoT) has become the spotlight area of research within the Industries and Academics. Primarily, IoT devices are characterized by small and nonscalable resources, including low processing capabilities, less internal memory, and short battery life. However, IoT applications demand extensive storage and faster response to ensure seamless and interoperable communication. Hence, Edge Computing and data/task offloading among the edge or cloud servers become promising, while also posing critical research challenges for edge-enabled large-scale IoT ecosystems. Several research activities have addressed the difficulties of determining an efficient and scalable data offloading strategy utilizing edge and cloud computing-supported technologies. This article focuses on the state-of-the-art edge IoT data offloading techniques, and optimization models in the heterogeneous IoT environment. We examine how edge and cloud-supported technologies can handle delay-sensitive IoT applications efficiently. Moreover, we introduce an IoT-based healthcare use case scenario to explain edge data execution and resource provisioning in IoT networks. Finally, we discuss several challenging issues and possible solutions to establish interoperable communication and computation for IoT applications.

Industry 5.0 aims to elevate industrial operations, businesses, and revolution to new heights by promoting sustainable, resilient, and human-centric practices. The popularity of Industry 5.0 is reflected in the increasing demand for real-time and near-edge processing in most latency-critical Industrial Internet of Things (IIoT) applications. However, designing an efficient task priority assignment strategy and accordingly executing tasks within the stipulated deadline is complex and challenging. Therefore, in this work, we design a novel Multi-device Edge Service Provisioning (MESP) framework for optimizing delay in Industry 5.0. At first, the MESP strategy classifies edge executable tasks using multi-nomial probability theory. Then, we prove that multi-device service demand at the edge devices is an NP-Hard problem, which requires approximate algorithms for finding near-optimal solutions. To follow this, we propose a game-theoretic approach where multiple IIoT devices request various services simultaneously while maximizing their mutual satisfaction. We also examine the structural property of the proposed game and show how this property helps in achieving the equilibrium point of the proposed game with finite improvement steps. Experimental analysis shows that MESP reduces computational overhead and end-to-end execution delay by 20-30% compared to standard algorithms.

The IPv6 over time-slotted channel hopping mode of IEEE 802.15.4e (6TiSCH) wireless protocol stack is released to offer high-throughput, low and bounded latency, energy efficient, and reliable communication in industrial Internet of Things (IoT). However, scheduling communicationcellamong the nodes for exchanging sensory data is not trivial in 6TiSCH networks when the network traffic is highly dynamic and unpredictable. The existing autonomous scheduling schemes suffer from static allocation, high end-to-end latency, and high energy consumption. To address the abovementioned problems, in this work, we proposeon-the-fly autonomous slot allocation(OASA)scheme to schedule slots for adaptive traffic in 6TiSCH networks autonomously and immediately. OASA also enables a lowradio-duty-cycleof the nodes when network traffic is less, which is not considered by any existing adaptive autonomous schedulers. To validate the effectiveness of OASA, we implemented it on Contiki-NG and performed testbed experiments on FIT IoT-LAB. The testbed experiment results demonstrate the effectiveness of OASA in terms of latency, packet delivery ratio, and energy consumption compared to the existing autonomous scheduling schemes.

Attack-resilience is essential to maintain continuous service availability in Internet of Vehicles (IoV) where critical tasks are carried out. In this paper, we address the problem of service outage due to attacks on the edge network and propose an attack-resilient mapping of vehicles to edge nodes that host different types of service instances considering resource efficiency and delay. The distribution of service requests (of an attack-affected edge node) to multiple attack-free edge nodes is performed with an optimal vehicle-to-edge (V2E) mapping. The optimal mapping aims to improve the user experience with minimal delay while considering fair usage of edge capacities and balanced load upon a failure over different edge nodes. The proposed mapping solution is used within a deep reinforcement learning (DRL) based framework to effectively deal with the dynamism in service requests and vehicle mobility. We demonstrate the effectiveness of the proposed mapping approach through extensive simulation results using real-world vehicle mobility datasets from three cities.

Network slicing is the critical enabler for next-generation mobile networks, which divides the infrastructure into multiple logical networks known as slices. Each logical network supports services with specific throughput and latency requirements. The fifth-generation(5G) and 5G-beyond networks employ more than two slices; hence, it has become necessary to deploy algorithms for efficient resource allocation. However, given the latency-sensitive applications, allocating resources to different slices based on the current demand would be a poor choice. Therefore, resource allocation needs to be performed in advance, which calls for forecasting algorithms predicting future demands. In this paper, we propose a novel predictive network slicing mechanism named NPRA to predict resources dynamically. NPRA predicts the future resource requirements using Unit Time Long Short-Term Memory (UT-LSTM). The predicted demand can be used as an input to any optimization algorithm for the timely allocation of resources. We also develop a 5G simulation testbed to generate datasets for performance study. The results presented demonstrate the effectiveness of NPRA.

In this work, we investigate an online service management problem in vehicular edge computing networks. To satisfy the varying service demands of mobile vehicles, a service management framework is required to make decisions on the service lifecycle to maintain good network performance. We describe the service lifecycle consists of creating an instance of a given service (scale-out), moving an instance to a different edge node (migration), and/or termination of an underutilized instance (scale-in). In this paper, we propose an efficient online algorithm to perform service management in each time slot, where performance quality in the current time slot, the service demand in future time slots, and the minimal observed delay by vehicles and the minimal migration delay are considered while making the decisions on service lifecycle. Here, the future service demand is computed from a gated recurrent unit (GRU)based prediction model, and the network performance quality is estimated using a deep reinforcement learning (DRL) model which has the ability to interact with the vehicular environment in real-time. The choice of optimal edge location to deploy a service instance at different times is based on our proposed optimization formulations. Simulation experiments using real-world vehicle trajectories are carried out to evaluate the performance of our proposed demand-prediction based online service management (DOSM) framework against different state-of-the-art solutions using several performance metrics.

Network Function Virtualization (NFV) has become a promising technology, which is used to replace the complex hardware implementation of various Network Functions. Virtual Network Functions (VNFs) are placed on the servers to realize these functions of a network, for example 5G network. Such an implementation reduces the expenditure and latency significantly. However, it also comes with its own set of challenges. One of the prominent challenges of NFV is the placement problem. The requests are in the form of a Service Function Chain (SFC), which is a pre-defined sequence of VNFs. When an SFC request comes in, resources must be allocated to the VNF instances to satisfy the requirements. In this paper, we propose an effective algorithm D3T (Double Deep Q-Network Decision Transformer) to optimize the SFC placement. The algorithm is designed using a Decision Transformer (DT) that is assisted by a Double Deep Q-Network (DDQN). We employ a DDQN model as the baseline algorithm to generate offline training data. The trajectory data in the Experience Reply Memory of DDQN will be processed into sequences and modeled by a transformer. The algorithm’s objective function considers end-to-end delay and rejection ratio as the objectives. Specifically, D3T combines the transformer with DRL to give an optimal solution. The results presented demonstrate the effectiveness of the proposed solution.

The service-based architecture (SBA) of 5G Core (5GC) introduces significant landscape changes to the modern communication and network system, and the network slicing enables different Network Functions (NFs) to meet diverse service requirements. However, with the broadening interface, some key NFs may become more vulnerable to internal hostile NFs or external malicious entities, which pose severe threats to the control-plane components in the 5GC network (5GCN). In this paper, we propose ADSeq-5GCN, a network-level anomaly detection framework based on modeling network traffic sequences. Our framework focuses on the control plane of 5GCN, where the network traffic is captured and analyzed for anomalies. We use a sequence model, Bidirectional Long Short Terms Memory (Bi-LSTM) networks, to learn normal NF-to-NF interactions and detect anomalies based on incorrect service event prediction. We evaluate our proposed framework on a 5GCN testbed with Free5GC and UERANSIM under various scenarios. Our results demonstrate the overwhelming performance of our proposed framework over the baseline models.

The 6TiSCH communication architecture provides delay-bounded packet delivery, energy efficient, and reliable data-delivering communication in mission-critical Internet of Things (IoT) applications. It uses IETF’s 6TiSCH minimal configuration (6TiSCH-MC) standard for resource allocation during network formation and routing using a routing protocol for low power and lossy network (RPL) as routing protocol. In RPL, the DODAG information solicitation (DIS) control packet is used to solicit routing information from the existing networks. However, it is observed that malicious transmission of this DIS packet can severely affect the 6TiSCH networks in terms of nodes’ network joining time and energy consumption. Therefore, designing countermeasures of DIS attack in 6TiSCH network has become critically important. Additionally, the existing works neither considered all the possible parameters together for detecting DIS attack nor energy efficient, and create control packet overhead. In this work, we model noncooperative gaming to determine the optimal probability of responding to a DIS packet. Subsequently, we design a trust model to detect malicious DIS transmission in 6TiSCH networks. Finally, we merge both the proposed gaming model and trust model to propose a scheme—gaming and trust-based countermeasure (GTCM) to reduce the effect of DIS attack in 6TiSCH networks. We implement the GTCM on Contiki-NG and validate it using open source FIT IoT-LAB testbed. Our experimental testbed results show that GTCM reduces the effect of DIS attack in terms of pledges’ (new nodes) joining time and energy consumption significantly.

Achieving network resilience in terms of attack tolerance and service availability is critically important for Internet of Vehicles (IoV) networks where vehicles require assistance in sensitive and safety-critical applications like driving. It is particularly challenging in time-varying conditions of IoV traffic. In this paper, we study an attack-resilient optimal service placement problem to ensure disruption-free service availability to the users in edge-enabled IoV network. Our work aims to improve the user experience while minimizing the delay and simultaneously considering efficient utilization of limited edge resources. First, an optimal service placement is performed while considering traffic dynamicity and meeting the service requirements with the use of a deep reinforcement learning (DRL) framework. Next, an optimal secondary mapping and service recovery placements are performed to account for the attacks/failures at the edge. The use of DRL framework helps to adapt to dynamically varying IoV traffic and service demands. In this work, we develop three integer linear programming (ILP) models and use them in the DRL based framework to provide attack-resilient service placement and ensure service availability with efficient network performance. Extensive numerical experiments are performed to demonstrate the effectiveness of the proposed approach.

Networks regularly face various threats and attacks that manifest in their communication traffic. Recent works proposed unsupervised approaches, e.g., using a variational autoencoder, that are not only effective in detecting anomalies in network traffic, but also practical as they do not require ground truth or labeled data. However, the problem of characterizing anomalies into different attack behaviors is still less explored; in this work, we study this specific problem. We develop APEX, a framework that employs data mining approaches in a semisupervised way to extract the attack patterns from anomalous traffic and links them to specific attack types. APEX comprises two levels of mining: the first level extracts patterns in anomalous network flows, and the second level characterizes behaviors in the extracted patterns into different attack classes. We carry out extensive experiments on real network traces obtained from the MAWI traffic archive. The evaluations demonstrate that APEX is effective in extracting distinguishable behaviors of network attacks from anomalous traffic, and provide useful insights to security analysts investigating the anomalies.

Password-based authentication system has been praised for its user-friendly, cost-effective, and easily deployable features. It is arguably the most commonly used security mechanism for various resources, services, and applications. On the other hand, it has well-known security flaws, including vulnerability to guessing attacks. Present state-of-the-art approaches have high overheads, as well as difficulties and unreliability during training, resulting in a poor user experience and a high false positive rate. As a result, a lightweight authentication compromise detection model that can make accurate detection with a low false positive rate is required.In this paper we propose – LOG-OFF – a behavior-based authentication compromise detection model. LOG-OFF is a lightweight model that can be deployed efficiently in practice because it does not include a labeled dataset. Based on the assumption that the behavioral pattern of a specific user does not suddenly change, we study the real-world authentication traffic data. The dataset contains more than 4 million records. We use two features to model the user behaviors, i.e., consecutive failures and login time, and develop a novel approach. LOG-OFF learns from the historical user behaviors to construct user profiles and makes probabilistic predictions of future login attempts for authentication compromise detection. LOG-OFF has a low false positive rate and latency, making it suitable for real-world deployment. In addition, it can also evolve with time and make more accurate detection as more data is being collected.

