Bandwidth Cost Research Articles

TVG-Streaming: Learning User Behaviors for QoE-Optimized 360-Degree Video Streaming

360-degree video streaming shows great potential to revolutionize the streaming market, by providing much better immersive experience than standard video streams. However, its wide adoption is hindered by the surging demand of network bandwidth due to multi-screen video transmission. To reduce the bandwidth cost, one promising approach is to predict a user’s field of view (FoV), and then prefetch video tiles that a user will view a few seconds ahead. The challenge lies in that user behaviors cannot be properly captured with very limited information, especially the viewing time spent on each tile and the FoV switching behavior are hard to predict. In this paper, we propose a novel 360-degree video streaming algorithm called <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TVG-Streaming</i> to optimize user experiences by learning user view behaviors. Different from previous approaches, our idea is to exploit tile-view graphs (TVGs) generated by real user behaviors and accurately estimate the probability that each tile falls in the FoV. With the tile view probability, we can determine the bitrate of each tile for delivery and buffering with limited bandwidth budget so as to maximize users’ quality of experience (QoE). For evaluation, we conduct extensive experiments using real traces and the results show that our proposed <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">TVG-Streaming</i> algorithm significantly outperforms other algorithms by at least 20% improvement in terms of users’ QoE.

The Error Performance Evaluation of Pulse Position Modulation with Hadamard Code in High Order Modulation of Optical Wireless System

This paper investigates the error probability of our proposed modulation technique which is produced by combining the existing pulse position modulation (PPM) technique with Hadamard code (HC). Moreover, the HC is also combined with differential PPM (DPPM). The error performances of existing technique and the proposed technique are compared at high-order modulation (HOM) and low-order modulation. The results show that the error performance of the proposed technique outperforms at HOM at the cost of higher bandwidth. This error performance is improved due to the bipolar nature of HC and it reduces the power requirements of the system.

OFaaS: OpenFlow Switch as a Service for Multi Tenant Slicing in SD-CDN

In Software Defined-Content Delivery Networks (SD-CDN), the policies of tenants such as Youtube, Netflix, Office 365, etc. are not the same due to having different 5G traffic requirements for such contents as enhanced Mobile Broadband (eMBB), ultra-reliable low-latency (URLLC) and massive machine-type communication (mMTC). This leads SD-CDN multi-tenant slicing to provide different services with limited network resources, where each tenant can functionally manage their own virtual slice of a physical component according to service level agreements (SLAs). However, they are not permitted to dynamically configure their own components. Therefore, the physical end-to-end configuration of all edge devices causes extra hardware and bandwidth costs. Although software as a service (SaaS) is more preferred to handle cost-efficiency on a switch configuration that increases forwarding throughput (Mbps) with a less number of physical components in SD-CDN, the edge devices can be only served as infrastructure as a service (IaaS) currently. This motivation leads us to serve the switch as a service that includes both IaaS and SaaS characteristics. Therefore, we propose an OpenFlow as a service (OFaaS) design where each tenant has flow management and switch configuration permissions on their own virtual slice. In flow management, we define a novel Service Oriented Architecture (SOA) to orchestrate OFaaS driven topology by isolating each tenant from physical complexity. Here, each tenant can dynamically alter QoS on a flow and load balance between a sub-set of contents via OpenFlow protocol. In switch configuration; a novel OFaaS Management Algorithm for a multi-tenant slicing increases the number of tenants served per OpenFlow switch thanks to OFaaS design. It enables an end-to-end configuration via the NETCONF protocol with a novel YANG model of OFaaS. According to performance evaluation; OFaaS has the same forwarding throughput as conventional IaaS based OpenFlow switch for a homogenous content, whereas it has 71% more forwarding throughput (Mbps) and it has 40% more cost-efficient than conventional one for a heterogeneous content with $17 savings per tenant.

