Data Center Networks Research Articles

EagerCC: An ultra-low latency congestion control mechanism in datacenter networks

With the rapid development of cloud applications, the workload pattern of datacenters presents the characteristics of mixed long and short flows and frequent micro-burst traffic, which puts forward new requirements for network transmission performance, including ultra-low latency, high throughput, and strong stability. At the same time, datacenters begin to deploy the low-diameter topology to accommodate these new requirements, and the high-performance datacenter (HPDC) comes into being. However, due to the complexity of the load, existing congestion control mechanisms cannot control dynamic delay in the network well, which significantly restricts the development of the HPDC. Therefore, it is necessary to deploy the congestion control mechanism for the HPDC. So we propose EagerCC, an ultra-low latency, low-overhead, and accurate congestion control mechanism based on In-Network-Telemetry (INT) information for various datacenter scenarios, especially for HPDCs. EagerCC uses switch-feedback, ACK-padding, ACK-first to reduce feedback delay and uses switch-calculation, probabilistic ACK-padding to reduce the overhead of congestion signals. We conduct a lot of experiments and the result shows that EagerCC performs well in various datacenter scenarios, especially for HPDCs. Specifically, EagerCC reduces the 99th-FCT and avg-FCT by 52.3% and 13.3% for the HPC workload NPB-CG compared to HPCC in Dragonfly. In addition, EagerCC significantly reduces the network’s feedback delay and queue occupancy.

Comparative analysis of optical data center switches

Abstract In the recent past, there has been an explosive growth in data traffic within data center systems. This rapid increase in data volume has put significant constraints on the speed of current data communication networks, which are predominantly based on electronic controllers and memory. The speed limitations of electronic devices hinder their ability to process data at higher rates efficiently. To address this issue and enable faster data processing, the use of optical communication emerges as the most promising solution. Optical communication leverages the capabilities of light signals and optical components, which can handle data at much higher rates compared to their electronic counterparts. One key element in optical communication systems is the utilization of fiber delay lines (FDLs) as optical storage for packets. FDLs offer the ability to temporarily store and delay optical signals, providing buffering capabilities within optical networks. These FDLs can be configured in various setups, including feed forward, feed backward, and recirculating configurations, each with its own set of advantages and disadvantages. In this context, this article explores several state-of-the-art arrayed waveguide grating (AWG)-based optical packet switch (OPSW) architectures. These switches are designed to efficiently route and manage optical packets within data center networks. Each architecture is summarized, highlighting its unique features and capabilities. To compare the different AWG-based OPSW architectures, various key parameters are considered, such as the cost, physical losses, bit error rate (BER), and the number of tunable components required. These parameters are critical in determining the performance, efficiency, and cost-effectiveness of the switches.

NDN-Based Coded Caching Strategy for Satellite Networks

To solve the transmission correlation issue arising from traditional named data networking (NDN) caching during segmenting contents, where users must obtain data blocks related to the requested file blocks to recover the source data, in this paper, we propose an NDN-based coded caching strategy for low Earth orbit (LEO) satellite networks, using coding operations to remove the transmission correlation of data. To achieve efficient content distribution, the satellite node’s coded package is strategically placed with the objectives of minimizing the backhaul link load and content acquisition latency. The optimization problem is solved by using a multi-colony ant colony algorithm, enabling fast content retrieval. We designed a cluster cooperative distribution mechanism to simplify the satellite network’s management and reduce the load of satellite links for content distribution when covered by multiple satellite nodes. Finally, we compare the multi-objective coded caching (CCMO) introduced in this paper with the most popular (MP) strategy and the random-based coded caching (CCR) strategy; the average cache hit ratio, backhaul link load, and content acquisition delay are compared and analyzed. The results show that CCMO performs better.