The growth of fifth-generation (5G) and edge computing has enabled the emergence of Internet of Vehicles (IoV). It supports different types of services with different resource and service requirements. However, limited resources at the edge, high mobility of vehicles, increasing demand, and dynamicity in service request types have made service placement a challenging task. A typical static placement solution is not effective as it does not consider the traffic mobility and service dynamics. Handling dynamics in IoV for service placement is an important and challenging problem which is the primary focus of our work in this article. We propose a deep reinforcement learning-based dynamic service placement (DRLD-SP) framework with the objective of minimizing the maximum edge resource usage and service delay while considering the vehicle’s mobility, varying demand, and dynamics in the requests for different types of services. We use SUMO and MATLAB to carry out simulation experiments. The experimental results show that the proposed DRLD-SP approach is effective and outperforms other static and dynamic placement approaches.

The emergence of technologies such as 5G and mobile edge computing has enabled provisioning of different types of services with different resource and service requirements to the vehicles in a vehicular network. The growing complexity of traffic mobility patterns and dynamics in the requests for different types of services has made service placement a challenging task. A typical static placement solution is not effective as it does not consider the traffic mobility and service dynamics. In this paper, we propose a reinforcement learning-based dynamic (RL-Dynamic) service placement framework to find the optimal placement of services at the edge servers while considering the vehicle’s mobility and dynamics in the requests for different types of services. We use SUMO and MATLAB to carry out simulation experiments. In our learning framework, for the decision module, we consider two alternative objective functions - minimizing delay and minimizing edge server utilization. We developed an integer linear programming (ILP) based problem formulation for the two objective functions. The experimental results show that 1) compared to static service placement, RL-based dynamic service placement achieves fair utilization of edge server resources and low service delay; and 2) compared to delay-optimized placement, server utilization-optimized placement utilizes resources more effectively, achieving higher fairness with lower edge-server utilization.

The classification of Internet-of-Things (IoT) devices into different types is of paramount importance, from multiple perspectives, including security and privacy aspects. Recent works have explored machine learning techniques for fingerprinting (or classifying) IoT devices, with promising results. However, the existing works have assumed that the features used for building the machine learning models are readily available or can be easily extracted from the network traffic; in other words, they do not consider the costs associated with feature extraction. In this work, we take a more realistic approach, and argue that feature extraction has a cost, and the costs are different for different features. We also take a step forward from the current practice of considering the misclassification loss as a binary value, and make a case for different losses based on the misclassification performance. Thereby, and more importantly, we introduce the notion of risk for IoT device classification. We define and formulate the problem of cost-aware IoT device classification. This being a combinatorial optimization problem, we develop a novel algorithm to solve it in a fast and effective way using the cross-entropy (CE)-based stochastic optimization technique. Using traffic of real devices, we demonstrate the capability of the CE-based algorithm in selecting features with minimal risk of misclassification while keeping the cost for feature extraction within a specified limit.

The fast-growing Internet-of-Things (IoT) market has opened up a large threat landscape, given the wide deployment of IoT devices in both consumer and commercial spaces. Attacks on IoT devices generally consist of multiple stages and are dispersed spatially and temporally. These characteristics make it challenging to detect and identify the attack stages using solutions that tend to be localized in space and time. In this work, we present Adept, a distributed framework to detect and identify the individual attack stages in a coordinated attack. Adept works in three phases. First, network traffic of IoT devices is processed locally for detecting anomalies with respect to their benign profiles. Any alert corresponding to a potential anomaly is sent to a security manager, where aggregated alerts are mined, using frequent itemset mining (FIM), for detecting patterns correlated across both time and space. Finally, using both alert-level and pattern-level information as features, we employ a machine learning approach to identify individual attack stages in the generated alerts. We carry out extensive experiments, with emulated and realistic network traffic; the results demonstrate the effectiveness of the proposed framework in terms of its ability in attack-stage detection and identification.

Software defined networking (SDN) is gaining the confidence of network operators, who are increasingly motivated to introduce it in their networks. However, since SDN is based on centralized control plane (decoupled from the data plane), it is incompatible with distributed control plane of legacy networks. As adopting SDN by complete overhaul of existing legacy networks is infeasible, several solutions propose to incrementally adopt SDN in legacy networks, resulting in hybrid SDN/legacy networks during the transition phase. While strategies to effectively introduce SDN switches in legacy networks have received significant attention in the literature, the same is not true for the associated SDN controller placement in hybrid SDN networks. In this paper, we introduce and formulate the multi-objective control plane dimensioning problem in hybrid SDN/legacy networks, considering network resilience, flow-setup latency and controller load balancing as objectives. We propose optimization models for both the single-objective as well as the multi-objective controller placements. We model both controller processing latency (based on queueing theory) and inter-controller network state synchronization latency (based on Steiner trees) without compromising linearity of the optimization formulations. A genetic algorithm-based heuristic is presented to efficiently deduce the approximate Pareto frontier for our problem. Extensive simulations over a large number of real networks establish the effectiveness of our approach in terms of several single-objective and multi-objective performance metrics.

The separation of control and data plane in Software Defined Networking (SDN) introduces new security threats. A compromised controller can leverage its position to perform attacks by installing malicious rules in switches while avoiding detection. Current approaches propose broadcast of flow-setup requests to multiple controllers simultaneously and to check consistency of forwarding rules to install the correct rule and identify compromised controllers. However, such approaches result in heavy load on the control plane, resulting in longer response times to requests and higher network cost to accommodate the increased load. To alleviate this issue, we propose a game-theory based framework to detect a malicious controller without overloading the control plane. Instead of broadcasting every request to multiple controllers, switches randomly broadcast requests on the basis of a randomization strategy obtained by the Stackelberg game, whose solution results in a randomization strategy that maximizes the detection probability of a malicious controller. We formulate a two-level optimization problem in the context of our game-theoretic framework that aims to maximize the attack detection probability among the set of controllers by mapping switches to controllers and obtaining randomization strategies for each controller. We developMidas(MalIcious controller Detection mApping Strategy), a heuristic algorithm to obtain an effective solution to the optimization problem in reasonable time.Midasachieves minimum detection probability within 12% of the optimal solution. Further, it achieves at least 80% of min-max ratio of load at the controllers, implying higher fairness in load distribution compared to optimal solution, a state-of-art algorithm and a baseline heuristic.

With the advances in modern communication technologies, the application scale of Internet of Things (IoT) has evolved at an unprecedented level, which on the other hand poses threats to the IoT ecosystem. As the intrusions and malicious actions are becoming more complex and unpredictable, developing an effective anomaly detection system, considering the distributed nature of IoT networks, remains a challenge. Moreover, the lack of sufficiently large amount of data samples of IoT traffic and data privacy pose further challenges in developing a behavior-based anomaly detection system. To address these issues, we present an unsupervised hierarchical approach for anomaly detection through cooperation between generative adversarial network (GAN) and auto-encoder (AE). The problems of data aggregation and privacy preservation are addressed by reconstructing a sampling pool at a centralized controller using a collection of generators from the individual IoT networks. Then, a centralized global AE is trained and passed to individual local networks for anomaly detection after a final adaptation with the local raw data from the IoT nodes. The performance is evaluated using the UNSW Bot-IoT dataset and the results demonstrate the effectiveness of our proposed approach which outperforms other approaches.

In a distributed Software Defined Networking (SDN) architecture, the Quality of Service (QoS) experienced by a traffic flow through an SDN switch is primarily dependant on the SDN controller to which that switch is mapped. We propose a new controller-quality metric known as the Quality of Controller (QoC) which is defined based on the controller's reliability and response time. We model the controller reliability based on Bayesian inference while its response time is modelled as a linear approximation of the M/M/1 queue. We develop a QoC-aware approach for solving (i) the switch-controller mapping problem and, (ii) control traffic distribution among the mapped controllers. Each switch is mapped to multiple controllers to enable resilience with the switch-controller mapping and control traffic distribution based on the QoC metric which is the combined cost of controller reliability and response time. We first develop an optimization programming formulation that maximizes the minimum QoC among the set of controllers to solve the above problem. Since the optimization problem is computationally prohibitive for large networks, we develop a heuristic algorithm - Qoc-Aware switch-coNTroller Mapping (QuANTuM) - that solves the problem of switch-controller mapping and control traffic distribution in two stages such that the minimum of the controller QoC is maximized. Through simulations, we show that the heuristic results are within 18% of the optimum while achieving a fair control traffic distribution with a QoC min-max ratio of up to 95%.

A 5G network slice is comprised of several service chains (SCs) - a chain of virtualized network functions (VNFs) each performing a function (e.g., firewall) to serve diverse traffic requirements. These VNFs are hosted on servers/physicalmachines in one or more data-centers. In this context, a fault/attack on a 5G network could be as fine grained as a VNF failure or as coarse-grained as the failure of a hosting physicalmachine (PM). The latter may lead to multiple VNF failures that may affect several SCs in one or more network slices. To enable resilience towards such failure/attacks, one approach is to protect all the SCs by replicating the VNFs on different PMs. However, such a protection approach will significantly increase the computational and operational costs opposing the 5G network objective. In this paper, we propose an approach for resilient network slice embedding such that the VNFs of a service chain are instantiated on appropriate PMs to minimize the number of affected SCs upon a PM failure. We develop an optimization programming formulation that minimizes the maximum of the affected number of weighted service chains while satisfying the VNF placement and routing constraints in the co located 5G network slices along with resource (bandwidth, computing) and slice-specific requirements. We evaluate the proposed resilient mapping approach through numerical analysis and show that it achieves a resilience factor of up to 0.88 while requiring 78% less number of VNF replicas on an average, when compared to the protection-based approach.

Recent developments in intelligent transport systems (ITS) based on smart mobility significantly improves safety and security over roads and highways. ITS networks are comprised of the Internet-connected vehicles (mobile nodes), roadside units (RSU), cellular base stations and conventional core network routers to create a complete data transmission platform that provides real-time traffic information and enable prediction of future traffic conditions. However, the heterogeneity and complexity of the underlying ITS networks raise new challenges in intrusion prevention of mobile network nodes and detection of security attacks due to such highly vulnerable mobile nodes. In this paper, we consider a new type of security attack referred to as crossfire attack, which involves a large number of compromised nodes that generate low-intensity traffic in a temporally coordinated fashion such that target links or hosts (victims) are disconnected from the rest of the network. Detection of such attacks is challenging since the attacking traffic flows are indistinguishable from the legitimate flows. With the support of software-defined networking that enables dynamic network monitoring and traffic characteristic extraction, we develop a machine learning model that can learn the temporal correlation among traffic flows traversing in the ITS network, thus differentiating legitimate flows from coordinated attacking flows. We use different deep learning algorithms to train the model and study the performance using Mininet-WiFi emulation platform. The results show that our approach achieves a detection accuracy of at least 80%.

With the proliferation of network devices and rapid development in information technology, networks such as Internet of Things are increasing in size and becoming more complex with heterogeneous wired and wireless links. In such networks, link faults may result in a link disconnection without immediate replacement or a link reconnection, e.g., a wireless node changes its access point. Identifying whether a link disconnection or a link reconnection has occurred and localizing the failed link become a challenging problem. An active probing approach requires a long time to probe the network by sending signaling messages on different paths, thus incurring significant communication delay and overhead. In this paper, we adopt a passive approach and develop a three-stage machine learning-based technique for link fault identification and localization (ML-LFIL) by analyzing the measurements captured from the normal traffic flows, including aggregate flow rate, end-to-end delay, and packet loss. ML-LFIL learns the traffic behavior in normal working conditions and different link fault scenarios. We train the learning model using support vector machine, multilayer perceptron, and random forest. We implement ML-LFIL and carry out extensive experiments using Mininet platform. Performance studies show that ML-LFIL achieves high accuracy while requiring much lower fault localization time compared to the active probing approach.

