Dynamic Bandwidth Allocation Research Articles

Machine Learning Techniques in Radio-over-Fiber Systems and Networks

The radio-over-fiber (RoF) technology has been widely studied during the past decades to extend the wireless communication coverage by leveraging the low-loss and broad bandwidth advantages of the optical fiber. With the increasing need for wireless communications, using millimeter-waves (mm-wave) in wireless communications has become the recent trend and many attempts have been made to build high-throughput and robust mm-wave RoF systems during the past a few years. Whilst the RoF technology provides many benefits, it suffers from several fundamental limitations due to the analog optical link, including the fiber chromatic dispersion and nonlinear impairments. Various approaches have been proposed to address these limitations. In particular, machine learning (ML) algorithms have attracted intensive research attention as a promising candidate for handling the complicated physical layer impairments in RoF systems, especially the nonlinearity during signal modulation, transmission and detection. In this paper, we review recent advancements in ML techniques for RoF systems, especially those which utilize ML models as physical layer signal processors to mitigate various types of impairments and to improve the system performance. In addition, ML algorithms have also been widely adopted for highly efficient RoF network management and resource allocation, such as the dynamic bandwidth allocation and network fault detection. In this paper, we also review the recent works in these research domains. Finally, several key open questions that need to be addressed in the future and possible solutions of these questions are also discussed.

Auction Mechanism for Dynamic Bandwidth Allocation in Multi-Tenant Edge Computing

Inspired by the shared infrastructure of colocation data centers and the growth of Mobile Edge Computing (MEC), colocation MEC businesses have thrived to offer an economical and low latency solution for MEC based tenants. In colocation MEC, a colocation service provider leases spaces at the base stations (BSs) to tenants for housing their servers. Although the tenants fully manage their own servers, they still need to purchase bandwidth from the network provider to via bandwidth purchasing market to serve their users. As a result, tenants should consider the trade-off between delay and energy consumption, which are related to bandwidth and server computing speed, respectively. In this paper, an auction framework is used to model the interaction between the network provider and tenants in the bandwidth market. The network provider determines the bandwidth amount and corresponding payment by tenants based on their submitted bids, in which the valuation of the bandwidth amount in each bid is based on the coupling between bandwidth and server computing speed. The networking provider also seeks winners in the social welfare maximization problem, which is NP-hard. To solve this problem, we propose a solution based on the randomized auction mechanism, which is proved to be approximately truthful, individually rational, and computationally efficient. Simulation results verify the effectiveness of our proposed randomized auction scheme in bandwidth trading.

Dynamic Memory Bandwidth Allocation for Real-Time GPU-Based SoC Platforms

Heterogeneous SoC platforms, comprising both general purpose CPUs and accelerators, such as a GPU, are becoming increasingly attractive for real-time and mixed-criticality systems to cope with the computational demand of data parallel applications. However, contention for access to shared main memory can lead to significant performance degradation on both CPU and GPU. Existing work has shown that memory bandwidth throttling is effective in protecting real-time applications from memory-intensive, best-effort (BE) ones; however, due to the inherent pessimism involved in worst-case execution time (WCET) estimation, such approaches can unduly restrict the bandwidth available to BE applications. In this article, we propose a novel memory bandwidth allocation scheme where we dynamically monitor the progress of a real-time application and increase the bandwidth share of BE ones whenever it is safe to do so. Specifically, we demonstrate our approach by protecting a real-time GPU kernel from BE CPU tasks. Based on profiling information, we first build a WCET estimation model for the GPU kernel. Using such model, we then show how to dynamically recompute on-line the maximum memory budget that can be allocated to BE tasks without exceeding the kernel’s assigned execution budget. We implement our proposed technique on NVIDIA embedded SoC and demonstrate its effectiveness on a variety of GPU and CPU benchmarks.