On estimating the interest satisfaction ratio in IEEE 802.15.4-based named-data networks

On estimating the interest satisfaction ratio in IEEE 802.15.4-based named-data networks

A review of modern caching strategies in named data network: overview, classification, and research directions

A review of modern caching strategies in named data network: overview, classification, and research directions

Optical Switching Data Center Networks: Understanding Techniques and Challenges

Optical Switching Data Center Networks: Understanding Techniques and Challenges

Decentralized Receiver-based Link Stability-aware Forwarding Scheme for NDN-based VANETs

The inter-vehicle communication in the vehicular ad-hoc network (VANET) providing ubiquitous mobility to mobile users is considered an essential component of intelligent transportation systems (ITS), enabling a wide range of safety and non-safety applications in which the vehicles seamlessly interact and co-operate with each other. The recent paradigm shift in VANETs protocols has led to the adoption of name data networking (NDN) as an underlying communication protocol in VANETs that replaces traditional address-based communication philosophy, treats content as a first-class citizen, and assigns unique names to contents consequently enabling name-based routing and forwarding, caching, and security. Although NDN has significantly improved the underlying communication issues in VANETs, a plethora of challenges such as message broadcast storms, in-efficient packet suppression mechanisms, intermittent connectivity owing to frequent topological changes, and reverse path partioning (RPP) require further attention from the research community. To cope with these issues, this paper proposes decentralized receiver-based link stability-aware forwarding (DRLSF) protocol. The DRLSF is a beacon-less receiver-based multi-hop protocol, well-suited for pull-based applications where end users request desired data by sending packets. The comparative performance analysis validates the effectiveness of the DRLSF protocol in different VANET environments and application scenarios.

Partial response THP for high-speed PAM-4 transmission

Intensity-modulation direct-detection (IM/DD) pulse-amplitude modulation (PAM-4) signals are promising for high-speed data center networks. However, high-speed PAM-4 signals always are subject to critical impairments due to the bandwidth constraint. In the conventional PAM-4 system, a direct detection-faster than Nyquist (DD-FTN) with high complexity is usually appended to the feed-forward equalizer (FFE) for colored noise suppression to maximize transmission performance. In order to yield a simple PAM-4 scheme with similar performance as that of the DD-FTN, we proposed a nonlinear differential coding-generalized Tomlinson-Harashima Precoding (NLDC-GTHP) PAM-4 system. The proposed system consists of a NLDC-GTHP at the transmitter and a simplified two-dimensional constellation distortion (S2DCD) module at the receiver. We experimentally demonstrated the proposed NLDC-GTHP PAM-4 system for 62.5-Gb/s PAM-4 signal transmission over 2-km standard single mode fiber (SSMF) with the 10-dB bandwidth of about 12 GHz. By introducing the NLDC-GTHP and S2DCD, we achieved the similar receiver sensitivity as that of the DD-FTN at the hard-decision forward-error-correction (HD-FEC) bit error rate (BER) threshold, with the decrease of 86% computational complexity.

Availability Model for Data Center Networks With Dynamic Migration and Multiple Traffic Flows

The evaluation of application availability is a crucial step in ensuring functionality of the data center network (DCN). The application of data center network is no more a chain of physical devices, but a chain of virtualized network function (VNF) dynamically installed in servers, i.e., the service function chain (SFC), due to the adaptation of network function virtualization (NFV). When dynamic migration is considered, the application availability evaluation would encounter problems since the state of VNF components is not independent. And in the application deployed in DCN, there is multiple traffic flow because of the serialization of 1+1 (active-standby) and N-way (load-sharing) redundant VNF, however only single flow is considered in current availability model. When dynamic migration and multiple traffic flows are considered, the application availability evaluation would encounter the following problems: 1) The application behavior is not described in the existing availability model; 2) The availability model could not quantify the influence of different traffic of each flow. This paper proposes an availability model, which consist of network evolution model (NEM) of DCN and the availability evaluation algorithm based on NEM (AEA-NEM), to evaluate the availability of different applications deployed on DCN. The NEM of DCN describes the dynamical process of migration and switch of three kinds of VNF (normal, 1+1 and N-way) in application, and the AEA-NEM evaluate the availability of application with multiple traffic flows. The simulation results show that, the steady state availability of the proposed model is consistent with it of generalized stochastic Petri net (GSPN) model. And the model could evaluate the application availability with dynamic migration and multiple traffic flows in DCN. Based on the current VNF placement and redundant strategy, the most available designing scheme is analyzed.