Identifying IoT devices connected to a network has multiple security benefits, such as deployment of behavior-based anomaly detectors, automated vulnerability patching of specific device types, dynamic attack mitigation, etc. In this paper, we look into the problem of IoT device identification at network level, in particular from an ISP's perspective. The simple solution of deploying a supervised machine learning algorithm at a centralized location in the network neither scales well nor can identify new devices. To tackle these challenges, we propose and develop a distributed device fingerprinting technique (DEFT), a distributed fingerprinting solution that addresses and exploits the presence of common devices, including new devices, across smart homes and enterprises in a network. A DEFT controller develops and maintains classifiers for fingerprinting, while gateways located closer to the IoT devices at homes perform device classification. Importantly, the controller and gateways coordinate to identify new devices in the network. DEFT is designed to be scalable and dynamic-it can be deployed, orchestrated, and controlled using software-defined networking and network function virtualization. DEFT is able to identify new device types automatically, while achieving high accuracy and low false positive rate. We demonstrate the effectiveness of DEFT by experimenting on data obtained from real-world IoT devices.

Cloud data centers nowadays play an important role in providing computing and network resources for online applications and services. Such applications obtain cloud resources by submitting resource requests in the form of virtual networks that are embedded in the cloud infrastructures, referred to as virtual network embedding (VNE). Developing an effective VNE algorithm is crucial since it affects the performance of data centers, such as rejection ratio, resource utilization, and revenue. The problem is much more challenging when considering ring optical data centers due to multiple issues: wavelength continuity constraint, wavelength selection, and physical path selection for a lightpath. In this paper, we first develop an optimization programming formulation, which is computationally prohibitive. We then develop a novel VNE algorithm that adopts the Web page ranking approach to evaluate the goodness of a top-of-the-rack (ToR)-based on its resources in correlation with that of other ToRs. We also develop efficient methods for wavelength and physical path selections for a lightpath dynamically created during the embedding. We evaluate the proposed algorithm through comprehensive simulations in comparing with the optimal results and baseline algorithms. The simulation results show that the proposed algorithm performs close to the optimal one. It significantly reduces the rejection ratio by at least 23% compared to the baseline algorithms, leading to an increase in revenue by at least 14%.

Given the rate at which communication technologies and protocols evolve, network operators are often cautious of fully migrating to a new technology like Software Defined Networking (SDN) at one go, and prefer to do so in phases. Consequently, the number of SDN switches and in turn the amount of SDN control traffic in the network increases gradually. This control traffic processing occurs at the SDN controller(s), which are, hence, required incrementally in such hybrid SDN/legacy networks. The placement of these controllers significantly affects several aspects such as control latency, resiliency, and load balancing. All existing controller placement strategies place controller(s) in a pure SDN network at a point in time, which is contrary to the SDN predominant migration scenario described above. Even when suitably adapted for hybrid networks, the existing controller placement strategies lead to inefficient placements. In this paper, we introduce and formulate the controller placement problem for hybrid SDN/legacy networks over a period of time, with an aim to maximize the switch-controller control channel resilience. We consider an SDN migration trajectory and deduce a resilient controller placement schedule using an optimization approach. Based on 138 real network topologies, we comprehensively evaluate our approach against a well-known existing resilient controller placement strategy (suitably adapted herein for hybrid SDN/legacy networks for a fair comparison), and demonstrate that our approach is more effective with up to 77% higher resiliency, while requiring up to 33% fewer controllers.

In software-defined networks, distributed controller architectures provide improved scalability and reliability by using multiple controllers, each managing a partition of the network. However, due to the dynamics of network control traffic, static switch-controller mapping causes load imbalance while dynamic mapping causes frequent switch migrations among controllers. In this paper, we present a novel game-theoretic switch-controller mapping approach that considers control traffic variations in distributed-controller software-defined networks. We formulate the problem as a Markov decision process that is a non-cooperative stochastic game in which the players are controllers. They compete to serve the switches within their processing capacity so as to maximize their reward based on the amount of traffic processed and their per-unit price while trying to reduce switch migrations due to traffic variations. We show the existence of a Markov perfect equilibrium for the game. We evaluate the performance of the proposed approach through comprehensive simulations, including comparisons with other alternatives. The results show that the switch-controller mapping solution obtained by the proposed approach is stable against the control traffic dynamics with good load balancing among controllers.

Software Defined Networking (SDN) enables programmability and flexibility in networks through the separation of control and data plane. SDN architecture poses new security threats to the network, especially in the control plane. A compromised controller can exhibit malicious behavior such as fraudulent rule installation while avoiding detection. Existing approaches deal with this issue by broadcasting every flow setup request to multiple controllers to check for forwarding rule consistency. This imposes heavy load on the control plane, leading to longer response time and increased cost. We propose a novel approach that can effectively detect a malicious controller without overloading the control plane. The proposed game-theoretic approach randomly selects switches to check for consistency of forwarding rules. We model the problem as a Stackelberg game and solve the corresponding optimization problem to obtain an effective randomization strategy. We also develop a heuristic algorithm to determine a set of actions for the defender to make the problem tractable. We consider load management at the controller and relative importance of switches while finding an optimal strategy for randomized checking. In comparison with three other heuristic approaches, our approach achieves up to 88% improvement in probability of detecting the malicious controller.

In almost a decade, Software Defined Networking (SDN) has transitioned from university laboratories to real-life networks. As networks evolve from legacy networks to hybrid OSPF/SDN networks to pure SDN networks, the associated challenge of appropriate controller placement - i.e. optimal count and location of controllers in an SDN network - need to be suitably addressed. Several studies investigated the controller placement problem at a point in time primarily from a latency minimization approach, in addition to various secondary objectives. However, considering long-term network growth, it is also important to evaluate the best timing to introduce a controller in an SDN network. In this paper, we introduce and formulate the incremental (multi- period) controller placement problem to add controllers in an SDN network in a phased manner over a finite planning horizon. We consider a practical scenario with increasing network traffic, as well as falling capital and operational expenditures of SDN controllers on account of vendor competition and technology maturity. We propose a multi-objective optimization model to derive the optimal multi- period controller roll-out plan. Using extensive simulations, we evaluate our solution in terms of various performance metrics and find that our multi-period placement schedule attains significant cost savings over single-period placement schedule at the cost of a small and acceptable increase in switch-controller latency.

Software-defined networking (SDN) is one of the most promising topics of research today. In this paper, we study the performance of three SDN controllers - ONOS, OpenMUL, and POX, that are implemented in three different programming languages, i.e., Java, C, and Python, respectively, using Mininet and Wireshark packet analysis. We compare the performance of the controllers analyzed using Mininet-Wireshark packet analysis with that of the benchmarking tool Cbench (that uses fake control packets generated from switch instances) in terms of latency and throughput and show that the latter significantly under-estimates the performance evaluation. Additionally, we also study a new performance metric, topology discovery time, that none of the current benchmarking tools capture.

In SDN, to enable resilience each switch can be mapped to multiple controllers (e.g., primary, secondary) and distribute the control traffic among them. The SDN controllers for each switch need to be chosen in a manner that does not compromise on network response time, while taking controller reliability into consideration. Given the switch-primary controller mapping, the objective of our work is to find an optimal switch-secondary controller mapping and control traffic distribution such that the reliability of mapped controllers is maximized. We mathematically model controller reliability, formulate and solve the optimization problem with the above objective. We implement a switch-controller mapping approach in SDN, wherein we map each switch to two active controllers and distribute the flow-setup requests (control traffic) between them. We implement our approach using Floodlight controllers and Mininet, and show that it improves reliability of mapping and achieves better recovery time upon controller failure by up to 27.3% as compared to the traditional master-slave setup.

The scale of connected devices in the modern communication network and its heterogeneous nature have made securing the network more challenging. However, with the advent of software defined networking (SDN), the algorithmic complexity is handled at a centralized control plane and the network elements perform only data forwarding based on control plane decisions. This enables researchers to design innovative security protocols at the control plane to dynamically defend against attacks. In this paper, we propose a dynamic attack-resilient routing (ARR) approach and develop an optimization formulation for fragmented multipath routing taking reliability and load into consideration for SDN-enabled networks. Though erasure encoding has been well studied for resilient data storage, it is rarely mentioned in the context of network routing owing to its complexity, redundancy, and difficulty of satisfying practical routing constraints. In this paper, we dynamically determine the optimal route for erasure-encoded fragments of the data, in terms of attack resilience, under the constraint on allowable encoding redundancy. Since the ARR algorithm is computationally prohibitive for larger networks, we develop a heuristic solution for the same using a multipath-tree. The proposed algorithm dynamically routes the data fragments along a set of reliable and lightly loaded paths to achieve multipath diversity and thereby improve data availability at the destination even in the presence of attacks. We demonstrate the effectiveness of our proposed approach in terms of weighted path reliability, resilience, and blocking performance through simulations.

The advances in fiber-optic technology and wavelength division multiplexing technique have led to the adoption of optical networks into cloud data centers to meet the ever-growing traffic demands. The hybrid optical-electrical data center network architecture becomes a transition solution from the all-electrical network architecture to the all-optical one since it is efficient in terms of the operational cost and resource efficiency in data centers. Network traffic can be routed through either the electrical network or optical network or migrated between the networks. Selection of the target network (optical or electrical) for embedding and decision of migration are challenging problems for providers since it depends on multiple inter-related constraints such as wavelength availability, computing resource availability and residual bandwidth on the links. A sub-optimal decision will affect the overall performance of data centers in terms of blocking ratio and bandwidth efficiency. We address the problem of virtual network embedding and migration in hybrid data centers in this paper. We first develop an optimization programming formulation that gives the embedding solution with target networks for a given set of virtual networks while maximizing the total bandwidth utilized. Since the formulation is computationally prohibitive, we then propose a heuristic algorithm, namely Time and Bandwidth-Aware Virtual Network Embedding and Migration (TBA-VNEM), that efficiently embeds and migrates virtual networks in hybrid data centers so as to improve bandwidth efficiency. We evaluate the proposed algorithm through comprehensive simulations. The results show that the proposed algorithm outperforms baseline algorithms by reducing the rejection ratio by up to 52% and increasing the bandwidth efficiency of the optical network alone by up to 20%.

Graph Processing has been widely used to capture complex data dependency and uncover relationship insights. Due to the ever-growing graph scale and algorithm complexity, distributed graph processing has become more and more popular. In this paper, we investigate how to balance performance and cost for large scale graph processing on configurable virtual machines (VMs). We analyze the system architecture and implementation details of a Pregel-like distributed graph processing framework and develop a system-aware model to predict the execution time. Consequently, cost effective execution scenarios are recommended by selecting a certain number of VMs with specified capability subject to the predefined resource price and user preference. Experiments using synthetic and real world graphs have verified that system-aware model can achieve much higher prediction accuracy than popular machine-learning models which treat graph processing framework as a black box. As a result, the recommended execution scenarios have comparable cost efficiency to the optimal scenarios.

Micro-blogging services like Twitter which allow users to post messages and follow activities are gaining in popularity. The content of the posted tweets is wide ranging, and sometimes includes private information like email addresses, physical addresses, birthdays and medical history. Such private data, if leaked through public posts, could be used by stalkers, foes, or unintended parties. Detecting the presence of private data in tweets is a first step towards analyzing the privacy risk associated with it. In this context, the purpose of our work is twofold. First, we categorize the tweets into private and non-private, based on whether they reveal any private information or not. Second, we try to gain more insights into categorized private tweets by identifying the type of private data being revealed. We train the model on novel features extracted from the labeled tweets and perform supervised classification. Our results show that detection of leak of private data can be achieved with an accuracy of about 80% and false positive rate of 18%. Furthermore, we are able to differentiate between private tweets by classifying them into different categories with high accuracy.

