Dynamic Traffic Demands Research Articles

Tapping the Knowledge of Dynamic Traffic Demands for Optimal CDN Design

The content delivery network (CDN) intensively uses cache to push the content close to end users. Over both traditional Internet architecture and emerging cloud-based framework, cache allocation has been the core problem that any CDN operator needs to address. As the first step for cache deployment, CDN operators need to discover or estimate the distribution of user requests in different geographic areas. This step results in a statistical spatial model for the user requests, which is used as the key input to solve the optimal cache deployment problem. More often than not, the temporal information in user requests is omitted to simplify the CDN design. In this paper, we disclose that the spatial request model alone may not lead to truly optimal cache deployment and revisit the problem by taking the dynamic traffic demands into consideration. Specifically, we model the time-varying traffic demands and formulate the distributed cache deployment optimization problem with an integer linear program (ILP). To solve the problem efficiently, we transform the ILP problem into a scalable form and propose a greedy diagram to tackle it. Via experiments over the North American ISPs points of presence (PoPs) network, our new solution outperforms traditional CDN design method and saves the overall delivery cost by 16% to 20%. We also study the impact of various traffic demand patterns to the CDN design cost, via experiments with both real-world traffic demand patterns and extensive synthetic trace data.

SAGECELL: Software-Defined Space-Air-Ground Integrated Moving Cells

Ultra-dense networks (UDNs) provide an effective solution to accommodate the explosively growing data traffic of multimedia services and real-time applications. However, the densification of large numbers of static small cells faces many fundamental challenges, including deployment cost, energy consumption and control, and so on. This motivates us to develop software-defined space-air-ground integrated moving cells (SAGECELL), a programmable, scalable, and flexible framework to integrate space, air, and ground resources for matching dynamic traffic demands with network capacity supplies. First, we provide a comprehensive review of state-of-the-art literature. Then the conceptual architecture of SAGECELL is elaborated in detail, and the technological benefits are emphasized. Next, we present four typical application cases of SAGECELL. A case study is conducted based on real-world road topology to validate the efficiency and flexibility of SAGECELL. Finally, we identify future research directions and challenges, and conclude this article.

Neural Network Meets DCN

The emerging optical/wireless topology reconfiguration technologies have shown great potential in improving the performance of data center networks. However, it also poses a big challenge on how to find the best topology configurations to support the dynamic traffic demands. In this work, we present xWeaver, a traffic-driven deep learning solution to infer the high-performance network topology online. xWeaver supports a powerful network model that enables the topology optimization over different performance metrics and network architectures. With the design of properly-structured neural networks, it can automatically derive the critical traffic patterns from data traces and learn the underlying mapping between the traffic patterns and topology configurations specific to the target data center. After offline training, xWeaver generates the optimized (or near-optimal) topology configuration online, and can also smoothly update its model parameters for new traffic patterns. We build an optical-circuit-switch-based testbed to demonstrate the function and transmission efficiency of our proposed solution. We further perform extensive simulations to show the significant performance gain of xWeaver, in supporting higher network throughput and smaller flow completion time.

Neural Network Meets DCN

The emerging optical/wireless topology reconfiguration technologies have shown great potential in improving the performance of data center networks. However, it also poses a big challenge on how to find the best topology configurations to support the dynamic traffic demands. In this work, we present xWeaver, a traffic-driven deep learning solution to infer the high-performance network topology online. xWeaver supports a powerful network model that enables the topology optimization over different performance metrics and network architectures. With the design of properly-structured neural networks, it can automatically derive the critical traffic patterns from data traces and learn the underlying mapping between the traffic patterns and topology configurations specific to the target data center. After offline training, xWeaver generates the optimized (or near-optimal) topology configuration online, and can also smoothly update its model parameters for new traffic patterns. The experiment results show the significant performance gain of xWeaver in supporting smaller flow completion time.

Self-organization models of urban traffic lights based on digital infochemicals

This paper presents a self-organizing model to design effective traffic signaling strategies in order to reduce traffic congestion in urban areas. The proposed traffic signaling system is based on a pattern model of self-organization, i.e., digital infochemicals (DIs), which are analogous to chemical substances that convey information between interactive elements mediated via the environment. In the context of traffic systems, the DIs refer to information generated by vehicles and dissipated by the urban transportation infrastructure. Based on the exploratory analysis with one single intersection, we demonstrate that the DI-based strategy performs significantly better than both the fixed and trigger-based scheduling strategies in terms of queue length and waiting time under both fixed and dynamic traffic demands.