Computational Overhead of a Novel Scheme for Identity Confidentiality on IoT Devices in 5G Mobile Networks

Mobile networks are becoming a preferred choice for the Internet of Things (IoT), due to its flexibility, broad coverage, increasing bandwidth, low latency and low subscription cost. However, a long-standing security issue in any mobile network across the various generations is identity confidentiality. In a recent technical specification standardised by 3rd Generation Partnership Project (3GPP) for 5G mobile network, a novel scheme called the Elliptic Curve Integrated Encryption Scheme (ECIES) is adopted to tackle the issue of identity confidentiality. While this mechanism seems to provide a reasonable solution for modern resource affluent smart phones, it’s suitability for resource constrained IoT devices needs to be analysed. In this paper, we study the computational overhead of the ECIES on IoT devices.

Highly efficient key agreement for remote patient monitoring in MEC-enabled 5G networks

Remote patient monitoring is one of the cornerstones to enable Ambient Assisted Living. Here, a set of devices provide their corresponding input, which should be carefully aggregated and analysed to derive health-related conclusions. In the new Fifth-Generation (5G) networks, Internet of Things (IoT) devices communicate directly to the mobile network without any need of proxy devices. Moreover, 5G networks consist of Multi-access Edge Computing (MEC) nodes, which are taking the role of a mini-cloud, able to provide sufficient computation and storage capacity at the edge of the network. MEC IoT integration in 5G offers a lot of benefits such as high availability, high scalability, low backhaul bandwidth costs, low latency, local awareness and additional security and privacy. In this paper, we first detail the procedure on how to establish such remote monitoring in 5G networks. Next, we focus on the key agreement between IoT, MEC and registration center in order to guarantee mutual authentication, anonymity, and unlinkability properties. Taking into account the high heterogeneity of IoT devices that can contribute to an accurate image of the health status of a patient, it is of utmost importance to design a very lightweight scheme that allows even the smallest devices to participate. The proposed protocol is symmetric key based and thus highly efficient. Moreover, it is shown that the required security features are established and protection against the most of the well-known attacks is guaranteed.

AI-Chain: Blockchain Energized Edge Intelligence for Beyond 5G Networks

Beyond fifth generation (B5G) networks have recently emerged as the advancement of existing communication systems. With the unprecedented proliferation of artificial intelligence (Ai), more intelligent B5G networks are projected to fuel the continuous development of Ai applications and the efficiency of communication technologies. More recently, the outgrowths of B5G have brought billions of devices throwing zillions of bytes of data to network edges. Training such huge data volumes using a centralized data center proves to be a challenging task, due to prohibitively heavy bandwidth costs, poor time efficiency, and high privacy leakages. These challenges have fueled the revolutionary shift of intelligence applications from a centralized data center to ubiquitous edges, denoted as edge intelligence. However, there is still a long way ahead before edge intelligence is able to fully mature. Governing and sharing the use of learning results efficiently, reliably, and safely are hampered by the heterogeneity and non-confidence among edges. in this article, we propose blockchain energized edge intelligence for B5G networks, called the Ai-Chain, fusing deep learning and blockchain. The Ai-Chain is a distributed and immutable record of learning results that is able to construct a new basis of sharing among edges. Additionally, we consider a novel learning-based consensus protocol in the Ai-Chain, denoted as proof of learning (PoL). in response to the huge computing power waste, PoL treats the training process as a working puzzle, rather than the meaningless hashing in a proof of work protocol. On the other hand, PoL fully unlocks the potential of sharing more advanced intelligence among edges. in order to demonstrate the effectiveness of the Ai-Chain, we employ it to solve a joint resource allocation problem in B5G networks. Experimental results prove the superior performance of the Ai-Chain to the current popular scheme.