Security in Software Defined Networks (SDNs) has been a major concern for its deployment. Byzantine threats in SDNs are more sophisticated to defend since control messages issued by a compromised controller look legitimate. Applying traditional Byzantine Fault Tolerance approach to SDNs requires each switch to be mapped to 3f + 1 controllers to defend against f simultaneous controller failures. This approach, on one hand, overloads the controllers due to multiple requests from switches. On the other hand, it raises new challenges concerning the switch-controller mapping and determining minimum number of controllers required in the network. In this paper, we present a novel primary-backup controller mapping approach in which a switch is mapped to only f + 1 primary and f backup controllers to defend against simultaneous Byzantine attacks on f controllers. We develop an optimization programming formulation that provides the switch-controller mapping solution and minimizes the total number of controllers required. We consider the controller processing capacity and communication delay between switches and controllers as problem constraints. Our approach also facilitates capacity sharing of backup controllers when two switches use the same backup controller but do not need it simultaneously. We demonstrate the effectiveness of the proposed approach through numerical analysis. The results show that the proposed approach significantly reduces the total number of controllers required by up to 50% compared to an existing scheme while guaranteeing better load balancing among controllers with a fairness index of up to 0.92.

Large amount of data is being generated at an alarming rate by various systems and devices such as computing systems, cameras and mobile devices. Owing to the huge volume of this data, its processing and analysis cannot be just limited to the place of origin but require to be done at multiple computing sites. A crucial problem is how to efficiently transfer and handle big data in a network, whose performance is affected by the transfer path and bandwidth allocated to the path. In this paper, we propose a bandwidth allocation scheme that flexibly and adaptively allocates bandwidth to big data transfer requests with an objective to maximize the acceptance ratio of the requests while satisfying the deadline constraints. We first develop an optimization programming formulation and then propose a heuristic algorithm to solve the problem due to its non-linear nature. We evaluate the performance of the proposed algorithm through comprehensive simulations on a realistic network topology for two routing scenarios: a pre-computed path scenario and a load-based routing scenario. In both scenarios, the proposed algorithm outperforms baseline algorithms by reducing the rejection ratio by at least 40% and increasing the data transferred by at least 21 TB in a day.

Distributed controller architectures in software defined networks raise the issue of switch-controller mapping. In a mapping approach where a switch distributes flow setup requests (traffic) to multiple controllers, a solution that finds the optimal switch-controller mapping and traffic distribution among the controllers for long term performance and responds effectively to network events such as short term traffic variation and controller failure is necessary. We develop a Multi-Controller Traffic Engineering (MCTE) scheme that: i) finds the long term switch-controller mapping and traffic distribution that minimizes flow setup time, ii) manages traffic distribution during short term variation, and iii) pre-computes backup controllers and traffic distribution upon controller failure. We formulate optimization problems for MCTE components and develop heuristic algorithms to obtain solutions in reasonable time. Numerical simulations show that the proposed algorithms achieve flow setup time within 2% of the lower bound and effectively manage traffic upon traffic variations and controller failures.

Increased power usage and network performance degradation due to best-effort bandwidth sharing significantly affect tenancy cost, cloud adoption, and data center efficiencies. In this article, we propose a novel Sliding-Scheduled Tenant request model which enables tenants to specify the required duration of their application within a certain window, in addition to its resource requirement graph. We investigate the sliding-scheduled application placement and routing problem, which selects the start-time of requests within their specified time-window to reserve both server and network resources for their required duration, and therefore provide resource guarantees with predictable performance. Using the multi-component utilization-based power model, we formulate the problem as an optimization problem that maximizes the acceptance rate while consuming as low power as possible. We develop fast online heuristics that adopt power, acceptance and adaptive spread-based scheduling policies while allocating the resources with the consideration of request duration and current shutdown-time of the devices. We demonstrate the effectiveness of the proposed algorithms in terms of power saving and acceptance rate, 1) for small data centers, by comparing their performance with the numerical results obtained from solving the optimization problem using CPLEX and 2) for large data centers using comprehensive simulation results.

With the evolution of software-defined networking (SDN) paradigm, traffic management in data center networks has become flexible and scalable. The existing solution using OpenFlow, the rate-guaranteeing mechanism, is inefficient as it limits the rate of the flows by dropping batches of packets to achieve the desired throughput. In this paper, we propose BASIS, a solution based on Bayesian inference for providing proportional Quality of Service (QoS) guarantees to tenants in a datacenter network. With BASIS, the bandwidth of an outgoing congested link will be shared among the competing flows in proportion to the weights chosen by them. We use Bayesian inference to capture the history of flow arrival rates and their offered load using a single queue, and estimate the differential drop probabilities of flows in a way that respects the weights assigned to them on arrival. Unlike the rate-limiting approach, BASIS proactively drops a packet of a flow probabilistically to achieve the desired throughput and avoids dropping batches of packets. We evaluate the proposed solution in an emulated SDN platform and show that BASIS achieves the desired throughput with lesser number of packet drops than the existing approaches.

Wavelength division multiplexed optical networks have become an attractive candidate to meet the ever-growing traffic demands in cloud data centers due to the features of large capacity and dynamic reconfiguration capability. While the bandwidth does not affect the makespan of compute-intensive and content-delivery-network applications, it has an impact on data-intensive applications that therefore require guaranteed bandwidth beside computing and storage resources for predictable performance. Motivated by this, we consider the problem of dynamically adjusting bandwidth so as to increase the acceptance of virtual networks embedded in optical data centers. We first develop an optimization programming formulation for the problem. We then develop a heuristic algorithm that efficiently embeds and adaptively allocates bandwidth to virtual networks such that the applications complete and release resources for future requests. We evaluate our algorithm through extensive simulations. The results show that our algorithm outperforms baseline algorithms by reducing rejections by up to 25%.

In this paper, we propose a Fragmentation-based Multipath Routing (FMR) model for Software Defined Networks (SDNs) to enable attack-resilient data transfer. With the use of erasure encoding to fragment a message, the fragments are routed along multiple paths such that no intermediate node receives enough fragments required for message decoding. This ensures that, any intruder on a compromised node does not infer the original data from the received fragments. We develop an optimization programming formulation of the problem to choose reliable paths that provide resilience to attacks. Using FMR, the SDN controller dynamically routes the data fragments along a set of most reliable paths to achieve multipath diversity and hence improve data availability at the destination even in the presence of an attack. We carry out performance studies and demonstrate the effectiveness of our approach in terms of weighted path reliability and blocking performance.

With more parallel and distributed applications moving to Cloud and data centers, it is challenging to provide predictable and controllable resources to multiple tenants, and thus guarantee application performance. In this paper, we propose an integrated QoS-aware resource provisioning platform based on virtualization technology for computing, storage and network resources. Coarse-grained CPU mapping and fine-grained CPU scheduling mechanisms are proposed to enable adjustable computing power. A hierarchical distributed scheduling mechanism is implemented on a scalable storage system to guarantee I/O throughput for individual tenants and applications. A network manager has also been developed to guarantee the data transmission rate. Web-based interface enables users to monitor real time resource utilization and to adjust resource QoS levels on the fly. According to our experimental results, the resource cost can be saved up to 45% without degrading the performance of a distributed data processing benchmark, and the performance of a parallel agent-based simulation can be improved by 91% using the same amount of resources.

In Software Defined Networks (SDNs), a reactive approach for failure recovery involves the centralized SDN controller which incurs long delay leading to packet losses. While a proactive approach enables fast failure recovery, it poses a new challenge concerning the number of additional forwarding rules required at every switch traversed by a flow on the primary and backup paths. These forwarding rules are stored in Ternary Content Addressable Memory (TCAM) which is limited in size and can hold only a few thousands of rules at a switch since it is expensive and power hungry. In this paper, we develop and analyze two proactive local rerouting algorithms namely Forward Local Rerouting (FLR) and Backward Local Rerouting (BLR) to compute backup paths for a primary path. By rerouting the failed traffic from the point of failure, local rerouting enables fast recovery. The proposed FLR and BLR algorithms choose backup paths so as to reduce the number of forwarding table entries with improved sharing of forwarding rules at the switches along the primary and backup paths. We evaluate the proposed algorithms through simulations on different topologies. The results show that the proposed algorithms reduce the average number of additional rules required to protect a flow by up to 75% compared to the existing approaches which do not take into account the limited size of TCAM. The results also show that the proposed algorithms are effective in terms of backup bandwidth sharing efficiency.

Increased power usage and network performance variation due to best-effort bandwidth sharing significantly affect tenancy cost, cloud adoption, and data center efficiencies. In this paper, we propose a novel time-aware request model which enables tenants to specify an estimated required time-duration, in addition to their required server resources for Virtual Machines (VMs) and network bandwidth for their communication. We investigate the VM-placement and routing problem, which allocates both server and network resources for the specified time-duration, to provide resource guarantees. Further, we exploit VM-migration while considering its power consumption overhead, to improve power saving and resource utilization. Using the multi-component utilization-based power model, we formulate the problem as an optimization problem that maximizes the acceptance rate while consuming as low power as possible. We develop fast online heuristics that allocate resources for requests, considering their duration and bandwidth demand. We also develop migration policies augmenting these heuristics. For migration heuristics, we propose server-migration and switch-migration approaches, which migrate the VMs between the powered-on servers only if their migrations result in turning-off at least one server and switch, respectively. We demonstrate the effectiveness of the proposed heuristics in terms of power saving, acceptance ratio, and migration overhead using comprehensive simulation results.

This article focuses on the problem of bandwidth allocation to users of Cloud data centers. An interesting approach is to use advance bandwidth reservation. Such systems usually assume all requests demand either bandwidth-guarantee (BG) or time-guarantee (TG), but not both. Hence the solutions are tailored for one type of requests. A BG request demands guarantee on bandwidth; whereas a TG request demands guarantee on time for transfer of data of specified volume. We define a new model that allows users to not only submit both kinds of requests, but also specify flexible demands. We tie up the problem of bandwidth allocation with differential pricing, that gives discounts to users based on the flexibility in their requests. We propose a two-phase, adaptive and flexible bandwidth allocator (A-FBA) that, in one phase admits and allocates minimal bandwidth to dynamically arriving user requests, and in another phase, allocates additional bandwidth for accepted requests maximizing revenue. The problem formulated in first phase is${\cal NP}$-hard, while the second phase can be solved in polynomial time. We show that, in comparison to a traditional deterministic model, the A-FBA not only increases the number of accepted requests significantly, but also does so by generating higher revenues.

Power efficiency and predictable performance have become major concerns for cloud service providers as they significantly affect cloud adoption and tenancy cost. Providing guaranteed resources for predictable performance in data centers drives the need for a request model which abstracts the traffic characteristics as well as the resource requirements of tenant applications. In this paper, we propose a novel Sliding Scheduled Tenant (SST) request model which enables tenants to request their resources for an estimated required time duration which can slide within a certain time-window. We investigate the power-efficient resource-guaranteed Virtual Machine (VM) -placement and routing problem for dynamically arriving SST requests. The problem requires provisioning of the specified resources in a data center for the required duration of requests by choosing an appropriate start- and end-time within their specified time-window, so as to maximize the number of accepted requests while consuming as low power as possible. We develop a mixed integer linear programming (MILP) optimization problem formulation based on the multi-component utilization-based power model. Since this problem which is a combination of VMplacement, scheduling and routing problems, is computationally rohibitive, we develop a fast and scalable heuristic algorithm. We demonstrate the effectiveness of the proposed algorithm and SST request model in terms of power saving and acceptance ratio through comprehensive simulation results.

With ever increasing traffic demands, data enter networks are envisioned to be a hybrid of both optical and electrical networks. In this context, we consider the recently proposed dynamic wavelength grouping (DWG) architecture for the optical network. This architecture can dynamically group wavelengths from different ports onto a single fiber carrying fixed number of wavelength groups. We focus on the joint problem of VM-placement and bandwidth allocation in such a hybrid optical-electrical data enter network with DWG capability. There are multiple challenges: (i) the number of edge-switches that can be simultaneously reached using optical paths from an edge-switch is limited by cost, and (ii) wavelength-group continuity constraint. Abstracting the requests of tenants as virtual networks, we study the novel problem of embedding virtual networks on this hybrid datacenter, which translates to the joint problem of bandwidth allocation and placement such that the requirements of virtual networks are satisfied. We develop and analyse two algorithms for embedding dynamically arriving virtual network demands on a hybrid datacenter with DWG capability. The performance studies demonstrate the effectiveness of exploiting existing optical paths as well as using electrical links in the face of multiple constraints to accept higher number of requests.

