Real Test Bed Research Articles

Servicing delay sensitive pervasive communication through adaptable width channelization for supporting mobile edge computing

Over the last fifteen years, wireless local area networks (WLANs) have been populated on large variety of pervasive devices hosting heterogeneous applications. Pervasive edge computing has accelerated more distributed network applications for these devices, eliminating the round-trip to help in achieving zero latency dream. However, These applications require significantly variable data rates for effective functioning, especially in pervasive computing. The static bandwidth of frequency channelization in current WLANs strictly restricts the maximum achievable data rate by a network station. This static behavior spawns two major drawbacks: under-utilization of scarce spectrum resources and less support to delay sensitive applications such as voice and video.To this point, if the computing is moved to the edge of WLANs to reduce the frequency of communications, the pervasive devices can be provided with better services during the communication and networking. Thus, we aim to distribute spectrum resources among pervasive devices based upon delay sensitivity of applications while simultaneously maintain the fair channel access semantics of medium access control (MAC) layer of WLANs. Henceforth, ultra-low latency, efficiency and reliability of spectrum resources can be assured.In this paper, two novel algorithms have been proposed for adaptive channelization to offer rational distribution of spectrum resources among pervasive edge nodes based on their bandwidth requirement and assorted ambient conditions. The proposed algorithms have been implemented on a real test bed of commercially available universal software radio peripheral (USRP) devices. Thorough investigations have been carried out to enumerate the effect of dynamic bandwidth channelization on parameters such as medium utilization, achievable throughput, service delay, channel access fairness and bit error rate. The achieved empirical results demonstrate that we can optimally enhance the network wide throughput by almost 30% by using channels of adaptable bandwidths.

Anomaly detection in substation networks

Fundamental components of the distribution systems of electric energy are primary and secondary substation networks. Considering the incorporation of legacy communication infrastructure in these systems, they often have in- herent cybersecurity vulnerabilities. Moreover, traditional intrusion defence strategies for IT systems are often not applicable. With the aim to improve cybersecurity in substation networks, in this paper we present two methods for monitoring SCADA system: the first one exploiting neural networks, while the second one is based on formal methods. To evaluate the effective- ness of the proposed methods, we conducted experiments on a real test bed representing the substation domain as close to real-world as possible. From this test bed we collect data during normal operation and during situations where the system is under attack. To this end several different types of attack are conducted. The data collected is used to test two versions of the mon- itoring system: one based on machine learning with a neural network and one using a model-checking approach. Moreover, the two proposed models are tested with new data to evaluate their performance. The experiments demonstrate that both methods obtain an accuracy greater than 90%. In particular, the methodology based on formal methods achieves better per- formance if compared to the one based on neural networks.

Priority-Based Bandwidth Management in Virtualized Software-Defined Networks

In Industrial Internet of Things (IoT) applications, when the network size increases and different types of flows share the bandwidth, the demand for flexible and efficient management of the communication network is compelling. In these scenarios, under varying workload and flow priorities, the combined use of Software-Defined Networking (SDN) and Network Virtualization (NV) is a promising solution, as such techniques allow to reduce the network management complexity. This work presents the PrioSDN Resource Manager (PrioSDN_RM), a resource management mechanism based on admission control for virtualized SDN-based networks. The proposed combination imposes bounds on the resource utilization for the virtual slices, which therefore share the network links, while maintaining isolation from each other. The presented approach exploits a priority-based runtime bandwidth distribution mechanism to dynamically react to load changes (e.g., due to alarms). The paper describes the design of the approach and provides experimental results obtained on a real testbed.

Stability analysis and nonlinear current‐limiting control design for DC micro‐grids with CPLs

In this study, a DC micro-grid consisting of multiple paralleled energy resources interfaced by both bidirectional AC/DC and DC/DC boost converters and loaded by a constant power load (CPL) is investigated. By considering the generic dq transformation of the AC/DC converters' dynamics and the accurate nonlinear model of the DC/DC converters, two novel control schemes are presented for each converter-interfaced unit to guarantee load voltage regulation, power sharing and closed-loop system stability. This novel framework incorporates the widely adopted droop control and using input-to-state stability theory, it is proven that each converter guarantees a desired current limitation without the need for cascaded control and saturation blocks. Sufficient conditions to ensure closed-loop system stability are analytically obtained and tested for different operation scenarios. The system stability is further analysed from a graphical perspective, providing valuable insights of the CPL's influence onto the system performance and stability. The proposed control performance and the theoretical analysis are first validated by simulating a three-phase AC/DC converter in parallel with a bidirectional DC/DC boost converter feeding a CPL in comparison with the cascaded PI control technique. Finally, experimental results are also provided to demonstrate the effectiveness of the proposed control approach on a real testbed.