Revocation in Vehicular Public Key Infrastructures: Balancing privacy and efficiency

Modern vehicular security architectures provision authorized vehicles with multiple short-term pseudonym certificates, so users can avoid tracking by rotating among certificates when signing messages. The large number of certificates in the system, however, makes revocation via Certificate Revocation Lists (CRLs) complex and/or inefficient. Two possible solutions for this issue are: (1) frequently provision non-revoked vehicles with few certificates, as done in the Cooperative Intelligent Transport Systems (C-ITS) standard; or (2) provision many encrypted certificates and periodically broadcast “activation codes” for controlling decryption by non-revoked vehicles, as proposed in Activation Codes for Pseudonym Certificates (ACPC), an extension of the Security Credential Management System (SCMS) standard. In this article, we compare such approaches in terms of computational efficiency and privacy preservation. We show that ACPC has advantages over both C-ITS (e.g., support for distributed caching) and CRLs (e.g., avoiding processing overheads on vehicles). We also discuss how ACPC can benefit from the unicast distribution of activation codes, with different trade-offs between privacy and bandwidth savings: getting as much privacy as C-ITS, but downloading 16-byte codes instead of hundreds of kilobytes, or fixing bandwidth costs (resp. privacy level) for a privacy degradation (resp. bandwidth usage) that grows logarithmically with the number of revocations.

An efficient approach for privacy preserving decentralized deep learning models based on secure multi-party computation

This paper aims to develop a new efficient framework named Secure Decentralized Training Framework (SDTF) for Privacy Preserving Deep Learning models. The main feature of the proposed framework is its capable of working on a decentralized network setting that does not need a trusted third-party server while simultaneously ensuring the privacy of local data with a low cost of communication bandwidth. Particularly, we first propose a so-called Efficient Secure Sum Protocol (ESSP) that enables a large group of parties to jointly calculate a sum of private inputs. ESSP can work not only with integer number but also with floating point number without any data conversion. We then propose a Secure Model Sharing Protocol that enables a group of parties securely train and share the local models to be aggregated into a global model. Secure Model Sharing Protocol exploits randomization techniques and ESSP to protect local models from any honest-but-curious party even n-2 of n parties colluding. Eventually, these protocols are employed for collaborative training decentralized deep learning models.We conduct theoretical evaluation of privacy and communication cost as well as empirical experiments on balance class image datasets (MNIST) and an unbalance class text dataset (UCI SMS Spam). These experiments demonstrate the proposed approach can obtain high accuracy (i.e. 97% baseline accuracy in only 10 training rounds with MNIST, 100 training rounds with SMS Spam) and robust to the heterogeneity decentralized network, with non-IID and unbalance data distributions. We also show a reduction in required rounds of training to achieve the accuracy baseline by 5× as compared to Downpour SGD. It is shown that the proposed approach can achieve both the privacy at the level of cryptographic approaches and efficiency at the level of randomization techniques, while it also retains higher model’s utility than differential privacy approaches.

Non-Orthogonal Wireless Backhaul Design for Cell-Free Massive MIMO: An Integrated Computation and Communication Approach

In cell-free massive multiple-input-multiple-output system with wireless backhaul, the distributed access points (APs) and the center processing unit (CPU) are connected via wireless links. Hence, the limited backhaul bandwidth becomes a critical challenge to uplink transmission. To save the bandwidth while maintaining high transmission accuracy, we propose to deploy non-orthogonal transmissions in backhaul link and jointly optimize the detection computation mappings at the APs and the CPU under the non-orthogonal backhaul. First, we formulate the joint design problem subject to backhaul bandwidth constraint aiming at a better end-to-end transmission accuracy. Then, the non-trivial problem is parameterized and solved with a novel model-driven deep neural network, where wireless backhaul is integrated as a neural computing layer by exploiting the reciprocity between non-orthogonal transmission and additive operation. Evaluations show that, the proposed integration method outperforms the conventional approaches by a margin in both backhaul bandwidth cost and the symbol error rate.

Which is the Optimal Commercial Mode for a Video Site: Paid, Free, or Hybrid?