With the ubiquitous adoption of Cloud services by both companies and consumers alike, lack of an efficient system to explicitly price and allocate limited bandwidth has severely impacted the performance of Cloud user-applications. In this context, we consider a two-tier pricing model — consisting of Reservation Phase and Dynamic Phase — that caters to the needs of different kinds of applications. While the Reservation Phase can be used by Cloud users to obtain guarantees on minimum bandwidth well ahead in time, Dynamic Phase can be used to demand and obtain (possibly) additional bandwidth dynamically. Bandwidth being a limited resource, we develop a unique multi-stage uniform price auction with supply uncertainty to dynamically allocate bandwidth to users in the Dynamic Phase. We study the proposed model using a game theoretical approach. Our results prove that proposed auction mechanism is a promising approach for bandwidth allocation. We show that the model promotes the dual advantage of market efficiency and maximum revenue for the Cloud provider. We also demonstrate the price stability using numerical simulations. We argue that for rational, payoff-maximizing tenants of Cloud, the price is stable over the long run which makes the mechanism suitable for practical use.

A promising development in the design of datacenters is the hybrid network architecture consisting of both optical and electrical elements. In this context, the joint problem of bandwidth allocation and VM-placement, a problem that only recently received attention in all-electrical datacenter networks, poses new and different challenges not addressed yet in hybrid datacenters. In particular, we foresee two issues: (i) the number of edge-switches that can be simultaneously reached using optical paths from an edge-switch is limited by the switch size, (ii) the dynamic creation of optical paths can potentially establish a constrained optical network topology leading to poor performance. We abstract the requests of tenants as virtual networks, and study the problem of embedding virtual networks on a hybrid datacenter, which translates to the joint problem of bandwidth allocation and placement such that the topology constraints of virtual networks are satisfied. We develop and analyse two algorithms for embedding dynamically arriving virtual network demands on a hybrid optical-electrical datacenter. Through simulations, we demonstrate the effectiveness of not only exploiting the already established optical paths, but also of using electrical network in embedding requests of virtual networks.

As bandwidth is shared in a best-effort way in today's data centers, traffic generated between a set of VMs (virtual machines) affect the traffic between another set of VMs (possibly belonging to another tenant) sharing the same physical links, leading to unpredictable performance of applications running on these VMs. This article addresses the problem of allocation of not only server resources (computational and storage) but also network bandwidth, to provide performance guarantees in multi-tenant data centers. Bandwidth being a critical shared-resource, we formulate the problem as an optimization problem that minimizes bandwidth demand between clusters of VMs of a tenant; and we prove it as NP-hard. We develop fast online heuristics as an integrated resource allocator (IRA) that decides on the admission of dynamically arriving requests, and allocates resources for the accepted ones. We also present a modified version of IRA, called B-IRA that bounds the cost of bandwidth allocation, while exploring smaller search space for solution. We demonstrate that, IRA accommodates significantly higher number of requests in comparison to a load-balancing resource allocator (LBRA) that does not consider reducing bandwidth between clusters of VMs. IRA also outperforms B-IRA when traffic demands of VMs in an input are not localized.

Variation in network performance due to the shared resources is a key obstacle for cloud adoption. Thus, the success of cloud providers to attract more tenants depends on their ability to provide bandwidth guarantees. Power efficiency in data centers has become critically important for supporting larger number of tenants. In this paper, we address the problem of time-aware VM-placement and routing (TVPR), where each tenant requests for a specified amount of server resources (VMs) and network resource (bandwidth) for a given duration. The TVPR problem allocates the required resources for as many tenants as possible by finding the right set of servers to map their VMs and routing their traffic so as to minimize the total power consumption. We propose a multi-component utilization-based power model to determine the total power consumption of a data center according to the resource utilization of the components (servers and switches). We then develop a mixed integer linear programming (MILP) optimization problem formulation based on the proposed power model and prove it to be N P-complete. Since the TVPR problem is computationally prohibitive, we develop a fast and scalable heuristic algorithm. To demonstrate the efficiency of our proposed algorithm, we compare its performance with the numerical results obtained by solving the MILP problem using CPLEX, for a small data center. We then demonstrate the effectiveness of the proposed algorithm in terms of power consumption and acceptance ratio for large data centers through simulation results.

We investigate how traffic splitting can be exploited for power-efficient connection provisioning in optical networks. We propose an algorithm for determining the most power-efficient Label-Switched Path (LSP) for a connection. A comparative study shows that traffic-splitting-enabled networks consume less power than non-traffic-splitting networks.

Bandwidth-sharing in data-center networks is an important problem that affects the growth of multi-tenant datacenters. A promising solution approach is the use of advance reservations. Such systems are usually based on deterministic models, assuming users to have precise knowledge of bandwidth they require, which is unlikely. This work takes a deviation and proposes a probabilistic model, where bandwidth requirements are specified along with some probabilities. As user-estimate of bandwidth depends on the cost they incur, we tie up the model with differential pricing; and formulate bandwidth allocation as a two-phase - static and dynamic - optimization problem. We show that the problem in each phase is NP-hard. We develop a bandwidth-allocator that processes requests and defines bandwidth profiles for accepted requests by solving the optimization problems. Numerical studies show that, in comparison to the deterministic model, our model brings down the number of rejected requests significantly, while allocating more bandwidth and thereby increasing revenue for providers.

The success of multi-tenant data-centers depends on the ability to provide performance guarantees in terms of the resources provisioned to the tenants. As bandwidth is shared in a best-effort way in today's data-centers, traffic generated between a set of VMs affect the traffic between another set of VMs sharing the same physical links. This paper proposes an integrated resource allocation scheme that considers not only server-resources, but also network-resource, to decide the mapping of VMs onto servers. We present a three-phase mechanism that finds the right set of servers for the requested VMs, with an aim of reducing the bandwidth on shared links. This mechanism provisions the required bandwidth for the tenants, besides increasing the number of tenants that can cohabit in a data-center. We demonstrate, using simulations, that the proposed scheme accommodates 10%–23% more requests in comparison to a load-balancing allocator that does not consider bandwidth requirements of VMs.

Fast Spanning Tree Reconnection (FSTR) is a newly proposed restoration method by us to handle single link failures in Metro Ethernet networks. Compared with other methods existing in the literature, FSTR has the advantages of lower signalling overhead, faster recovery and lesser rerouted traffic. In this work, we propose to use partial spatial protection (PSP) mechanism to provision differentiated reliability (DiR) and, as a result, to decrease the blocking probability of traffic requests in FSTR- based Metro Ethernet networks. In PSP, a traffic request is protected using backup bandwidth against the failure of only a subset of the links on its route. We develop a mixed integer linear programming (MILP) model for PSP to minimize the bandwidth reserved by an incoming traffic request while satisfying the protection grade required by the traffic request. The numerical results demonstrate that the proposed solution can reduce the blocking probability of traffic requests by up to 83% in a 4×4 grid network, compared to that in full protection.

Ethernet is becoming a preferred technology to be extended to metropolitan area networks (MANs) due to its low cost, simplicity, and ubiquity. However, current Ethernet lacks a fast failure recovery mechanism as it reconstructs the spanning tree after the failure is detected, which commonly requires tens of seconds. Some fast failure-handling methods based on multiple spanning trees have been proposed in the literature, but these approaches are either centralized or require periodic message broadcasting over the entire network. In this paper, we propose a local restoration mechanism for metro Ethernet using multiple spanning trees, which is distributed and fast and does not need failure notification. Upon failure of a single link, the upstream switch locally restores traffic to preconfigured backup spanning trees. We propose two restoration approaches, connection-based and destination-based, to select backup trees. We formulate the tree preconfiguration problem that includes working spanning tree assignment and backup spanning tree configuration. We prove that the preconfiguration problem is NP-complete and develop an integer linear programming model. We also develop heuristic algorithms for each restoration approach to reduce the computation complexity. To evaluate the effectiveness of our heuristic algorithms, we carry out the simulation on grid and random networks. The simulation results show that our heuristic algorithms have comparable performance close to the optimal solutions, and both restoration approaches can efficiently utilize the network bandwidth to handle single link failures.

In Optical Burst Switching (OBS) network architecture, edge nodes assemble bursts and send them to the core network arbitrarily, which will lead to inevitable collisions and low network utilization in the core network. Our objective is to achieve high performance burst transmission for bursty traffic while avoiding the use of wavelength converters and excessive optical buffers. To achieve this objective, we propose a novel optical burst chain switching (OBCS) mechanism, which combines the merits of optical circuit switching and optical burst switching with affordable signaling overhead. In the proposed mechanism, switching unit is a burst chain which consists of multiple non-periodic and non-consecutive bursts in one wavelength. Theoretical analysis for the throughput and queuing delay of the proposed scheme is carried out and is also verified by simulation results. We present extensive simulation results to demonstrate its superior performance over OBS networks with/without wavelength converters.

Ethernet is becoming a preferred technology to be deployed in metro domain due to its low cost, simplicity and ubiquity. However, traditional spanning tree based Ethernet protocol does not meet the requirement for metro area networks in terms of network resilience. In the work of Qiu et al. (2009), we proposed a fast spanning tree reconnection (FSTR) mechanism for metro Ethernet networks to handle single link failures. Upon failure of a link on a spanning tree, FSTR mechanism activates a reconnect-link to reconnect the broken spanning tree. FSTR mechanism has the features of fast recovery, simplicity, and guaranteed protection. However, when more than one link fail in the network, the FSTR mechanism would generate unexpected loops and cannot function properly. In this paper, we propose a fast spanning tree reconnection mechanism to handle double-link failures with protection grade guarantees. The mechanism is distributed and can alleviate the problem in previous FSTR mechanism. We formulate the reconnect-link pre-configuration problem for double-link failures as an integer linear programming problem. Through numerical results we demonstrate that the proposed mechanism can satisfy the protection grade required for each connection by efficiently utilizing the network capacity.

Ethernet is becoming a preferred technology to be deployed in metro domain due to its low cost, simplicity and ubiquity. However, spanning tree based Ethernet protocol does not meet the requirement for Metro Area Networks in terms of network resilience, despite the advancement of Ethernet standardization and commercialization. In this paper, we propose a fast spanning tree reconnection (FSTR) mechanism for Metro Ethernet networks to handle any single link failure, which has features of fast recovery, backup capacity guarantees and ease of implementation. Upon failure of a link on a spanning tree, a distributed failure recovery protocol is activated to reconnect the broken spanning tree using a reconnect-link not on the spanning tree. We present the details of the protocol, including failure notification and forwarding table reconfiguration procedures. The pre-configuration of the reconnect-links to reconnect each spanning tree is formulated as an integer linear programming (ILP) problem. The optimization results show that with lower implementation cost, fast spanning tree reconnection mechanism can achieve comparative or considerably better performance than other resilient mechanisms for Metro Ethernet networks.

Protected working lightpath envelope (PWLE) is a promising path-oriented protection strategy developed by us to provision survivable services for dynamic traffic. Partitioning the network capacity into a working layer and a protection layer by means of lightpath-protecting p-cycles, PWLE possesses the advantages of high capacity efficiency, avoidance of wavelength conversion, good blocking performance and guaranteed optical transmission quality. Nevertheless, the impact of network connectivity on the actual utilization of the working layer has been overlooked. Hence, we propose connectivity aware protected working lightpath envelope (CAPWLE) to augment the actual utilization of the working layer whilst retaining its high capacity efficiency. To fulfill the goal, we propose effective envelope, developed based on the maximum concurrent flow problem (MCFP), to evaluate the protection capability of lightpath-protecting p-cycles from a combined perspective of connectivity and volume. By virtue of effective envelope, CAPWLE is optimized by a mixed integer linear programming (MILP) model. Numerical results show that the actual utilization of the working layer is enhanced by CAPWLE under both dynamic stationary and non-stationary traffic patterns.