Path Diversity and Survivability for the HyperX Datacenter Topology

Network survivability has been recognized as an issue of a major importance in terms of security, stability and prosperity. This paper studies fundamental properties of HyperX, an emerging topology for connecting supercomputing and datacenter networks. We focus on the establishment of paths with guaranteed survivability, allowing path availability even upon a restricted number of link failures. We first examine the availability of disjoint paths connecting a pair of input nodes. Disjoint paths guarantee path existence even upon a bounded number of link failures. We explore the inherent tradeoff between allowing slightly longer paths and the ability to extend available sets of mutually disjoint paths. Second, we study the availability of paths in a HyperX topology that already observed link failures. Such failures can increase the length of available paths or even eliminate connectivity between network parts. We also compare basic properties of shortest paths in HyperX to other datacenter topologies. Last, we provide an evaluation to illustrate path availability along with the potential impact of failures.

FoggyEdge: An Information-Centric Computation Offloading and Management Framework for Edge-Based Vehicular Fog Computing

Recent advances aiming to enable in-network service provisioning are empowering a plethora of smart infrastructure developments, including smart cities and intelligent transportation systems (ITSs). Although edge computing, in conjunction with roadside units, appears to be a promising technology for proximate service computations, the rising demands for ubiquitous computing and ultralow-latency requirements from consumer vehicles are challenging the adoption of ITSs. Vehicular fog (VF) computing, which extends the fog computing paradigm in vehicular networks by utilizing either parked or moving vehicles for computations, has the potential to further reduce computation offloading transmission costs. Therefore, with a precise objective of reducing latency and delivering proximate service computations, we integrate VF computing with roadside edge computing and propose a four-layer framework named FoggyEdge. The FoggyEdge framework is built at the top of named data networking and employs microservices to perform in-network computations and offloading. A real-world Simulation of Urban Mobility-based preliminary performance comparison validates FoggyEdge’s effectiveness. Finally, a few future research directions on incentive mechanisms, security and privacy, optimal VF location, and load balancing are summarized.

Disjoint Paths Construction and Fault-Tolerant Routing in BCube of Data Center Networks

BCube is a promising structure of data center network, as it can significantly improve the performance of typical applications. With the expansion of network scale and increasement of complexity, reliability and stability of networks have become more essential. In this paper, we study the fault-tolerant routings in BCube. Firstly, we design a fault-tolerant routing algorithm based on node disjoint multi-paths. The proposed multi-path routing has stronger fault tolerance, since each path has no other common nodes except the source node and the destination node. Secondly, we investigate an effective fault-tolerant routing based on routing capabilities algorithm for BCube. The proposed algorithm has higher fault tolerance and success rate of finding feasible routes, since it does not limit the faults number. Thirdly, we present an adaptive path finding algorithm for establishing virtual links between any two nodes in BCube, which can shorten the diameter of BCube. Extensive simulation results show that the proposed routing scheme outperforms the existing popular algorithms. Compared with the state-of-the-art fault-tolerant routing algorithms, the proposed algorithm has a 21.5% to 25.3% improvement on both throughput and packet arrival rate. Meanwhile, it reduces the average latency of 18.6% and the maximum latency of 23.7% in networks.

Association of serotonin reuptake inhibitors with asthma control.

Background: Clinical trials demonstrated that selective serotonin reuptake inhibitors (SSRI) can improve asthma control in patients with comorbid major depressive disorder (MDD) and that this effect may be greater than the effect of SSRIs on depression. These findings suggest that SSRIs may improve asthma control in patients without MDD. Objective: The current retrospective study examined the effect of SSRIs and serotonin and norepinephrine reuptake inhibitors (SNRI) on asthma control in adult patients. We hypothesized that patients would have fewer asthma exacerbations after treatment with an SSRI or SNRI. Methods: Electronic health record data of adult patients (N = 592) who were seen at a University of Texas Southwestern (UTSW) hospital or clinic and had (1) an SSRI or SNRI prescription, (2) a previous asthma diagnosis, and (3) no mood disorder diagnosis were extracted by using the UTSW Clinical Data Exchange Network. Wilcoxon signed rank tests were used to compare oral corticosteroid prescriptions and asthma-related emergency department (ED) visits and hospitalizations in the 12 months before and after the start of an SSRI/SNRI. Results: Therapy with SSRIs/SNRIs was associated with a significant decrease in oral corticosteroid use (p = 0.003), ED visits (p = 0.002), and hospitalizations (p < 0.001). Conclusion: Results from the current study add to the existing literature by demonstrating a reduced rate of severe exacerbations in patients with asthma by using an SSRI/SNRI without limiting the analytic sample to a high-illness-severity subgroup defined by symptoms of asthma or depression. Future work should include a prospective, placebo controlled study with individuals who have asthma and no comorbid mental health condition, verified by a mental health professional.