Network Topology Reconfiguration for FSO-Based Fronthaul/Backhaul in 5G+ Wireless Networks

Using free-space optics (FSO) in the mobile fronthaul/backhaul networks is a promising solution that can potentially enhance the capacity of mobile networks. The main challenge of establishing FSO-based fronthaul/backhaul networks is the dynamic feature of the networks, i.e., the fragile links under adverse weather conditions and the dynamic traffic demands among small cells. Topology reconfiguration is proposed to dynamically rearrange the FSO links. In this paper, two types of network reconfigurations are investigated, in which the proactive network reconfiguration is to proactively optimize the topology of the FSO-based fronthaul/backhaul networks for each specific time period based on the statistics of the network, and the reactive reconfiguration is to actively adjust the topology when traffic demands and link states are changed. The proactive network reconfiguration is formulated as a mixed integer nonlinear programming (MINLP) problem, which jointly optimizes the network throughput and power consumption of the FSO-based fronthaul/backhaul networks. A greedy algorithm is proposed to derive the solution of the proposed problem. After the network topology having been reconfigured by solving the MINLP problem, two reactive reconfiguration algorithms are designed to optimally adjust the network topology once a link failure and demand explosion occur, respectively, to further enhance the network throughput and reduce the power consumption. The performance of the algorithms is demonstrated via extensive simulations.

Green Link Weights for Disruption-Free Energy-Aware Traffic Engineering

Energy-aware traffic engineering (ETE) has been gaining increasing research attentions due to the cost reduction benefits that it can offer to network operators and for environmental reasons. While numerous approaches exist which attempt to provide energy reduction benefits by intelligently manipulating network devices and their configurations, most of them suffer from one fundamental shortcoming: however, minor adaptations to a given IP network topology configuration all lead to temporal service disruptions incurred by routing reconvergence, which makes these schemes less appealing to network operators. The more frequently the IP topology reconfigurations take place in order to optimize the network performance against dynamic traffic demands, the more frequently service disruptions will occur to end users. Motivated by the essential requirement for network operators to enable seamless service assurance, we put forward a framework for disruption-free ETE, which leverages on selective link sleeping and wake-up operations in a disruption-free manner. The framework allows for maximizing the opportunities for disruption-free reconfigurations based on intelligent IGP link weight settings, assisted by a dynamic scheme that optimizes the reconfigurations in response to changing traffic conditions. As our simulation-based evaluation show, the framework is capable of achieving significant energy saving gains while at the same time ensuring robustness in terms of disruption avoidance and resilience to congestion.

Migration From Fixed to Flexible Grid Optical Networks With Sub-Band Virtual Concatenation

A novel strategy is presented to operate a mixed fixed/flexible grid optical network which may be used to gradually migrate from a fixed to a flexible mode of operation. Routing and spectrum allocation (RSA) in an optical network where fixed and flexible grids co-exist using a multi-path sub-band virtual concatenation (VCAT) or split spectrum (SS) technique is considered. Mixed integer linear programming models and an efficient heuristic algorithm based on spectrum window planes are proposed for the RSA optimization. The results obtained for the static traffic demand indicate that it is operationally more convenient to use the multi-path VCAT only by itself in a mixed grid optical network to guarantee performance comparable to that of the joint multi-path and single-path VCAT case. Network performance is also evaluated in terms of bandwidth blocking probability (BBP) under a dynamic traffic demand. Simulation results show that the multi-path VCAT scheme can efficiently utilize the overall spectrum resources with low blocking. The results of studies with both static and dynamic traffic demands also confirm that migration from a pure fixed grid optical network to a pure flexible grid will be desirable for better network capacity utilization.