In this paper, we investigate the problem of multipath traffic engineering in optical burst switching (OBS) networks. The main goal of this work is to minimize burst loss rate in the network by adaptively balancing the burst traffic among multiple paths based on the measurement and analysis of path congestion. We develop three distributed traffic splitting algorithms by considering the special features of OBS networks. Among these three algorithms, one algorithm considers burst level traffic distribution, and the other two algorithms consider flow level traffic distribution to avoid packet reordering problem. The proposed multipath traffic engineering algorithms achieve significant performance improvement in reducing burst loss probability in a distributed manner in OBS networks. Further, through theoretical analysis and simulations, we show that the proposed algorithms converge quickly on a sudden traffic increase in the network.

We consider Label Switched Path (LSP) protection for connections with various protection grade requirements in multi-protocol label switching (MPLS) over wavelength division multiplexing (WDM) optical networks. In full protection, bandwidth needs to be reserved for the backup LSP to protect the failure of any fiber along the primary LSP. In this paper, we study the problem of partial spatial-protection (PSP) where bandwidth is reserved for the backup LSP to protect the failure of a subset of fibers traversed by the primary LSP to satisfy the specified protection grade. We formulate the optimal LSP PSP problem as an ILP and identify three suboptimal problems. For each suboptimal problem, an exhaustive search algorithm and a heuristic are developed. We analyze the probability that a connection can be restored upon a fiber failure and find that it is higher than or equal to the protection grade specified. We also develop a failure recovery protocol which specifies recovery operations upon failure and determines whether a connection can be recovered if the failed fiber is one of its unprotected fibers. Through extensive simulation experiments, we demonstrated that the proposed LSP PSP algorithms perform better than shared LSP protection in blocking probability and resource efficiency.

This paper deals with the problem of routing and wavelength assignment (RWA) of fault-tolerant sliding scheduled lightpath demands (FSSLDs) in WDM optical mesh networks. Sliding scheduled traffic model allows the service provider and end-users to negotiate the starting time and ending time of the demands. We have developed a time conflict resolving algorithm that exploits the time disjointness that could exist among FSSLDs by rearranging the demands and then dividing them into time-independent windows. We then present, two RWA algorithms to efficiently route scheduled lightpath demands from time-independent windows. The proposed algorithms schedule both primary and end-to-end protection routes and also assign wavelengths for the duration of the demands. Extensive simulations are conducted on ARPANET, NSFNET, USANET, and Mesh 8×8, 10×10, 12×12 networks. By rearranging the demands and exploiting time-disjointness across demands, the proposed algorithms can reuse the wavelengths and hence reduces the amount of global resources required and blocking probability.

Protected working lightpath envelope (PWLE) is a promising scheme developed by us to provision survivable services for dynamic traffic. As a a path-oriented protection strategy based on lightpath-protecting p-cycles, PWLE has the advantages of high capacity efficiency, avoidance of wavelength conversion, good blocking performance and guaranteed optical transmission quality compared with conventional schemes. It partitions the total network capacity into a working layer and a static protection layer in which a set of lightpath-protecting p-cycles are configured. The lightpath-protecting p-cycles in PWLE are not designed particularly for a set of pre-defined paths. Instead, they are selected to protect an envelope of working channels that can be used flexibly for working routing. The protected working channels are grouped based on the related attach nodes which are defined as the nodes one-hop away from the on-cycle nodes. To design PWLE, pre-computation of a subset of candidate cycles and cycle selection within the subset are crucial. Due to the uniqueness of the lightpath- protecting p-cycle, no existing algorithms can be applied directly for cycle selection for PWLE. Moreover, existing algorithms for pre-computation of candidate cycles are mostly span-protection- oriented. Therefore, we propose the attachnode-based cycle generation (ANCG) algorithm for pre-computation of candidate cycles and three algorithms for lightpath-protecting p-cycle selection for PWLE. Numerical results show that ANCG can generate a small subset of cycles with high capacity efficiency. Meanwhile, the algorithms for cycle selection work well with much reduced computational time.

The next-generation dynamically-reconfigurable intelligent GMPLS-based WDM optical networks are expected to operate at extremely high data rates (40 Gbps or beyond) and will be used to carry voice, data, and various other multimedia applications. In these networks a signalling protocol (such as RSVP-TE), which is used to dynamically set up lightpaths, should not only be fast and efficient, but must also be scalable and should try to maximize the number of accepted connections at an acceptable level of control overhead. In this paper, we propose a fast and efficient segmented signalling protocol (SSP) based on a novel concept of intermediate destinations (IDs), for wavelength reservation in GMPLS/WDM optical networks. In SSP, as the route is divided into a number of smaller independent segments on which RSVP-TE runs simultaneously, it reduces the setup time. Simulation experiments on 10times10 grid network show that the proposed protocol is fast with respect to setup time, efficient in terms of average call acceptance ratio and resource utilization.

Optical burst switching (OBS) is a promising technique for wavelength division multiplexing (WDM) networks. In practice, wavelength converters (WCs) are either absent or only sparsely deployed in WDM networks due to economic and technical limitations. Thus, wavelength assignment is expected to be an important component of OBS networks. In this paper, an offline wavelength assignment scheme in OBS networks without wavelength conversion capability is proposed. The key idea of the scheme is to decide the wavelength searching order of each traffic connection at edge nodes according to the wavelength priorities determined by the calculated burst loss probabilities on different wavelengths. Simulation results indicate that the proposed scheme can reduce the network-wide burst loss probability significantly compared with other schemes. It is also illustrated that the performance of the proposed scheme can be further enhanced by a larger number of wavelengths per link and a reasonable delay bound at edge nodes.

This paper addresses the problem of dynamic routing of survivable sub-lambda traffic in heterogeneous grooming optical networks wherein sub-lambda client layer connections are multiplexed onto lambda connections on the optical layer. We present a differentiated survivability framework, which includes multi-layer survivability approaches with improved resource sharing methods based on inter-layer and inter-class resource sharing techniques. In addition to this, a new graph based network model is proposed, which supports both 1) the heterogeneity in a network, and 2) the coexistence of inter-layer and inter-class based resource sharing methods. The suitability of the model for a critical must-use grooming port scenario is presented. A tradeoff phenomenon between transceiver-usage and reserved links, which is inherent in the inter-layer sharing methods, is illustrated. Simulation experiments are carried out, and the performance variation and the tradeoff phenomenon are investigated.

In this paper, we consider the problem of scheduling sensor activity to maximize network lifetime while maintaining both discrete targets coverage and network connectivity. We assume that each sensor can distinguish and select up to L targets in its sensing range to observe and sends all the observed-data back to the sink via single or multiple hop communications. We develop a polynomial-time algorithm which can achieve optimal solution based on the theory of linear programming and integer theorem. We then develop an flexible efficient heuristic algorithm and demonstrate the effectiveness of its performance through extensive simulation results.

We propose a new scheme, called protected working lightpath envelope (PWLE), for dynamic provisioning of survivable services without requiring any wavelength conversion. PWLE is a path-oriented protection strategy based on lightpath-protecting rho-cycles. It partitions the total network capacity into a static protection layer and a working layer available for dynamic routing. To deal with the new issues raised by lightpath-protecting p-Cycles, we propose compatible grouping so that the PWLE can be formulated as an mixed integer linear programming (MILP) model. Based on compatible grouping, we further propose compatible group routing (CGR), a distributed routing algorithm tailored for PWLE. Compared with rho-Cycle-based protected working capacity envelope (PWCE), PWLE has the advantages of high capacity efficiency, avoidance of wavelength conversion, good blocking performance and guaranteed optical transmission quality. We evaluate the performance improvement of PWLE in capacity efficiency and blocking performance through the numerical results obtained from CPLEX and simulations.

In order to meet the requirements of real-time applications, Optical Burst Switched backbone networks need to provide quantitative edge-to-edge loss guarantees to traffic flows. For this purpose, there have been several proposals based on the relative differentiation quality of service (QoS) model. However, this model has an inherent difficulty in communicating information about internal network states to the edge in a timely manner for making admission control decisions. In this paper, we propose an absolute QoS framework to overcome this difficulty. The key idea is to offer quantitative loss guarantees at each hop using a differentiation mechanism and an admission control mechanism. The edge-to-edge loss requirement is then translated into a series of small per-node loss probabilities that are allocated to the intermediate core nodes. The framework includes a preemptive differentiation scheme, a node-based admission control scheme and an edge-to-edge reservation scheme. The schemes are analyzed and evaluated through simulation. It is shown that the framework can effectively offer quantitative edge-to-edge loss guarantees under various traffic conditions.

In this paper, we consider the reliability of a connection as a parameter to denote different levels of fault- tolerance and propose a segment-based partial protection scheme for routing reliability-guaranteed connections in a resource efficient manner. In this scheme we try to provide a partial protection path instead of an end-to-end protection path to improve the reliability of a connection. To provide partial protection path, we need to identify primary segments which are less (more) reliable (vulnerable) to failures and select a suitable primary segment to provide a protection segment. However, identifying less-reliable primary segments which really contribute to achieve the required reliability and selection of resource-efficient protection segment among several possible segments are not trivial problems. In this paper, we develop efficient methods to address these problems. Apart from providing the reliability guarantee, the proposed scheme is able to recover all failures immediately, except the failures which are not covered by the protection segment. In this case the failed connections cannot be rerouted on to the protection segment and we initiate our proposed recovery process which handles all possible failure scenarios. We conduct extensive simulation experiments to evaluate the effectiveness of the proposed scheme on different network configurations with respect to three performance metrics: average spare wavelength usage, average recovery time, and average recovery ratio.

We consider the problem of offline route optimization in optical burst switching (OBS) networks to determine a route layout for a given traffic demand to minimize the overall burst loss. Route selection based on the traditional Erlang B formula is not efficient because of the unique features of OBS networks such as streamline effect. We analyze the streamline effect and propose a more accurate loss estimation formula which takes the streamline effect into consideration. Based on this formula, we develop a mixed integer linear programming (MILP) formulation for the problem. Since the MILP-based solution is computationally intensive, we develop a heuristic algorithm. We evaluate the effectiveness of the proposed algorithms through the numerical results obtained from CPLEX and simulation results.

In this paper, we investigate the problem of multipath dynamic traffic routing in optical burst switching (OBS) networks. In this work, traffic is split among the multiple link-disjoint paths between each source and destination pair. Taking into consideration the packet reordering problem, we propose a new adaptive proportional routing scheme that operates on flow level. The objective of the proposed scheme is to reduce the overall burst loss in the network and at the same time avoid the packet out-of-sequence arrival problem. This is achieved by adaptively adjusting the traffic flow proportions assigned to the multiple paths between a source and destination pair based on the measurement of the impact of the offered traffic load on each of them in the past. When a new flow arrives, an appropriate path is chosen for it based on the calculated flow proportions. In addition, burst assembly time between different paths can also be adjusted to further enhance the performance. Through extensive simulations, we show that the proposed flow-based multi-path traffic routing approach not only provides significantly better burst loss performance than the basic equal-proportion and hop-length based traffic routing algorithms, but is also void of any packet re-orderings.