Optical Signal Processing Aided by Optical Frequency Combs

This paper highlights certain optical signal processing (OSP) functions that can be achieved when using optical frequency combs. An optical frequency comb can provide many narrow-linewidth, mutually coherent, and equidistant optical carriers, making it a potentially useful tool for facilitating efficient signal processing functions. In this paper, we explore the following general topics: (a) tailoring a comb for enhanced usability in OSP, such as inserting more lines, generating Nyquist pulse trains, and regenerating a comb at a distance; (b) achieving a tunable linear and nonlinear filter for correlation, equalization, and Volterra filtering, and (c) enabling various transmission and networking functions, such as multicasting, format conversion, and dynamic bandwidth allocation.

Satisfaction-based Dynamic Bandwidth Reallocation for multipath mobile data offloading

Mobile data offloading is a key solution to mitigate heavy traffic load through cellular networks to the WiFi networks. To motivate the mobile users to transmit data over the cellular/WiFi networks under the control of the operator smartly, many incentive pricing frameworks have been proposed with the execution of dynamic bandwidth reallocation schemes to improve network efficiency. In this paper, a Satisfaction-based Dynamic Bandwidth Reallocation (SDBR) scheme is proposed to improve user satisfaction in term of the blocking probability and the user payment in data offloading environments. In SDBR, the operator can efficiently reallocate the bandwidth to accommodate more potential users; the overall operator revenue therefore increases as well. To demonstrate the effectiveness of our proposed SDBR scheme, we first assess the blocking probability with and without serving users having WiFi connections and discuss the benefits of the dynamic bandwidth allocation. Then, we elaborate a satisfaction-based pricing scheme for data downloading to motivate both mobile operator and users to work with SDBR, which brings the increment of operator overall revenue and reduction of individual user payment. Finally, simulation results show that our proposed SDBR scheme outperforms the four existing bandwidth allocation schemes in both time-based and volume-based data transmission.

DBA dis-aggregation in PON-OLT with scalability enhancement by REPORT/GATE compression [Invited

The dis-aggregated passive optical network (PON) optical line terminal (OLT) architecture is attracting attention in academia and industry. This paper first reviews the evolution in dynamic bandwidth allocation (DBA) dis-aggregation. We further enhance the scalability of the dis-aggregated PON-OLT by reducing the bandwidth overhead of REPORT/GATE, the time division multiplexing (TDM) control information transferred between dis-aggregated DBA and the hardware platform. Our experiment verifies that the data compression based on basic integer coding can improve the scalability by 27% with acceptable processing time.

Enhanced Dynamic Bandwidth Allocation Algorithm for Intelligent Home Networks

Internet of Things (IoT) has been seen playing a tremendous change in the Information Technology (IT) environments, and thus its importance has also been realized and played a vital role within Intelligent Home Networks (IHNs). This is because IoT establishes a connection between things and the Internet by utilizing different sensing devices to implement the intelligence to deal with the identification and management of the connected things. IHNs use intelligent systems to perform their daily operations. Meanwhile, these networks ensure comfort, safety, healthcare, automation, energy conservation, and remote management to devices and users. Apart from that, these networks provide assistance in self-healing for faults, power outages, reconfigurations, and more. However, we have realized that more and advanced devices and services continue to be introduced and used in these networks. This has led to competitions of the limited available network resources, services, and bandwidth. In this paper, therefore, we present the design and implementation of a Novel Dynamic Bandwidth Allocation (NoDBA) algorithm to solve the performance bottleneck incurred with IHNs. The proposed algorithm deals with the management of bandwidth and its allocation. In the proposed algorithm, this study integrates two algorithms, namely; Offline Cooperative Algorithm (OCA) and Particle Swarm Optimization (PSO) to improve Quality of Service (QoS). PSO defines the priority limits for subnets and nodes in the network. Meanwhile, OCA facilitates dynamic bandwidth allocation in the network. The Network Simulator-2 (NS-2) was used to simulate and evaluate the NoDBA and it showed improved results compared to the traditional bandwidth allocation algorithms. The obtained results show an average throughput of 92%, average delay of 0.8 seconds, and saves energy consumption of 95% compared to Dynamic QoS-aware Bandwidth Allocation (DQBA) and Data-Driven Allocation (DDA). Keywords: IHNs, Dynamic Bandwidth Allocation, PSO, OCA, QoS