ABLA: Application-Based Load-Balanced Approach for Adaptive Mapping of Datacenter Networks

Cloud-based services are growing more rapidly than ever, and so does the management challenge on their providers’ side. Cloud-based datacenter networks are built with nodes of huge processing power, connected by high bandwidth capacities to carry their interior traffic requirements. However, such cloud networks still have limits that are imposed not necessarily by their physical components, but by the schemes of resource management being deployed. Traditionally, for an institute to provide services, it needs to have its own datacenter facility that interconnects its servers through a topology that matches its desired administrative policies and scaling objectives. With the theme of cloud-based IaaS, such datacenter topologies can be created virtually over the cloud. Nowadays, a significant part of those institutes who provide us with our daily services have their infrastructures hosted over cloud ones. Therefore, resources of such cloud networks need to be efficiently utilized, in order to keep their performance and hosting prices competitive. A typical datacenter network mainly consists of server nodes and network links. Besides the resources of the server nodes, the network bandwidth resources are considered a crucial key determinant for the whole datacenter performance. Indeed, a server without sufficient bandwidth capacities is almost useless. Proposals in the literature present schemes for resource utilization on either side of the problem at a time: the nodes or the links. Working in isolation can never deliver efficient mapping solutions. ABLA is an Application-Based, and Load balancing Approach for adaptive mapping proposal. ABLA’s methodology tackles both sides of the datacenter, the nodes and links. It starts by (1) breaking down the node’s resource requirement for the requested applications to be hosted over the virtual server machines besides (2) reading the topological connectivity and bandwidth requirements for each virtual node to all other nodes in the virtual datacenter topology. Compared to other models in the literature, the simulation results show that our proposed ABLA model provides for complete mapping services via load-balanced hosting networks. This allows for competitive hosting prices, with higher performance and service satisfaction rates.

A Learning-based Data Placement Framework for Low Latency in Data Center Networks

Abstract: The challenge of achieving low-latency data services has become increasingly crucial for data center applications. In modern distributed storage systems, efficient data placement plays a vital role in minimizing data movement delays, thereby significantly reducing service latency. While previous solutions for data placement have often relied on pre-assumed data request distributions or traced analysis, the dynamic nature of network conditions and evolving user request patterns poses a complex online decision-making challenge. Traditional static model-based approaches are limited in their ability to cope with such dynamic system dynamics. To address these challenges and consider both data movement and analytical latency, we introduce a novel framework named DataBot+. This framework leverages reinforcement learning techniques to autonomously learn optimal data placement policies. DataBot+ employs neural networks that are trained using a modified version of Q-learning. The input to these networks comprises real-time data flow measurements, while the output is a value function that estimates the imminent future latency. To ensure timely decision-making, DataBot+ is designed with a two-fold asynchronous architecture, consisting of production and training components. This separation guarantees that the training process does not introduce additional overhead to the handling of data flows. Our approach is validated through extensive evaluations using real-world traces, which convincingly demonstrate the efficacy of our proposed design.