Optical burst switching (OBS) is a promising technology to transmit bursty data traffic over optical networks. For QoS routing, an accurate estimation of the amount of residual capacity on a link is necessary. In IP or ATM networks, the residual bandwidth of a link can be computed as the total bandwidth subtracted by the effective bandwidth of each flow. However, this method is based on the assumption that each node has a large buffer. This method is not applicable to OBS networks, where there is no buffer at the core nodes. We propose a new metric, called residual admission capacity (RAC), to determine the amount of residual capacity on the links in OBS networks. To show the effectiveness of the new metric, we propose a routing algorithm called RAC-WSP, which is a widest shortest path (WSP) algorithm with RAC as the measurement of the residual capacity on a link. Experimental results show that RAC-WSP can admit more loss-guaranteed flows than other routing schemes, including the shortest path algorithm and other WSP algorithms where the residual capacity is evaluated by other metrics

This paper addresses the problem of dynamic routing of sub-lambda connections with differentiated survivability using multi-layer protection and inter-layer backup resource sharing in IP/MPLS-over-WDM networks. When providing the differentiated survivability services, mission-critical high priority connections which require optical layer protection are more likely to be rejected than connections which require IP/MPLS layer protection because of the difference in protection granularity. To address this problem, we propose two re routing-based dynamic routing schemes for reducing the blocking of mission-critical connections. In the schemes, rerouting operation is done with the use of lightpaths called potential lightpaths, and an efficient heuristic algorithm is proposed for choosing them. Further, the schemes employ inter-layer backup sharing for the benefit of high priority connections and they adopt strategies which consider critical issues in finding and utilizing the potential lightpaths. Our schemes have several important and attractive features: 1) the schemes do not cause any interruption for ongoing traffic during normal operation, 2) the schemes improve the performance of mission-critical traffic without affecting the performance of other traffic significantly, and 3) the schemes are affordable in terms of computational intensity since the rerouting operation is done only when a mission-critical connection is not honored. Through simulation experiments we investigate the performance of the proposals and verify their effectiveness

Network lifetime is one of the critical issues in sensor networks. An effective approach to prolong the network lifetime is to schedule the active states of sensors: only a subset of the deployed sensors that can maintain both sensing coverage and network connectivity is scheduled to be active. In this paper, we consider the connected target coverage (CTC) problem with the objective of maximizing the network lifetime by scheduling sensors into multiple sets, each of which can maintain both target coverage and connectivity among all the active sensors and the sink. We model the CTC problem as a maximum cover tree (MCT) problem and prove that the MCT problem is NP-complete. We give an upper bound on lifetime of the MCT problem and develop a heuristic algorithm called communication weighted greedy cover (CWGC) algorithm to solve it. We study the performance of CWGC algorithm comparing it with other algorithms that consider the coverage and connectivity problems independently. Simulation results show that CWGC algorithm performs much better than others in terms of the network lifetime and the lifetime obtained by our algorithm is close to the upper bound

This paper addresses the problem of achieving a balance in satisfying a network service provider's requirements: maintaining an acceptable call acceptance rate, satisfying protection requirements of requests, and controlling the signaling overhead in case of a component failure, in IP-over-WDM networks in a dynamic traffic arrival scenario. Satisfying all these aspects is a difficult task especially when the traffic pattern is dynamic in nature and a single layer protection approach is followed. In this work, first we propose a multi-layer protection scheme for achieving a better and acceptable tradeoff between blocking performance and signaling overhead in IP-over-WDM networks. We define various operational settings in the proposed scheme and investigate their impacts on the performance of the scheme. An important feature of this scheme is that, these settings allow a network service provider to select a suitable operational strategy for achieving the desired tradeoff based on network's policy and traffic demand. Then we propose an adaptive protection approach for dynamic traffic in consideration of providing better protection to requests as much as possible while considering blocking performance and signaling overhead. Through simulation experiments we evaluate the performance of the proposals and demonstrate their effectiveness.

We propose a Probabilistic Preemptive Burst Segmentation (PPBS) scheduling scheme to provide priority classes with different Quality-of-service (QoS) requirements in WDM Optical Burst Switching (OBS) networks. The PPBS scheme enables high priority bursts to preempt and segment low priority bursts in a probabilistic fashion. The PPBS scheme achieves 100% isolation among priority classes, and burst loss probabilities can be controlled by tunable parameters. Our PPBS queueing model also implicitly enables us to obtain several results related to the Erlang's loss function which may be useful in the role of this function arising in some areas of queueing theory, as well as characterizing a loss rate region that work conserving burst scheduling techniques can achieve. As an application to providing service differentiation, we show how the PPBS scheme can minimize the total sum of loss rates and achieve proportional loss differentiation. Finally, we also demonstrate the effectiveness of the PPBS scheme empirically using realistic Internet traffic model, e.g., long range dependent traffic model.

Sensor networks comprise large number of sensor nodes densely located in an area for sensing purposes. A key consideration in sensor networks is that the sensor nodes have limited battery resources that determine the lifetime of the network. Motivated by the fact that geographical routing is localized, scalable and nearly-stateless, we propose three energy-aware geographical forwarding schemes. The aim is to improve the network lifetime by considering the residual energy of neighbours in deciding the next-hop while preserving the localized, scalable and nearly-stateless property of geographical routing. In view of a possible heavy traffic along the perimeter of the hole, which may cause nodes along this perimeter to die faster, a hole avoidance method is proposed to reduce the traffic along the hole to remove this possibility. Topology control algorithm is proposed to be executed at each node to remove undesirable geographical neighbours determined by a link cost. It removes geographical neighbours without causing unnecessary holes and still maintains the connectivity in the network. The main aim of removing undesirable geographical neighbours is to achieve the global objective of having a longer network lifetime. The effectiveness of the three proposed forwarding schemes and the topology control algorithm are verified through extensive simulation results

In this paper, we describe and study a phenomenon unique to bufferless optical burst switched (OBS) networks called the streamline effect. This is the phenomenon wherein bursts within an input stream to a core node only contend with those from other input streams but not among themselves. It causes the burst loss probability at a link to depend strongly on the number of input streams to the link and their relative burst rates. We analyse the streamline effect and provide a burst loss probability formula that is more accurate than the traditionally used Erlang B formula. The formula is then used in a load balancing scheme for reservation-based quality of service (QoS) traffic to give a better link cost function. Through extensive simulation experiments for different traffic scenarios, we show that the proposed load balancing scheme utilising the streamline effect performs better than the shortest path routing scheme and the load balancing scheme without considering the streamline effect.

This paper addresses the problem of quality of service (QoS) differentiation of sub-wavelength level label switched path (LSP) requests considering both differentiated survivability services and transmission quality in terms of end-to-end delay requirements of requests. We define different classes of LSP requests and propose a novel QoS differentiation scheme for IP-over-WDM networks. The scheme adopts a multi-layer protection approach consists of dedicated lightpath protection, shared lightpath protection, and shared LSP protection methods. The inter-class backup resource sharing technique is used to improve the performance of different classes of traffic. The routing strategies consider efficient resource utilization and end-to-end delay requirements of different priority requests. In this work, LSPs belonging to different classes with different protection requirements are allowed to traverse the same lightpath to use resources effectively. This grooming approach is efficient in terms of resource usage and it poses several challenges: 1) lightpath protection can be used or modified to satisfy the protection requirement of a newly traversed LSP, and 2) when and how the protection resources are released or updated when an LSP is released while preserving the other LSPs' protection needs. An effective way of distributing signaling messages for failure recovery should also be considered as different protection methods are used. We address all these issues in the proposed scheme. Through simulation experiments, we investigate the impacts of the backup sharing strategies and the routing strategies on connections acceptance rate for different classes of traffic and end-to-end delay requirements of different priority requests. Based on simulation results we demonstrate how the selection of a suitable sharing strategy and a routing strategy can be done that gives the best performance for all the classes and show the effectiveness of the scheme in satisfying both users' QoS requirements and network services providers' requirements.

In WDM optical networks, depending on the offered services the service provider will have precise information for some traffic demands such as the number of required lightpaths and the instants at which these lightpaths must be set-up and torn-down, known as scheduled lightpath demands (SLDs). It may so happen that in a given set of SLDs, some of the demands are not simultaneous in time, and hence the same network resource could be used to satisfy several demands at different times. In this paper we develop two complementary algorithms-independent sets algorithm (ISA) and time window algorithm (TWA), based on circular arc graph theory, which respectively capture time-disjointness or time-overlap that could exist among SLDs. We compare and evaluate the algorithms based on the number of wavelengths required, number of reused wavelengths, average call acceptance ratio, and the reuse factor. The numerical results obtained from simulation experiments indicate that TWA reuses significant number of wavelengths (reuse factor up to 53%) followed by ISA (reuse factor up to 14%).

We study the dynamic survivable lightpath provisioning problem in wavelength-division multiplexing (WDM) optical mesh networks. Dynamic provisioning implies routing of connection requests that arrive one at a time with no a priori knowledge of future arrivals. We propose a new protection approach to provision lightpath requests according to their differentiated joint quality-of-service (QoS) requirements: connection reliability and recovery time, in a resource efficient manner. In our scheme, a probabilistic failure environment is assumed. We provide differentiated protection to the requests according to their differentiated reliability requirements, and at the same time guarantee the specified connection recovery time if failures occur and the connection is recoverable from these failures. We compare the proposed scheme with several sample full or partial protection schemes. Simulation results show that both QoS parameters have serious impact on the network performance and the proposed joint-QoS protection scheme significantly outperforms the other sample full or partial protection schemes in terms of connection blocking probability.

This paper presents an efficient protection scheme to provide differentiated survivability services for dynamic traffic in mesh WDM optical networks. In the deterministic quality of protection (QoP) framework, connections with different reliability requirements are classified by assigning a continuous spectrum of QoP grades. The backup bandwidth of those connections with low QoP grades is treated as low-speed traffic and may be multiplexed into one backup lightpath for the purpose of efficient bandwidth utilization. In this way, the bandwidth can be efficiently allocated by two-level backup bandwidth aggregation: backup bandwidth sharing and backup bandwidth multiplexing. To allocate the connections with different QoP grades, we propose two efficient routing algorithms: minimum delay QoP routing (MDQR) and maximum aggregation QoP routing (MAQR). Through simulation experiments, we demonstrate that our algorithms work well in terms of blocking performance and the ability of providing differentiated survivability services.

Protection schemes for WDM mesh networks are typically based on the single-failure scenario. However, the single-failure model is not realistic since the failure of network components is probabilistic. On the other hand, different applications or end users need different levels of fault tolerance. In this work, we assume a probabilistic failure environment and choose the connection reliability as a parameter to denote the level of fault tolerance. We propose a partial segment-based protection approach to accommodate lightpath requests according to their differentiated reliability requirements. We also consider incorporating backup sharing in such an environment. We conduct extensive simulation experiments to evaluate the effectiveness of the proposed scheme. Simulation results show that establishing connections according to their differentiated reliability requirements can effectively improve network blocking performance and the proposed partial segment-based protection scheme outperforms the partial path-based protection scheme proposed in the literature in terms of connection blocking probability.

Context information produced by various interconnected sensors and software sources needs to be efficiently identified and located by a context-aware application. In this paper, we propose an inter-space context discovery platform, called Orion, which allows context to be discovered and retrieved from multiple smart spaces. Orion employs semantic overlay network peer-to-peer architecture to provide connectivity between the smart spaces, allowing lookup query to be forwarded to the destined spaces. On the other hand, to overcome data heterogeneity, semantic Web ontology-driven approach is used for the context modeling and matchmaking.

In this paper, we consider the problem of routing and wavelength assignment (RWA) of fault-tolerant scheduled lightpath demands (FSLDs) in all optical wavelength division multiplexing (WDM) networks under single component failure. In scheduled traffic demands, besides the source, destination, and the number of lightpath demands between a node-pair, their set-up and tear-down times are known, in this paper, we develop integer linear programming (ILP) formulations for dedicated and shared scheduled end-to-end protection schemes under single link/node failure for scheduled traffic demand with two different objective functions: 1) minimize the total capacity required for a given traffic demand while providing 100% protection for all connections; and 2) given a certain capacity, maximize the number of demands accepted while providing 100% protection for accepted connections. The ILP solutions schedule both the primary and end-to-end protection routes and assign wavelengths for the duration of the traffic demands. As the time disjointness that could exist among fault-tolerant scheduled lightpath demands is captured in our formulations, it reduces the amount of global resources required. The numerical results obtained from CPLEX indicate that dedicated scheduled (with set-up and tear-down times) protection provides significant savings (up to 33 %) in capacity utilization over dedicated conventional (without set-up and tear-down times) end-to-end protection scheme; shared scheduled protection provides considerable savings (up to 21 %) in capacity utilization over shared conventional end-to-end protection schemes. Also the numerical results indicate that shared scheduled protection achieves the best performance followed by dedicated scheduled protection scheme, and shared conventional end-to-end protection in terms of the number of requests accepted, for a given network capacity.