Channel-Aware Polling-Based MAC Protocol for Body Area Networks: Design and Analysis

Body area networks (BANs) enable wearable/implanted devices to exchange information or collect monitored data. The channel quality of a link in a BAN is typically highly dynamic, since sensors equipped on a human body usually move with gesture, posture, or mobility. Therefore, existing sleep-wake-up scheduling mechanisms used in traditional static sensor networks could be very inefficient in a BAN, because they do not consider channel fluctuation of body sensors. Sensors might be waked up to transmit during bad channel conditions, leading to transmission failures and energy waste. To remedy this inefficiency, this paper proposes a Channel-aware Polling-based MAC protocol CPMAC. Our design only wakes sensors up and triggers them to transmit when the channel is strong enough to ensure fast and reliable transmissions. We further analyze the energy consumption and derive a queueing model to estimate the probability of completing all data transmissions of all sensors in our CPMAC. Benefiting from these analyses, we are able to optimize energy efficiency of our CPMAC by adapting the number of polling periods in a superframe to dynamic traffic demands and channel fluctuation. Our simulation results show that, as compared with TDMA-based scheduling and the IEEE 802.15.6 CSMA/CA protocol, CPMAC significantly improves energy efficiency and, meanwhile, keeps the latency short.

A universal frequency reuse scheme in LTE‐A heterogeneous networks

Effective inter-cell interference mitigation has been extensively studied because of its outstanding cell-edge signal quality improvement capability. Conventional static inter-cell interference coordination strategies, including fractional frequency reuse and soft frequency reuse, have received much attention owing to their effectiveness in mitigating interference and low complexity in implementation. However, they are less effective when dealing with dense uneven traffic distributions and dynamic traffic demands and thus incur low spectrum utilization in some cells and spectrum shortage in others. This paper proposes a universal frequency reuse scheme in a two-layer Long Term Evolution-Advanced heterogeneous network to ensure good throughput for all user equipment (UE), especially UEs at cell edge. The proposed scheme allows each cell to use all the spectrum resources, limited by an orderly regulation of all sub-bands. This scheme minimizes the potential occurrence probability of inter-cell co-sub-band interference through an intra-cell sub-band resource management. Furthermore, a graph-theoretic based sub-band allocation algorithm is developed to optimize UE throughput performance, especially for the cell-edge low signal to interference noise ratio UEs. A comprehensive performance comparison among different frequency reuse schemes is conducted by considering performance metrics, including cell-edge throughput, average throughput, and signal to interference noise ratio cumulative distribution function. Simulation result shows that the universal frequency reuse scheme outperforms other two schemes significantly. Copyright © 2016 John Wiley & Sons, Ltd.

Optical Power Control in GMPLS Control Plane

The exponential traffic growth in optical networks has triggered the evolution from fixed-grid to flex-grid technology. This evolution allows better spectral efficiency and spectrum usage over current optical networks in order to facilitate huge dynamic traffic demands. The promise of flex-grid technology in terms of increasing the number of optical channels established over optical links, however, may not be sustainable because of the associated increase in optical amplification power. In this work, we detail a power control process that takes advantage of link optical power and channel optical signal-to-noise ratio (OSNR) margins to allow network operators to support this optical power increase while maintaining the use of legacy optical amplifiers. New generalized multiprotocol label switching (GMPLS) protocol extensions are proposed on which to integrate the optical power control process in the control plane. The performance of the process is evaluated in terms of the blocking ratio and network throughput over fixed-grid and flex-grid networks. Results show that controlling optical power benefits from the flex-grid technology in terms of spectrum and capacity gain and reduces optical connection blocking.

Green cell planning and deployment for small cell networks in smart cities

In smart cities, cellular network plays a crucial role to support wireless access for numerous devices anywhere and anytime. The future 5G network aims to build the infrastructure from mobile internet to connected world. Small Cell is one of the most promising technologies of 5G to provide more connections and high data rate. In order to make the best use of small cell technology, smart cell planning should be implemented to guarantee connectivity and performance for all end nodes. It is particularly a challenging task to deploy dense small cells in the presence of dynamic traffic demands and severe co-channel interference. In this paper, we model various traffic patterns using stochastic geometry approach and propose an energy-efficient scheme to deploy and plan small cells according to the prevailing traffic pattern. The simulation results indicate that our scheme can meet dynamic traffic demands with optimized deployment of small cells and enhance the energy efficiency of the system without compromising on quality-of-service (QoS) requirements. In addition, our scheme can achieve very close performance compared with the leading optimization solver CPLEX and find solutions in much less computational times than CPLEX.