A Bibliometric Analysis and Visualization of Passive Optical Network Research in the Last Decade

Passive optical networks (PONs) are the most successful broadband access solutions for its wide bandwidth, low-cost deployment worldwide, and maintenance. In the last decade, the diverse and prevalent research related to PONs sustains 330+ science citation index (SCI) articles per year, with 9.18 average citations per article, making it "big literature data". This article, for the first time, presents the graphical representation of knowledge base, knowledge domain, and knowledge evolution of PON research using co-citation analysis based on 3381 SCI publications worldwide from 2010 to 2019 in bibliometric visualization tool- CiteSpace. The bibliographic analysis focused on 528 important research articles, 34 hotspots, 155 research frontier keywords, 481 core authors, 185 cited authors, 58 countries, 254 institutions, and 118 journals. Based on the frequency of keywords, Modulation, WDM PON, Transmission, Semiconductor optical amplifier, and Dynamic bandwidth allocation are the dominant and turning points in the structural basis of PON research. The most prominent research hotspots appearing in recent years are Energy efficiency, NG-EPON, and Chaotic encryption. The trend of future research on PONs is from TWDM PON issues, fronthaul implementation in NG-PON2, 5G-PON and encryption methods for physical layer security. The leading core authors, institutes, countries, and journals were also identified related to the PON research. This analysis would serve as a comprehensive guide on the status and future research trends on PONs for the aspirant researchers.

Fiber-Wireless Testbed using Software Defined Radio for Protocol and Algorithm Testing

Fiber-Wireless (FiWi) network can provide abundant bandwidth capacity and mobility to the end-users. It also eliminates the need of having complete tedious end-to-end fiber installation from the central office to the users, which saves tremendous capital expenditure. However, FiWi is still progressing. Researchers worldwide are still developing experimental works for improvement on the network reliability, quality-of-services and security. Almost all recently proposed testbed designed for FiWi are using hardware that lacks in programmability feature, making it challenging to implement any protocols and algorithms. A testbed must be flexible, scalable and reprogrammable so that various experiments and testing can be implemented easily for testing purposes. In this paper, a reprogrammable FiWi testbed using software-defined radio (SDR) is proposed. One of the most prominent SDR available in the market is Universal Software Radio Peripheral (USRP). It is chosen to be used in this paper as it is equipped with a user-friendly programming platform; LabVIEW. To test the testbed’s reprogrammability feature, two algorithms are implemented for proof-of-concept; collision avoidance and dynamic bandwidth allocation. The collision avoidance algorithm is implemented in the wireless side of the testbed using the concept of Carrier Sensing Multiple Access/Collision Avoidance. At the fiber domain, a dynamic bandwidth allocation-limited scheduling is incorporated in the testbed. The results show that algorithms implemented in the testbed are in-line with the expected results. It proves that the testbed can be used for future algorithm testing for research purposes.

All-optical virtual private network utilizing bit allocation in OFDM-PON system

We propose a new all-optical virtual private network (AOVPN) using bit allocation technique in an orthogonal frequency-division multiplexing-based passive optical network (OFDM-PON) for the first time. The proposed system makes full use of the fundamental advantages of OFDM such as high spectral efficiency, dispersion tolerance, and flexibility in bandwidth allocation. It enables the simultaneous transmission of upstream (U/S) data and VPN data by multiple users. Since the U/S signal and the VPN signal are separated into bits within a high-order symbol, flexible resource management is possible according to real-time requirements. The feasibility of the proposed system is demonstrated in a 20-km single-fiber loopback link as a proof of concept. The experiment is focused on the optical transmission of VPN and U/S data. The results include performance variations of VPN transmission according to the bit allocation in a full-band uplink without multiple access and verification of dynamic bandwidth allocation (DBA) considering the multiple access by multi-users

Compact Frequency and Bandwidth Reconfigurable Microwave Filter

This paper details the construction and working of a compound reconfigurable filter capable of frequency and bandwidth reconfiguration in the frequency range from 2 to 3 GHz. The switching between frequency and bandwidth reconfiguration is inherited by two PIN diodes. Bandwidth tuning is facilitated by tuning two transmission zeros individually using varactor diodes, giving flexibility in reconfiguring the upper and lower pass edges. The two transmission zeros are obtained using simple concentric square loop resonators. The maximum bandwidth obtained is 1.5 times the minimum bandwidth offered by the filter. Hence the filter can be used for dynamic bandwidth allocation. This prototype is fabricated and validated in real-time. The simulated and measured results are analogous to each other.