A monopolistic video site provides three commercial modes for viewers: paid, free, and hybrid. In paid mode, the video site provides high-definition video with no advertising (“paid video”), but viewers must pay for watching; in free mode, the video site provides free video with high or standard definition, and viewers have to watch advertisements; in hybrid mode, the video site provides both paid and free video. Our objective is to determine the optimal price, advertising amount, and free video definition to maximize the site’s profits according to viewer characteristics. In doing so, we take into account bandwidth costs, which are correlated with video definition. We study the optimal profit of the video site in different cases and find that in free mode, the video site should provide high-definition free video. Through numerical examples, we find that the system parameters determine whether the optimal profit can be obtained by providing paid, free, or hybrid mode.

Optimal averaging time for improving observer accuracy of stochastic dynamical systems

In the problem of remote estimation by a centralized observer, improvements to the accuracy of observer estimates come at a cost of higher communication bandwidth and energy consumption. In this article we improve observer estimation accuracy by reducing the measurement variance on the sensor node before its transmission to the centralized observer node. The main contribution is to show that measurement variance is a trade-off between dynamical system variance and sensor variance. As a result there is an optimal averaging time that minimizes measurement variance, providing more accurate measurement to the observer. The optimal averaging time is computable by solving a univariate optimization problem.

Optimal Control-Aware Transmission for Mission-Critical M2M Communications Under Bandwidth Cost Constraints

In this paper, we consider a mission-critical control system, where a dynamic plant is monitored by a mobile device (MD), and the monitored signal is transmitted to a remote controller via heterogeneous cellular and Wi-Fi networks. We propose an optimal control-aware machine-to-machine (M2M) transmission strategy for mission-critical control applications, in which control performance is measured by remote estimation error and system stability while limited by bandwidth cost. Specifically, the problem of minimizing estimation error, subject to the constraints of cellular usage costs and system stability, is formulated as an infinite-horizon constrained Markov decision process (CMDP), where the MD has options to transmit through Wi-Fi or cellular, or to stay idle. We solve the problem by utilizing the Lagrange multiplier approach, and prove that the optimal strategy is a randomized mixture of two threshold structure strategies. Furthermore, to estimate the structured optimal strategy, we present an algorithm called simultaneous perturbation stochastic approximation (SPSA), in which the complexity is $O(|\mathcal {A}|)$ lower than a non-structured one with $|\mathcal {A}|$ being the number of the actions.

Identifier and Locator Separation Based Site Multi-homing Path Failure Recovery

Multi-homing is one of the effective ways to defeat path failure and increase the reliability of site network service. However, limited by the TCP/IP architecture, multi-homing has not been well deployment. In this paper, we proposed a sit multi-homing path failure recovery method based on “Locator/Identifier Split” by adding mapping service and RTT cooperative detection mechanism to the edge router. In order to reduce the probing’s cost of the router’s resource and link bandwidth, we also optimized the probing algorithm by reducing unnecessary active probing. The results of simulation test proved that this method could effectively detect the path failure and performance drop including packets delay and packets loss, and made a fast path switch to ensure the normal running of the upper application services (e.g., FTP). We also gave a theory analysis of the overhead brought by the sit multi-homing path failure recovery method. It shows that the addition bandwidth cost is 0.036% of the simulation link bandwidth and our approach is practicable.

Machine Learning Algorithm Based Bandwidth Prediction for Internet Usage

The applicability of framework structure and affiliation arranging recognize a basic activity in the bandwidth prediction. The procedure for predicting the framework use is to see the basic transmission limit with respect to future periods. This prediction helps with utilizing the techniques workplaces in the saint way. Thinking about the fundamental cost of bandwidth, at top hours of a framework traffic we can follow an amazing sort of plan to purchase. In this paper, the past use data of FWDR organize centers is at risk to univariate direct time plan ARIMA model after precise change is used to calculate necessary bandwidth limit concerning future needs. The anticipated data is veered from the obvious data gained from a for all intents and purposes indistinguishable framework and the foreseen data has been viewed as inside ten percent MAPE. This design reduction the MAPE by eleven point seventy-one percentage and fifteen point forty-two percent of self-rulingly when stood separated from the non-able changed ARIMA model at ninety-nine percent CI. The outcome show that the suitably changed ARIMA design has improved show when meandered from non-intentionally changed ARIMA model. Increasingly significant dataset can be passed on with season alterations and thought of expanded length groupings, for dynamically unequivocal and longer term needs.