Low Latency Big Data Processing without Prior Information

Job scheduling plays an important role in improving the overall system performance in big data processing frameworks. Simple job scheduling policies, such as Fair and FIFO scheduling, do not consider job sizes and may degrade the performance when jobs of varying sizes arrive. More elaborate job scheduling policies make the convenient assumption that jobs are recurring, and complete information about their sizes is available from their prior runs. In this paper, we design and implement an efficient and practical job scheduler for big data processing systems to achieve better performance even without prior information about job sizes. The superior performance of our job scheduler originates from the design of multiple level priority queues, where jobs are demoted to lower priority queues if the amount of service consumed so far reaches a certain threshold. In this case, jobs in need of a small amount of service can finish in the topmost several levels of queues, while jobs that need a large amount of service to complete are moved to lower priority queues to avoid head-of-line blocking. Our new job scheduler can effectively mimic the shortest job first scheduling policy without knowing the job sizes in advance. To demonstrate its performance, we have implemented our new job scheduler in YARN, a popular resource manager used by Hadoop/Spark, and validated its performance with experiments in both real testbeds including Amazon EC2 and large-scale trace-driven simulations. Our experimental and simulation results have strongly confirmed the effectiveness of our design: our new job scheduler can reduce the average job response time of the Fair scheduler by up to 45 percent and achieve better fairness at the same time.

JPAS: Job-progress-aware flow scheduling for deep learning clusters

Deep learning (DL) is an increasingly important tool for large-scale data analytics and DL workloads are also common in today's production clusters due to the increasing number of deep-learning-driven services (e.g., online search and speech recognition). To handle ever-growing training datasets, it is common to conduct distributed DL (DDL) training to leverage multiple machines in parallel. Training DL models in parallel can incur significant bandwidth contention on shared clusters. As a result, the network is a well-known bottleneck for distributed training. Efficient network scheduling is essential for maximizing the performance of DL training. DL training is feedback-driven exploration (e.g., hyper-parameter tuning, model structure optimization), which requires multiple retrainings of deep learning models that differ in terms of their configuration. The information at the early stage of each retraining can facilitate the direct search for high-quality models. Thus, reducing the early-stage time can accelerate the exploration of DL training. In this paper, we propose JPAS, which is a flow scheduling system for DDL training jobs that aims at reducing the early-stage time. JPAS uses a simple greedy mechanism to periodically order all DDL jobs. Each host machine sets priorities for its flows using the corresponding job order and offloads the flow scheduling and rate allocation to the underlying priority-enabled network. We evaluate JPAS over a real testbed that is composed of 13 servers and a commodity switch. The evaluation results demonstrate that JPAS can reduce the time to reach 90% or 95% of the converged accuracy by up to 38%. Hence, JPAS can remarkably reduce the early-stage time and thus accelerate the search for high-quality models.

Data augmentation for self-paced motor imagery classification with C-LSTM

Objective. Brain–computer interfaces (BCI) are becoming important tools for assistive technology, particularly through the use of motor imagery (MI) for aiding task completion. However, most existing methods of MI classification have been applied in a trial-wise fashion, with window sizes of approximately 2 s or more. Application of this type of classifier could cause a delay when switching between MI events. Approach. In this study, state-of-the-art classification methods for motor imagery are assessed offline with considerations for real-time and self-paced control, and a convolutional long-short term memory (C-LSTM) network based on filter bank common spatial patterns (FBCSP) is proposed. In addition, the effects of several methods of data augmentation on different classifiers are explored. Main results. The results of this study show that the proposed network achieves adequate results in distinguishing between different control classes, but both considered deep learning models are still less reliable than a Riemannian MDM classifier. In addition, controlled skewing of the data and the explored data augmentation methods improved the average overall accuracy of the classifiers by 14.0% and 5.3%, respectively. Significance. This manuscript is among the first to attempt combining convolutional and recurrent neural network layers for the purpose of MI classification, and is also one of the first to provide an in-depth comparison of various data augmentation methods for MI classification. In addition, all of these methods are applied on smaller windows of data and with consideration to ambient data, which provides a more realistic test bed for real-time and self-paced control.