Bandwidth Utilization and Management Algorithms (BUMAs) for NG-EPON

An upcoming capable technology named as Next generation Ethernet passive optical network (NG-EPON); to handle enormous bandwidth and effective distribution requirements for prospective Internet based services. Standardization procedures for NG-EPON are in progress. Transmission of data traffic from Optical Network Units (ONUs) towards Optical Line Terminal (OLT) in NG-EPON is carried out through four wavelengths each operating at a data rate of 25Gbps. Dynamic wavelength and bandwidth allocation (DWBA) algorithms are needed for the efficient arbitration of the bandwidth resources between the subscribers. In this paper, we have proposed two different DWBA algorithms for NG-EPON; DWBA-I and DWBA-II. DWBA-I is a Quality-of-Service based algorithm that has been designed to manage the different types of data traffic in NG-EPON. DWBA-I performs better than the different existing DBA algorithms in terms of packet delay, completion time and packet drop ratio. DWBA-II is an algorithm designed to meet the system requirement specifications of NG-EPON and handle traffic requirements of subscribers in a cost effective manner. DWBA-II algorithm is comparatively analyzed and evaluated with First-Fit-DWBA and Modified-IPACT algorithms. Performance of DWBA-II is better on the basis of average delay, grant utilization, completion time and packets drop ratio. Performance of DWBA-I and DWBA-II has been evaluated through simulation results.

A QoE-Aware Energy Supply Scheme over a FiWi Access Network in the 5G Era.

Integrated fiber-wireless (FiWi) should be regarded as a promising access network architecture in future 5G networks, and beyond; this due to its seamless combination of flexibility, ubiquity, mobility of the wireless mesh network (WMN) frontend with a large capacity, high bandwidth, strong robustness in time, and a wavelength-division multiplexed passive optical network (TWDM-PON) backhaul. However, the key issue in both traditional human-to-human (H2H) traffic and emerging Tactile Internet is the energy conservation network operation. Therefore, a power-saving method should be instrumental in the wireless retransmission-enabled architecture design. Toward this end, this paper firstly proposes a novel energy-supply paradigm of the FiWi converged network infrastructure, i.e., the emerging power over fiber (PoF) technology instead of an external power supply. Then, the existing time-division multiplexing access (TDMA) scheme and PoF technology are leveraged to carry out joint dynamic bandwidth allocation (DBA) and provide enough power for the sleep schedule in each integrated optical network unit mesh portal point (ONU-MPP) branch. Additionally, the correlation between the transmitted optical power of the optical line terminal (OLT) and the quality of experience (QoE) guarantee caused by multiple hops in the wireless frontend is taken into consideration in detail. The research results prove that the envisioned paradigm can significantly reduce the energy consumption of the whole FiWi system while satisfying the average delay constraints, thus providing enough survivability for multimode optical fiber.

Small Cell Planning: Resource Management and Interference Mitigation Mechanisms in LTE HetNets

In an attempt to address the huge data demand of indoor mobile users and poor signal strength from outdoor base stations to indoor environments, opertaors have started deploying variety of small cells likes Femtos, picos and micro cells. In this work, we used Femtos as small cells. Femto cells are low-cost, low-power consuming cellular base stations which operate only in licensed spectrum and are designed for both outdoor and indoor communication. Although small cells can be used for enhancing network capacity and coverage, arbitrary deployment of large number of small cells can lead to increase in operators expenditure and may create severe interference issues for cell edge users. In this paper, we look into optimal placement of small cell solutions to improve data rates of users in LTE HetNets using Femtos. Besides, these solutions address the main concerns of interference and resource management by proposing mechanisms for optimal placement of Femtos (OptFP, MinNF), dynamic power control and bandwidth allocation in Femtos (SOPC) and dynamic offloading. We provide a comparison of placement solutions and the applicability of each proposal keeping the operators’ revenue in mind.