FANET optimization: a destination path flow model

Closed-loop routing in flying ad hoc networks (FANET) arises as a result of the quick changes of communication links and topology. As such, causing link breakage during information dissemination. This paper proposed a destination path flow model to improve the communication link in FANET. The models utilized Smell Agent Optimization and Particle Swarm Optimization algorithms in managing link establishment between communicating nodes. The modeled scenario depicts the practical application of FANET in media and sports coverage where only one vendor is given the license for live coverage and must relay to other vendors. Three different scenarios using both optimization Algorithms were presented. From the result obtained, the SAO optimizes the bandwidth costs much better than PSO with a percentage improvement of 10.46%, 4.04% and 3.66% with respect to the 1st, 2nd and 3rd scenarios respectively. In the case of communication delay between the FANET nodes, the PSO has a much better communication delay over SAO with percentage improvement of 40.89%, 50.26% and 68.85% in the first, second and third scenarios respectively.

SDN-Enabled Energy-Efficient Routing Optimization Framework for Industrial Internet of Things

The traditional Internet architecture relies on the best-effort principle, which is not suitable for critical industrial Internet of Things (IIoT) applications such as healthcare systems with stringent quality-of-service (QoS) requirements. In this article, a software-defined network (SDN) based on an analytical parallel routing framework is proposed by using the massive processing power of a graphics processing unit (GPU) for dynamically optimizing multiconstrained QoS parameters in the IIoT. The framework considers three types of QoS applications for smart healthcare traffic: loss-sensitive, delay-sensitive, and jitter-sensitive. A QoS-enabled routing optimization problem is formulated as a max-flow min-cost problem, while a greedy heuristic that dispatches the path calculation task concurrently to the GPU for calculating optimal forwarding paths considering the QoS requirement of each flow is proposed. The results show that the proposed scheme efficiently utilizes the limited bandwidth cost in terms of energy and bandwidth while satisfying the QoS requirement of each flow with maximizing the network resources for future IIoT traffic flows. Comparative analysis of simulation results with shortest path delay, Lagrangian relaxation-based aggregated cost, and Sway schemes indicate a reduced violation in the service-level agreement by 17%, 19%, and 4%, respectively, by using the AttMpls topology, while it is 48%, 44%, and 7% when the Goodnet topology is used. Moreover, SEQOS is seen to be energy efficient and eight times faster than the benchmark algorithms in large IIoT networks.

Traffic Differentiated Clustering Routing in DSRC and C-V2X Hybrid Vehicular Networks

Vehicles equipped with sensors can participate in mobile crowdsourcing applications. Vehicular Ad Hoc Networks (VANETs) based on Dedicated Short Range Communication (DSRC) are used to carry sensing data. However, multi-hop transmissions for gathering data to Road Side Units (RSUs) in VANETs suffer from low data rate and long end-to-end delay, which can hardly meet the QoS requirements of delay-sensitive services. This triggers the consideration of constituting a DSRC and Cellular-Vehicle-to-Everything (C-V2X) hybrid vehicular network. Nevertheless, using cellular links to carry traffic can cause high cellular bandwidth costs. In this paper, we propose a Traffic Differentiated Clustering Routing (TDCR) mechanism in a Software Defined Network (SDN)-enabled hybrid vehicular network. The proposed mechanism includes a centralized one-hop clustering approach and a data delivery optimization method. Particularly, the optimization is to make a tradeoff between cellular bandwidth cost and end-to-end delay, for Cluster Heads (CHs) delivering their aggregated data either by multi-hop Vehicle-to-Vehicle (V2V) transmissions or by cellular networks. Since the problem is proven to be NP-hard, a two-stage heuristic algorithm is designed. We carry out simulations to evaluate the performance of our data collection scheme and the results show that it performs better than traditional mechanisms.