Comparison of Management and Orchestration Solutions for the 5G Era

5G is considered to be the technology that will accommodate the development and management of innovative services with stringent and diverse requirements from end users, calling for new business models from the industry. In this context, the development and efficient management of Network Services (NS) serving specific vertical industries and spanning across multiple administrative domains and heterogeneous infrastructures is challenging. The main challenges regard the efficient provision of NSs considering the Quality of Service (QoS) requirements per vertical industry along with the optimal usage of the allocated resources. Towards addressing these challenges, this paper details an innovative approach that we have developed for managing and orchestrating such NSs, called SONATA, and compare it with OSM and Cloudify, which are two of the most known open-source Management and Orchestration (MANO) frameworks. In addition to examining the supported orchestration mechanisms per MANO framework, an evaluation of main operational and functional KPIs is provided based on experimentation using a real testbed. The final aim is the identification of their strong and weak points, and the assessment of their suitability for serving diverse vertical industry needs, including of course the Internet of Things (IoT) service ecosystem.

Collaborative data acquisition and processing for post disaster management and surveillance related tasks using UAV-based IoT cloud

Rescue and recovery operations are very critical for post disaster management. For post disaster management and surveillance activities, there is a need of acquiring the information about the ground situation through sensing or observing data and identify the locations, which is a challenging task. These data may be sensed or observed through different types of sensors deployed in the area of interest. This can be achieved through a collaborative way of data acquisition and processing. In this context, this paper introduces the framework for data acquisition based on collaborative processing using unmanned aerial vehicles and internet of things network. The major contribution of this work is that the real time test bed using actual UAVs and IoT devices for data acquisition, clustering of different sensors and devices deployed for it and localising the sensors and different events or situations arising in the region of interest is detailed and demonstrated. The framework is evaluated using real test bed with drones (UAVs) and integrated with cloud platform in IoT-based environment for data storage of the acquired data for further analysis and effective decision making in disaster situations.

Analysis of Distributed and Autonomous Scheduling Functions for 6TiSCH Networks

The 6TiSCH architecture is expected to play a significant role to enable the Internet of Things paradigm also in industrial environments, where reliability and timeliness are of paramount importance to support critical applications. Many research activities have focused on the Scheduling Function (SF) used for managing the allocation of communication resources in order to guarantee the application requirements. Two different approaches have mainly attracted the interest of researchers, namely distributed and autonomous scheduling. Although many different (both distributed and autonomous) SFs have been proposed and analyzed, a direct comparison of these two approaches is still missing. In this work, we compare some different SFs, using different behaviors in allocating resources, and investigate the pros and cons of using distributed or autonomous scheduling in four different scenarios, by means of both simulations and measurements in a real testbed. Based on our results, we also provide a number of guidelines to select the most appropriate SF, and its configuration parameters, depending on the specific use case.

High Performance Fault Tolerant Resource Scheduling in Computational Grid Environment

Virtual resources team up to create a computational grid, which is used in computation-intensive problem solving. A majority of these problems require high performance resources to compute and generate results, making grid computation another type of high performance computing. The optimization in computational grids relates to resource utilization which in turn is achieved by the proper distribution of loads among participating resources. This research takes up an adaptive resource ranking approach, and improves the effectiveness of NDFS algorithm by scheduling jobs in those ranked resources, thereby increasing the number of job deadlines met and service quality agreements met. Moreover, resource failure is taken care of by introducing a partial backup approach. The benchmark codes of Fast Fourier Transform and Matrix Multiplication are executed in a real test bed of a computational grid, set up by Globus Toolkit 5.2 for the justification of propositions made in this article.