BwSlicer: A bandwidth slicing framework for cloud data centers

Bandwidth allocation is an important and influential factor in enhancing the performance of the data centers’ nodes. In this paper we propose bwSlicer, a framework for bandwidth slicing in cloud data centers, that sheds light on the virtues of effective dynamic bandwidth allocation on improving the system performance and energy efficiency. Three algorithms are investigated to deal with this issue. In the first algorithm, called Fair Bandwidth Reallocation (FBR), two virtual machines co-hosted on the same node conditionally exchange bandwidth slices based on their requirements. The second algorithm, called Required Bandwidth Allocation (RBA), periodically monitors the co-hosted virtual machines and adds/removes bandwidth slices for each of them based on their bandwidth utilization. The third algorithm, called Divide Bandwidth Reallocation (DBR), divides the bandwidth of the virtual machine into slices once it finishes its execution, and distributes the slices among the co-hosted running virtual machines according to a specific policy. The proposed bandwidth slicing algorithms are emulated in a virtualized networking environment using the Mininet network emulator. The emulation results demonstrated a promising improvement ratio in execution time and energy consumption reaching up to 30%. These results present a call for action for further research into bandwidth slicing and reallocation as a viable complement to other energy-saving techniques for enhancing the energy consumption in cloud data centers.

ROTOS: A Reconfigurable and Cost-Effective Architecture for High-Performance Optical Data Center Networks

Fast and high capacity optical switching techniques have the potential to enable low latency and high throughput optical data center networks (DCNs) to afford the rapid increasing traffic boosted by multiple applications. Flexibility of the DCN is of key importance to provide adaptive and dynamic bandwidth and capacity to handle the variable traffic patterns of heterogeneous applications. Aiming at improving the network performance and the system flexibility of optical DCNs, we propose and investigate a novel optical DCN architecture named ROTOS based on reconfigurable optical top of rack (ToR) and fast optical switches. In the proposed DCN architecture, the novel optical flexible ToRs employing multiple transceivers (TRXs) and a wavelength selective switch (WSS) are reconfigured by the software-defined networking (SDN) control plane. The bandwidth can be dynamically allocated to the dedicated optical links on-demand according to the desired oversubscription (OV) and intra-/inter-cluster traffic matrix. Numerical investigations of the novel architecture under realistic traffic model indicate that dynamically allocating the TRXs and elastically controlling the WSS, a packet loss below 1E-5 and a server-to-server latency lower than 3 μs can be guaranteed for different traffic patterns at load of 0.4. With respect to the DCN with static interconnections, the average packet loss of ROTOS decreases two orders of magnitude and the average server-to-server latency performance improves by 21.5%. Scalability investigation to a large number of servers shows limited (11%) performance degradation as the network scale from 2560 to 40960 servers. Additionally, the dynamic bandwidth allocation of the DCN is experimentally validated. Network performance results show a packet loss of 0.05 and 5.85 μs end-to-end latency at the load of 0.8. Finally, investigations on the cost and power consumption confirm that the ROTOS DCN architecture has 28.4% lower cost and 35.0% better improvement for power efficiency with respect to the electrical switch based DCNs.

FlowTrace: Maximizing the Service Payoff of Heterogeneous Communications Networks

Massive Internet of Things devices with rapidly growing traffic make heterogeneous communications network a commercial infrastructure. As a critical network metric that represents the value created by the network, the service payoff of the network is positively correlated with the network throughput only when the quality of service (QoS) delivered to data-flows is guaranteed. To maximize the service payoff of the network, we propose a so-called FlowTrace as an QoS guarantee scheme. We first model the upper bound of the service payoff in a network as the maximum posterior service payoff (MPSP). To facilitate online bandwidth allocation and approximate the upper bound of the service payoff, we formulate a dynamic weight-based online bandwidth allocation (DWOBA) problem by designing dynamic weight as the flow metric. To relieve the computational pressure in the centralized network controller and the following bottleneck effect, we then decompose the DWOBA problem to a two-step DWOBA problem. Moreover, we propose a semi-centralized bandwidth allocation algorithm based on parallel edge computing for the two-step DWOBA problem. Simulation results demonstrate that the proposed FlowTrace outperforms prevalent online methods with a 3 $\sim$ 10x service payoff improvement under heavy traffic load and the optimization result reaches 70 $\%$ of the MPSP.