Recently, some absolute QoS schemes have been proposed to offer quantitative loss guarantees at core nodes in OBS networks. By moving the admission control process to core nodes, which have the most updated information about their own traffic conditions, these schemes can provide reliable and robust per-link loss guarantees. However, the issue of how to provide edge-to-edge guarantees for burst flows over the entire edge-to-edge paths have not been investigated in any of the proposals. In this paper, we propose a signalling and reservation scheme to coordinate the reservation and teardown process over the edge-to-edge path in the absolute QoS model. The key idea is to judiciously divide the edge-to-edge loss requirement into a series of small loss probabilities that are allocated to the intermediate links. The performance of the scheme is evaluated through simulation.

We consider the problem of routing and wavelength assignment (RWA) of segmented protection lightpaths in all-optical wavelength division multiplexing (WDM) networks under single link and node failures for static traffic demand. We develop integer linear programming (ILP) formulations for dedicated and shared segmented protection schemes under single link/node failure for static traffic demand with two different objective functions: (1) minimize the total capacity required for a given traffic demand while providing 100% protection for all the traffic demands; (2) given a certain capacity, maximize the number of demands accepted while providing 100% protection for accepted connections. The numerical results obtained from CPLEX indicate that shared segmented protection (SSP) provides significant savings (up to 41%) in capacity utilization over dedicated and shared end-to-end protection schemes; dedicated segmented protection (DSP) provides marginal savings (up to 39%) in capacity utilization over dedicated and shared end-to-end protection schemes. The numerical results also indicate that shared segmented protection achieves the best performance followed by dedicated segmented protection and shared end-to-end protection, w.r.t. the number of requests accepted, given the network capacity.

We propose a probabilistic preemptive burst segmentation (PPBS) scheme to provide traffic classes with different quality-of-service (QoS) requirements in optical burst switching (OBS) networks. PPBS enables high priority bursts to preempt and segment low priority bursts in a probabilistic fashion. By tuning a preemptive parameter that can be instrumented locally on an OBS node, our scheme provides a flexible method to achieve service differentiation on an OBS switch for multiple prioritized traffic classes. We develop an analytical queueing model that captures the notion of increased bandwidth and parallel service using multiple wavelengths. It achieves 100% isolation between high and low priority classes and low loss probabilities in the single and multiple wavelength case. Specifically, our queueing model can achieve a proportional differentiated service (PDS) in terms of loss. Finally, we also compare the performance of PPBS with previous heuristic methods in achieving proportional loss differentiation using long range dependent traffic models.

This paper presents an efficient protection scheme to dynamically allocate restorable bandwidth-guaranteed paths in integrated IP-over-WDM networks. A restorable bandwidth-guaranteed path implies an active path and another link-disjoint backup path both with the required bandwidth. Dynamic traffic where connection requests arrive one by one with no knowledge about future requests is assumed. We consider integrated IP/WDM networks with generalized multi-protocol label switching (GMPLS) capabilities. To optimize the network resources, we provide the protection at the multi-protocol label switching (MPLS) layer and backup resources are judiciously shared among different requests at the label switched path (LSP) level. To allocate the restorable bandwidth-guaranteed paths, we propose two integrated routing algorithms: hop-based integrated routing algorithm (HIRA) and bandwidth-based integrated routing algorithm (BIRA). We use connection blocking probability and number of optical-electrical-optical (o-e-o) conversions as performance metrics to evaluate these two algorithms. Through extensive simulation experiments, we demonstrate that our algorithms perform better in terms of the above two metrics than other existing approaches.

In this paper, we consider the problem of fast and efficient dynamic scheduling of bursts that belong to different classes of priority in wavelength-division multiplexing (WDM)-based optical burst switching (OBS) networks with limited optical buffers. In OBS networks, control and data components of a burst are sent separately with a time gap to ensure that resources such as wavelengths and fiber delay lines (FDLs) are reserved at various nodes before the data burst arrives. A scheduling algorithm with attractive features such as computational simplicity and efficient resource utilization is mandatory to quickly handle dynamic burst traffic and reduce burst dropping probability. While void filling algorithms achieve good burst dropping performance they are computationally complex. We propose burst rescheduling as an alternative to void filling which can do fast scheduling without requiring to examine and fill voids and at the same time can achieve good performance. Burst rescheduling uses two mechanisms known as wavelength reassignment and lasthop FDL reassignment. We develop a scheduling algorithm using the above rescheduling mechanisms called burst rescheduling with wavelength and last-hop FDL reassignment (BR-WFR) which is computationally simpler than a void filling algorithm. We then discuss the signalling overhead and feasibility of implementing burst rescheduling. Through simulation experiments we demonstrate the effectiveness of the proposed burst rescheduling algorithm.

A control scheme which is used to set up and tear down lightpaths, should not only be fast and efficient, must also be scalable, and should try to minimize the number of blocked connections; while satisfying the requested level of fault-tolerance. In this work we propose a distributed control scheme based on preferred link approach for establishing reliability-constrained least-cost lightpaths, by choosing the reliability of a lightpath to denote the level of fault-tolerance required by the connection request. Four heuristics are proposed and their performance is studied through extensive simulation experiments. The simulation results suggest that our heuristics provide better performance compared to other distributed protocols available, in terms of average call acceptance rate, average path cost, average routing distance, and average connection setup time; when the connection requests with different levels of fault-tolerance requirements arrive to and depart from the network randomly.

Integrated IP over wavelength division multiplexed (WDM) networks, with generalized multiprotocol label switching (GMPLS) control and management capabilities, are promising candidates for the next generation optical Internet networks. By using a unified control plane, such networks make more efficient use of network resources both at the IP layer and the WDM optical layer. We consider prioritized routing of bandwidth-guaranteed label switched paths (LSPs) providing service differentiation between classes of high and normal priority traffic. The QoS delay requirements are assumed to be translated into bandwidth and opto-electronic-opto (OEO) conversion requirements. We develop a graphical representation of the integrated network state which is different from other conventional graphical representations in that it models the cost of usage of ports and OEO conversions in addition to the usage of wavelength and bandwidth resources. We then develop a threshold-protection-based routing algorithm which admits high-priority LSPs in preference over normal-priority LSPs and satisfies the bandwidth and OEO constraint requirements. Through extensive simulation experiments, we demonstrate the service differentiation and effectiveness of our algorithm.

Wavelength-division multiplexed (WDM) networks using wavelength routing are emerging to be the right choice for the future transport networks. In a WDM-based transport network, the optical layer provides circuit-switched lightpath services to the client layer such as IP, SONET, and ATM. The set of lightpaths in the optical layer defines the virtual topology. Since the optical switches (cross-connects) are reconfigurable, the virtual topology can be reconfigured in accordance with the changing traffic demand pattern at the client layer in order to optimize the network performance. On the other hand, changing the virtual topology can be disruptive to the network since the traffic at each node must be buffered or re-routed while the topology is being reconfigured. We develop a reconfiguration algorithm to reduce the cost of virtual topology reconfiguration in WDM optical ring networks. The algorithm is based on the concept of splitting and merging existing lightpaths, together with cost-benefit analysis to reduce the network reconfiguration cost. Our objective is to reduce the number of lightpaths that need to be reconfigured, while ensuring that the network congestion is low. The performance of the algorithm is verified through simulation experiments.

We consider the problem of establishing dependable connections in WDM networks with dynamic traffic demands. We call a connection with fault-tolerant requirements a dependable connection (D-connection). We consider the single-link failure model in our study and recommend the use of a proactive approach, wherein a D-connection is identified with the establishment of the primary lightpath and a backup lightpath at the time of honouring the connection request. We develop algorithms to select routes and wavelengths to establish D-connections with improved blocking performance. The algorithms use the backup multiplexing technique to efficiently utilize the wavelength channels. To further improve channel utilization, we propose a new multiplexing technique called primary-backup multiplexing. Here, a connection may not have its backup lightpath readily available throughout its existence. We develop algorithms based on this technique to route D-connections with a specified restoration guarantee. We present an efficient and computationally simple heuristic to estimate the average number of connections per link that do not have backup lightpaths readily available upon a link failure. We conduct extensive simulation experiments on different networks to study the performance of the proposed algorithms.

This paper considers the problem of dynamically establishing dependable connections (D-connections) with specified failure restoration guarantees in wavelength-routed wavelength division multiplexed (WDM) networks. We call a connection with fault-tolerant requirements a D-connection. We recommend using a proactive approach to fault-tolerance wherein a D-connection is identified with the establishment of a primary and a backup lightpath at the time of honoring the connection request. However, the backup lightpath may not be available to a connection throughout its existence. Upon occurrence of a fault, a failed connection is likely to find its backup path available with a certain specified guarantee. We develop algorithms to select routes and wavelengths to establish D-connections with specified failure restoration guarantees. The algorithms are based on a technique called primary-backup multiplexing. We present an efficient and computationally simple method to estimate the average number of connections per link for which the backup paths are not readily available upon occurrence of a link failure. This measure is used for selecting suitable primary and backup lightpaths for a connection. We conduct extensive simulation experiments to evaluate the effectiveness of the proposed algorithms on different networks. The results show that the blocking performance gain is attractive enough to allow some reduction in guarantee. In particular, under the light load conditions, more than 90% performance gain is achieved at the expense of less than 10% guarantee reduction.

Optical networks based on wavelength division multiplexing (WDM) and wavelength routing are considered to be potential candidates for the next generation of wide area networks. One of the main issues in these networks is the development of efficient routing algorithms which require a minimum number of wavelengths. We focus on the permutation routing problem in multistage WDM networks which we call 2-multinets. We present a simple, oblivious probabilistic approach which solves the permutation routing problem on 2-multinets with very high probability (in the usual theoretical sense) using O(log/sup 2/N/loglogN) wavelengths, where N is the number of nodes in the network, thereby improving the previous result due to Pankaj and Gallager (1995) that requires O(log/sup 3/ N) wavelengths. Our approach is advantageous and practical as it is simple, oblivious, and suitable for centralized as well as distributed implementations. We also note that O(logN) wavelengths will suffice with good probabilistic guarantee for the case of dynamic permutation routing where requests arrive and terminate without any relation to each other. The above results are for networks with wavelength converters and we show that the use of converters can be eliminated at the expense of a factor of log N more wavelengths. We also show how our approach can be used to solve the dynamic permutation routing problem well (in practice), using O(1) wavelengths on the hypercube and O(logN) wavelengths on the Debruijn network. These improve the previous known bounds of O(logN) and O(log/sup 2/N), respectively.

The computing resources offered by a collection of workstations can be enormous. To locate and utilize idle workstations efficiently has been a problem requiring much study. A process migration facility is one such method aimed at better utilization of the vast computing potential that allows the user of a workstation to off-load programs onto idle workstations, thereby providing the user with access to computational resources far beyond that provided by his or her personal workstation. PMDOS: Process Migration on a network of DOS workstations is a distributed processing system comprising a checkpoint/restart mechanism and a process migration mechanism at user level. It was designed and implemented on a network of DOS machines (PCs) to create an environment wherein temporarily unused workstations can be accessed by other users on the network to perform distributed or parallel computations or execute background or non-interactive jobs. An owner of a workstation need not be aware that the machine is being used during idle times; the machine is immediately returned when the owner begins to work again. The system gives good throughput, fault tolerance and efficiency for parallel and distributed applications.