Software-defined hyper-cellular architecture for green and elastic wireless access

To meet the surging demand of increasing mobile Internet traffic from diverse applications while maintaining moderate energy cost, the radio access network (RAN) of cellular systems needs to take a green path into the future, and the key lies in providing elastic service to dynamic traffic demands. To achieve this, it is time to rethink RAN architectures and expect breakthroughs. In this article, we review the state-of-art literature which aims to renovate RANs from the perspectives of control-traffic decoupled air interface, cloud-based RANs, and software-defined RANs. We then propose a software-defined hyper-cellular architecture (SDHCA) that identifies a feasible way of integrating the above three trends to enable green and elastic wireless access. We further present key enabling technologies to realize SDHCA, including separation of the air interface, green base station operations, and base station functions virtualization, followed by our hardware testbed for SDHCA. Besides, we summarize several future research issues worth investigating.

Dynamic Routing and Spectrum Allocation in Elastic Optical Networks With Mixed Line Rates

Elastic optical networks (EONs) are considered a very promising solution for next-generation optical networks. Elastic spectral bandwidth allocation promotes spectrum utilization efficiency and thus increases the network capacity. One of the fundamental problems in EONs is the routing and spectrum allocation (RSA) problem. Since real EONs may provide only a few regular line rates, we comprehensively study in this paper the dynamic RSA problem in EONs with several mixed line rates. To solve the dynamic RSA problem efficiently, we decompose the problem into RSA subproblems. For the routing subproblem, we propose an efficient multiconstrained routing algorithm named sorted feasible paths searching (SFPS) to find the shortest feasible paths for dynamic traffic demands. The completeness, optimality, and complexity of SFPS are proved. For the spectrum allocation subproblem, we propose two spectrum allocation strategies named fixed segmentation and adaptive segmentation to assign spectrum for the noncommensurate traffic demands of EONs with mixed line rates. Simulation results prove that the proposed dynamic RSA algorithms are time efficient and perform better than existing dynamic RSA algorithms in terms of bandwidth blocking probability and spectrum fragmentation ratio in EONs with mixed line rates.

Research on Optimized Control Model of Freeway Based on Dynamic Traffic Demand Estimation

According to the estimation information of dynamic traffic demands, a novel optimal control model of freeway was established on the basis of the hierarchical concept. There are four control modules in this model. The OD prediction module predicts the total traffic demands in a long time and determines the upper bound of the future queuing length in advance; the global optimal control module predicts the future traffic state and establishes the coordination constraints for each ramp in the network; the traffic demand estimation module estimates the real-time traffic conditions for each ramp; the local adaptive control module regulates ramp metering rate according to the estimated information of the real-time traffic conditions and the results optimized by the global optimal control module. The simulation results show that this control system is of a good dynamic performance. It coordinates the benefits of various ramps and optimizes the overall performance of the freeway network.

Flex-grid optical networks: spectrum allocation and nonlinear dynamics of super-channels

Flex-grid optical networks have evolved as a near-future deployment option to facilitate dynamic and bandwidth intense traffic demands. These networks enable capacity gains by operating on a flexible spectrum, allocating minimum required bandwidth, for a given channel configuration. It is thus important to understand the nonlinear dynamics of various high bit-rate super-channel configurations, and whether such channels should propagate homogenously (uniform channel configuration) or heterogeneously (non-uniform channel configuration), when upgrading the current static network structure to a flex-grid network. In this paper, we report on the spectrum allocation strategies based on the impact of inter-channel fiber nonlinearities, for PM-16QAM channels (240Gb/s, 480Gb/s and 1.2Tb/s) -termed as super-channels, propagating both homogenously, and heterogeneously with 120Gb/s PM-QPSK, 43Gb/s PM-QPSK, and 43Gb/s DPSK traffic. In particular, we show that for high dispersion fibers, both homogenous and heterogeneous spectrum allocation enable similar performance, i.e. the nonlinear impact of hybrid traffic is found to be minimal (less than 0.5dB relative penalties). We further report that in low dispersion fibers, the impact of spectrum allocation is more pronounced, and heterogeneous traffic employing 120Gb/s PM-QPSK neighbors enables the best performance, ~0.5dB better than homogenous transmission. However, the absolute nonlinear impact of co-propagating traffic is more significant, compared to high dispersion fibers, with maximum performance penalties up to 1.5dB.