Exploring Path Computation Techniques in Software-Defined Networking: A Review and Performance Evaluation of Centralized, Distributed, and Hybrid Approaches

Software-Defined Networking (SDN) is a networking paradigm that allows network administrators to dynamically manage network traffic flows and optimize network performance. One of the key benefits of SDN is the ability to compute and direct traffic along efficient paths through the network. In recent years, researchers have proposed various SDN-based path computation techniques to improve network performance and reduce congestion.&#x0D; This review paper provides a comprehensive overview of SDN-based path computation techniques, including both centralized and distributed approaches. We discuss the advantages and limitations of each approach and provide a critical analysis of the existing literature. In particular, we focus on recent advances in SDN-based path computation techniques, including Dynamic Shortest Path (DSP), Distributed Flow-Aware Path Computation (DFAPC), and Hybrid Path Computation (HPC).&#x0D; We evaluate three SDN-based path computation algorithms: centralized, distributed, and hybrid, focusing on optimal path determination for network nodes. Test scenarios with random graph simulations are used to compare their performance. The centralized algorithm employs global network knowledge, the distributed algorithm relies on local information, and the hybrid approach combines both. Experimental results demonstrate the hybrid algorithm's superiority in minimizing path costs, striking a balance between optimization and efficiency. The centralized algorithm ranks second, while the distributed algorithm incurs higher costs due to limited local knowledge. This research offers insights into efficient path computation and informs future SDN advancements.&#x0D; We also discuss the challenges associated with implementing SDN-based path computation techniques, including scalability, security, and interoperability. Furthermore, we highlight the potential applications of SDN-based path computation techniques in various domains, including data center networks, wireless networks, and the Internet of Things (IoT).&#x0D; Finally, we conclude that SDN-based path computation techniques have the potential to significantly improvement in-order to improve network performance and reduce congestion. However, further research is needed to evaluate the effectiveness of these techniques under different network conditions and traffic patterns. With the rapid growth of SDN technology, we expect to see continued development and refinement of SDN-based path computation techniques in the future.

Q-ICAN: A Q-learning based cache pollution attack mitigation approach for named data networking

The Cache Pollution Attack (CPA) is a recent threat that poses a significant risk to Named Data Networks (NDN). This attack can impact the caching process in various ways, such as causing increased cache misses for legitimate users, delays in data retrieval, and exhaustion of resources in NDN routers. Despite the numerous countermeasures suggested in the literature for CPA, many of them have detrimental effects on the NDN components. In this paper, we introduce Q-ICAN, a novel intelligent technique for detecting and mitigating cache pollution attacks in NDN. More specifically, Q-ICAN uses Q-Learning as an automated CPA prediction mechanism. Each NDN router integrates a reinforcement learning agent that utilizes impactful metrics such as the variation of the Cache Hit Ratio (CHR) and the interest inter-arrival time to learn how to differentiate between malicious and legitimate interests. We conducted several simulations using NDNSim to assess the effectiveness of our solution in terms of Cache Hit Ratio (CHR), Average Retrieval Delay (ARD) and multiple artificial intelligence evaluation metrics such as accuracy, precision, recall, etc. The obtained results confirm that Q-ICAN detects CPA attacks with a 95.09% accuracy rate, achieves a 94% CHR, and reduces ARD by 18%. Additionally, Q-ICAN adheres to the security policy of the NDN architecture and consumes fewer resources from NDN routers compared to existing state-of-the-art solutions.

A Review of Blockchain Technology in Knowledge-Defined Networking, Its Application, Benefits, and Challenges