General QoS-aware scheduling procedure for passive optical networks

Increasing volume, dynamism, and diversity of access traffic have complicated the challenging problem of dynamic resource allocation in passive optical networks. We introduce a general scheduling procedure for passive optical networks, which optimizes a desired performance metric for an arbitrary set of operational constraints. The proposed scheduling has a fast and causal iterative implementation, where each iteration involves a local optimization problem followed by a recursive update of some status information. The generality of the platform enables a proper description of the diverse quality of service requirements, while its low computational complexity makes agile tracking of the network dynamism possible. To demonstrate its versatility and generality, the applications of the scheme for service-differentiated dynamic bandwidth allocation in time- and wavelength-division-multiplexed passive optical networks are discussed. To further reduce the computational complexity, a closed-form solution of the involved optimization in each iteration of the scheduling is derived. We directly incorporate transmission delay in the scheduling and show how the consumed power is traded for the tolerable amount of transmission delay. Furthermore, a 50% power efficiency improvement is reported by exploiting the inherent service diversity among subscribers. The impact of service prioritization, finite buffer length, and packet drops on the power efficiency of the scheme are also investigated.

DEC TDMA: A delay controlled and energy efficient clustered TDMA mechanism for FiWi access network

Integration of passive optical network and wireless access network is referred as FiWi access network. It provides larger bandwidth and flexibility for end users to improve last mile performance. With rapid growth in internet traffic and expansion in ICT, reduction in delay and efficient energy utilization have become major concerns in FiWi network. There has been lot of work done on delay and energy efficiency separately for optical back end and wireless front end to provide desired QoS. But this paper presents a cooperative mechanism namely DEC TDMA which is based on the concept of clustered TDMA. In DEC TDMA cooperation between front end and back end is achieved by providing synchronization using time stamping. Analytical expressions are also derived for energy efficiency, cycle time and delay for proposed and existing methods. Results found that proposed mechanism reduces delay very significantly specially at heavy load. It also provides better energy efficiency for large number of ONUs. Moreover DEC TDMA exhibits better delay performance due to implementation of dynamic bandwidth allocation while providing data slots in TDMA. Therefore proposed mechanism offers delay and energy efficient solution for large sized FiWi access network demanding high bandwidth applications in heavy load conditions.

A Cost Effective Architecture and Throughput Efficient Dynamic Bandwidth Allocation Protocol for Fog Computing Over EPON

Edge and fog networks call for high speed and energy-efficient networking platforms. Passive Optical Networks (PON) have been identified as a promising candidate for such platforms. In conventional dynamic bandwidth allocation (DBA) schemes for PON, the transmitter of an optical network unit (ONU) remains idle until the completion of upstream transmission of all other ONUs which indicates poor utilization of the transmitter. Scheduling fog services during these free intervals of time would lead to improved transmitter utilization. In this work, we exploit this opportunity and present a DBA protocol and its architecture, for enabling fog services. Our architecture employs a single line card at the ONUs providing considerable savings in terms of cost and energy. The results show that the cost per ONU could be reduced by 87$ and power savings of 5W per ONU could be achieved compared to proposals using separate transceivers at the ONU. The proposed DBA improves the load tolerance by ~300% towards maintaining delay bound of 10 ms and an improvement in throughput by ~20% at high loads when compared to existing philosophies, which is highly desirable for a delay-sensitive fog network. Placing fog node at remote node allows the fog node to operate at higher loads while maintaining delay bound when compared to its placement at OLT.