Decentralized Bargain: A Two-Tier Market for Efficient and Flexible Dynamic Spectrum Access

Market mechanisms have been exploited as important means for spectrum acquisition and access in cognitive radio networks. In this paper, we propose a two-tier market for decentralized dynamic spectrum access. In the proposed Tier-1 market, spectrum is traded from a primary user (PU) to secondary users (SUs) in a relatively large time scale to reduce signaling overhead. Then, driven by dynamic traffic demands, SUs set up the Tier-2 market to redistribute channels among themselves in a small time scale. More specifically, we use a Nash bargain game to model the spectrum acquisition of SUs in the Tier-1 market and derive the equilibrium prices. We then employ a strategic bargain game to study the spectrum redistribution in the Tier-2 market, where SUs can exchange channels with low overhead through random matching, bilateral bargain, and the predetermined market equilibrium price. We investigate how various factors, such as the availability of channels and bargain partners, matching strategies, and traffic dynamics, affect the market relationships. This work provides new understanding on the spectrum market and valuable guidelines to primary and secondary network operators.

Delay-Jitter aware slot assignment for Real-Time applications in wireless multimedia ad hoc sensor networks

This paper presents a distributed time slot assignment algorithm which adopts TDMA as medium access control, specially suited to support applications with tight delay, jitter and throughput requirements in convergecast traffic network, e.g. wireless multimedia surveillance sensor network. The proposed algorithm basically consists of two parts. The first part is the slot assignment for single node, which minimizes delay and jitters for dynamically arrived message streams and optimizes stream acceptation ratio. The second part is the slot assignment for multi-hop network, which guarantees every node having a path to the base station for its data delivery. In the path, sufficient resource is reserved and QoS can be guaranteed. For different applications, we use a dynamic programming technique to optimize TDMA frame length in accordance with their QoS requirements. The proposed algorithm is further optimized for dealing with unreliable links and dynamic traffic demands. It is evaluated and compared with existing slot assignment algorithms. The results show that our algorithm provides lower end-to-end delay and jitter, but higher throughput.

A Novel Optimization and Control Method for Signalized Traffic Network

For the dynamic optimization and control problem of the signalized arterial network in urban, a dynamic traffic flow model based on multi-phase control is firstly formulated, in which the total number of the retained vehicles through the arterial during the control period is adopted as the optimization objective and the green times and offsets as the control variables. Then a hybrid optimization method based on real-coded genetic algorithm and local search technique is designed to solve the optimization problem. For examining the validity of the optimization and control method proposed, it is applied to a case study with dynamic traffic demands and a large number of simulations show that the dynamic optimization and control method proposed in this paper can work well for the signalized arterial network.

ROADM optimization in WDM ring networks

Reconfigurable optical add–drop multiplexers (ROADMs) and tunable transceivers play an important role in optical network reconfigurability and cost saving. ROADMs provide support to dynamic traffic demands, enable new services, and save capital expenditure (CapEx) along with the operational expenditure (OpEx). Limiting the range of wavelengths that can be accessed at a node reduces costs, and possibly reduces flexibility in reconfigurability. The limited range (band size) of wavelengths is a result of the limited add/drop range of the ROADMs and/or limited tuning range of transceivers. In either case, limited reconfigurability introduces a new constraint in the wavelength assignment of connections called as the wavelength termination constraint. Limited ROADMs (L-ROADMs) can add and drop only a subset of (contiguous) wavelengths (a waveband) and are less costly than fully reconfigurable OADMs (F-ROADMs). In order to quantify the trade-offs between the number of L-ROADMs and their band sizes and compare those with F-ROADMs, optimization solutions must be obtained. In this paper, we consider two models of L-ROADMs and present optimization problems specific to each model under all-to-all traffic for ring networks. We consider two ring topologies, uni- and bi-directional rings. We give integer linear programming formulations and present heuristic solutions to minimize the worst-case band size and the total number of wavebands for the two L-ROADM types. For the uni-directional ring in the first model, we present a lower bound of ⌈ ( W + 2 ) / 2 ⌉ for the band size of ROADMs, where W is the number of wavelengths. For the upper bound, our results show that the worst-case band size of L-ROADMs is only about 65% of the full range for large network sizes. For the bi-directional ring of N nodes, our heuristic algorithm, Algorithm-Min-Bi, produces a worst-case band size of b ( w ) = N 2 − 4 N + 19 8 . This produces a reduction of 10% to 20% in the band size for typical network sizes. When the band size b s is fixed under the second model, we observe the trade-off between the total band size of L-ROADMs and the number of L-ROADMs. For both uni- and bi-directional ring networks, our results are close to optimal. We also examine the performance of our models under dynamic traffic.