AyoPeer: The adapted ayo-game for minimizing free riding in peer-assisted network

Peer-assisted networks have recently emerged as an excellent tool for distributing video contents across the Internet. High peers’ participation and cooperation in peer-assisted networks can drastically reduce the bandwidth cost, increased scalability and ease of deployment. However, the main factor which limits effective distribution of resources in this network is free riding; a situation in which a peer consumes resources freely but refuses to service other peers. This paper presents a new peer-assisted streaming model (AyoPeer) using game theoretics with the aim of minimizing free riding.The interactions between peers are fashioned after a famous Yoruba game called “Ayo”. AyoPeer provides a reward and punishment mechanism that ensures the free-riders fail and cooperating peers succeeds with download. The experimental results revealed that AyoPeer mechanism outperform the existing scheme by ensuring fair distribution of resources, high cooperation among peers and resistance to free-riding peers.

Empirical Feasibility Analysis for Energy Harvesting Intravehicular Wireless Sensor Networks

Vehicle systems currently utilize wired networks for power delivery and communication between nodes. Wired networks cannot practically accommodate nodes in moving parts and with the increasing functional complexity in vehicles, they require kilometer-long harnesses, significantly increasing fuel consumption, manufacturing, and design costs. Alternatively, energy harvesting intravehicular wireless sensor networks (IVWSNs) can accommodate nodes in all locations and they obviate the need for wiring, significantly lowering costs. This article empirically analyzes the feasibility of such an IVWSN framework via extensive in-vehicle measurements for communications at 2.4 GHz, ultrawideband (UWB), and millimeter-wave (mmWave) together with radio frequency (RF), thermal, and vibration energy harvesting. Our analyses indicate mmWave performs best for short Line-of-Sight (LoS) links in the engine compartment with performance close to UWB for LoS links in the chassis and passenger compartments considering worst case signal-to-interference-and-noise ratio (SINR). For non-LoS links, which appear mostly in the engine compartment and chassis, UWB provides the highest security and reliability. 2.4 GHz suffers heavily from interference in all compartments while UWB utilizes narrowband suppression techniques at the cost of lower bandwidth; mmWave inherently experiences very low interference due to its propagation characteristics. On the other hand, RF energy harvesting provides up to 1 mW of power in all compartments. Vibration and thermal energy harvesters can supply nodes consuming <; 10 mW in the engine compartment and <; 5 mW nodes in the chassis. In the passenger compartment, thermal harvesting is not available due to low temperature gradients, but vibration and RF sources can supply <; 1 mW nodes.

NewMCOS: Towards a Practical Multi-Cloud Oblivious Storage Scheme

Encryption alone is not enough to protect data privacy, because access pattern leaks some sensitive information. Oblivious RAM (ORAM), the solution to this problem, is still far from practical deployment for heavy storage and communication/ computation overhead. To reduce them, an insightful idea was proposed to utilize non-colluding clouds to shift client computation and client-cloud communication to the clouds. The proposed multi-cloud ORAM achieved $O$ O (1) client-cloud bandwidth cost and removed most of client computation. In this paper, we exploit “ disconnected ORAM operation ” and design “ two-layer encryption ” to further reduce these overheads. Experiments show that our proposed scheme, NewMCOS, significantly reduces evict cache size from GB/MB to KB level with about 2-3 times lower response time and 20 percent savings in bandwidth for clouds, compared to other schemes. Theoretically speaking, we reduce evict cache size from $O(\sqrt{N})$ O ( N ) to $O(ZK)$ O ( Z K ) , where $N$ N is the number of real data blocks, $K$ K is the number of clouds ( $2 2 K N ), and $Z$ Z is the number of real blocks uploaded from the client for eviction. By employing “ lazy eviction operation ”, the write frequency is reduced by $O(Z)$ O ( Z ) , the shuffling bandwidth cost is reduced by $\Omega (Z\; \log Z)$ Ω ( Z log Z ) . Meanwhile, NewMCOS is proved to be secure.