Knowledge-Defined Networking (KDN) necessarily consists of a knowledge plane for the generation of knowledge, typically using machine learning techniques, and the dissemination of knowledge, in order to make knowledge-driven intelligent network decisions. In one way, KDN can be recognized as knowledge-driven Software-Defined Networking (SDN), having additional management and knowledge planes. On the other hand, KDN encapsulates all knowledge-/intelligence-/ cognition-/machine learning-driven networks, emphasizing knowledge generation (KG) and dissemination for making intelligent network decisions, unlike SDN, which emphasizes logical decoupling of the control plane. Blockchain is a technology created for secure and trustworthy decentralized transaction storage and management using a sequence of immutable and linked transactions. The decision-making trustworthiness of a KDN system is reliant on the trustworthiness of the data, knowledge, and AI model sharing. To this point, a KDN may make use of the capabilities of the blockchain system for trustworthy data, knowledge, and machine learning model sharing, as blockchain transactions prevent repudiation and are immutable, pseudo-anonymous, optionally encrypted, reliable, access-controlled, and untampered, to protect the sensitivity, integrity, and legitimacy of sharing entities. Furthermore, blockchain has been integrated with knowledge-based networks for traffic optimization, resource sharing, network administration, access control, protecting privacy, traffic filtering, anomaly or intrusion detection, network virtualization, massive data analysis, edge and cloud computing, and data center networking. Despite the fact that many academics have employed the concept of blockchain in cognitive networks to achieve various objectives, we can also identify challenges such as high energy consumption, scalability issues, difficulty processing big data, etc. that act as barriers for integrating the two concepts together. Academicians have not yet reviewed blockchain-based network solutions in diverse application categories for diverse knowledge-defined networks in general, which consider knowledge generation and dissemination using various techniques such as machine learning, fuzzy logic, and meta-heuristics. Therefore, this article fills a void in the content of the literature by first reviewing the diverse existing blockchain-based applications in diverse knowledge-based networks, analyzing and comparing the existing works, describing the advantages and difficulties of using blockchain systems in KDN, and, finally, providing propositions based on identified challenges and then presenting prospects for the future.

Parameterized deep reinforcement learning with hybrid action space for energy efficient data center networks

To ensure the delivery of high-performance and reliable services, data center networks (DCNs) are often over-provisioned for peak workload and traffic bursts. However, in real-world data centers, network traffic seldom reaches peak capacity of the network, resulting in significant energy waste. Traditional energy conservation approaches either suffer from high computational complexity and low solution quality, or their strategies cannot be dynamically adjusted to accommodate changes in data center network traffic. Deep reinforcement learning (DRL) provides an effective way to deal with these issues. However, most of the existing DRL-based schemes only consider either a continuous action space or a discrete action space, which greatly restricts the optimality of decisions. To solve these problems, this paper proposes a novel DRL-based DCN energy optimization framework, named SmartDCN. Specifically, SmartDCN consists of a traffic prediction module (TPM) and an energy optimization module (EOM). TPM incorporates an improved LSTM model JANET with an attention mechanism providing a high prediction accuracy, while EOM integrates our newly proposed parameterized DRL algorithm, named PAS-DQN, combining with the discrete-continuous hybrid action space. PAS-DQN implements a two-level control mechanism for the network, using TPM to predict future traffic in the data center as input. It is devoted to dynamically aggregating current traffic and makes tradeoffs between energy efficiency, performance, and robustness to optimize the network’s power consumption by dynamically calculating the minimum required network subset and turning off the non-involved network devices to achieve power savings. Experimental results show that SmartDCN significantly outperforms the existing state-of-the-art schemes in terms of energy savings under various network conditions.

An efficient scheme for in-orbit remote sensing image data retrieval

The conflict between real-time retrieval of remote sensing image data (RSID) and the limited satellite network transmission resources is becoming more and more prominent. To address it, this paper proposes a novel scheme for RSID retrieval from the remote sensing satellite network, which exploits the spatio-temporal characterization (STC) of sensing images to improve data retrieval efficiency. In particular, based on the spatio-temporal characterization, a data content naming scheme is introduced and an algorithm is accordingly proposed for RSID segmentation in satellite nodes. With the naming scheme, RSID retrieval can be implemented by making use of the Named Data Networking (NDN) architecture. The study aims to investigate the effectiveness of making use of the STC of RSID for the retrieval under the NDN architecture. To the best of our knowledge, this paper is the first to exploit STC to establish connections between user demands, RSID, and data naming for NDN. As a result, the proposed scheme makes the network aware of RSID’s STC and consequently improves the data retrieval performance of the satellite network using NDN. The evaluation results show that the proposed scheme delivers about 65% and 7% higher retrieval hit rates than a retrieval scheme based on the original NDN proposal and a recently proposed scheme — CAP, respectively. In addition, comparing with these two schemes, the proposed one reduces transmission latency by about 20% and 15%, and traffic load by about 48% and 40%, all respectively. These imply that the proposed scheme outperforms in terms of higher data retrieval efficiency and better link utilization.