Energy-Efficient Real-Time Multi-Core Assignment Scheme for Asymmetric Multi-Core Mobile Devices

The big.LITTLE architecture has been extensively integrated into smart mobile devices for better performance and higher energy efficiency. However, the desired energy savings obtained by the big.LITTLE architecture is not sufficiently achieved because the LITTLE cores are not fully utilized while running real-time user applications. In this study, an energy efficient big.LITTLE core assignment algorithm is proposed to reduce the energy consumption of the mobile device by utilizing the LITTLE core as much as possible while guaranteeing the real-time performance of the mobile application. By applying the proposed multi-core assignment technique on a real test-bed of an off-the-shelf smartphone, we prove that the proposed technique can improve the energy saving effect while guaranteeing real-time performance. The energy efficiency of the proposed scheme is compared with that of the legacy scheduler in various environments. In addition, we propose a machine learning-based method to predict the expected processing time more accurately for a task before assigning to one of multi-cores. The presented prediction method is expected to reduce the chances of missing a deadline when employed on the proposed multi-core assignment scheme.

Indoor device-free passive localization with DCNN for location-based services

With the increasing demand of indoor location-based services, such as tracking targets in a smart building, device-free localization technique has attracted great attentions because it can locate the targets without employing any attached devices. Due to the limited space and complexity of the indoor environment, there still exist challenges in terms of high localization accuracy and high efficiency for indoor localization. In this paper, for addressing such issues, we first convert the received signal strength (RSS) signals into image pixels. The localization problem is then formulated as an image classification problem. To well handle the variant RSS images, a deep convolutional neural network is then structured for classification. Finally, for validating the proposed scheme, two real testbeds are built in the indoor environments, including a living room and a corridor of an apartment. Experimental results show that the proposed scheme achieves good localization performance. For example, the localization accuracy can reach up to 100% in the scenario of living room and 97.6% in the corridor. Moreover, the proposed approach outperforms the methods of the K-nearest-neighbor and the support vector machines in both the noiseless and noisy environments.

DA&FD–Deadline-Aware and Flow Duration-Based Rate Control for Mixed Flows in DCNs

Data center has become an important facility for hosting various applications. For data center networks, deadline missing rate and average flow completion time are two main metrics for the performance of applications. In this paper, we find deadline-aware methods can only reduce the percentage of flows missing deadline, while flowsize-aware and information-cumulative methods can only optimize the average flow completion time. However, traffic in data center is the mixture of various flows and focusing on the single goal is not enough. We advocate to incorporate deadline and flow duration time into flow rate control. Then we design DA&FD (Deadline-Aware and Flow Duration) based rate control mechanism and analyze its performance in theory. At last, we evaluate DA&FD under different topologies, real world traffic and load scenarios, both by simulation and in real testbed. Our results show that DA&FD performs close to D2TCP and about 15%, 25%, 30%, 35% better than Ameon, L2DCT, Karuna, DCTCP on deadline missing rate. For average FCT, the performance of DA&FD is similar to L2DCT and compared with Ameon, D2TCP, Karuna, DCTCP, DA&FD can reduce average FCT by 10%, 15%, 20%, 25%.

Energy Modeling of Neighbor Discovery in Bluetooth Low Energy Networks.

Given that current Internet of Things (IoT) applications employ many different sensors to provide information, a large number of the Bluetooth low energy (BLE) devices will be developed for IoT systems. Developing low-power and low-cost BLE advertisers is one of most challenging tasks for supporting the neighbor discovery process (NDP) of such a large number of BLE devices. Since the parameter setting is essential to achieve the required performance for the NDP, an energy model of neighbor discovery in BLE networks can provide beneficial guidance when determining some significant parameter metrics, such as the advertising interval, scan interval, and scan window. In this paper, we propose a new analytical model to characterize the energy consumption using all possible parameter settings during the NDP in BLE networks. In this model, the energy consumption is derived based on the Chinese remainder theorem (CRT) for an advertising event and a scanning event during the BLE NDP. In addition, a real testbed is set up to measure the energy consumption. The measurement and experimental results reveal the relationship between the average energy consumption and the key parameters. On the basis of this model, beneficial guidelines for BLE network configuration are presented to help choose the proper parameters to optimize the power consumption for a given IoT application.

ECSNeT++ : A simulator for distributed stream processing on edge and cloud environments

The objective of Internet of Things (IoT) is ubiquitous computing. As a result many computing enabled, connected devices are deployed in various environments, where these devices keep generating unbounded event streams related to the deployed environment. The common paradigm is to process these event streams at the cloud using the available Distributed Stream Processing (DSP) frameworks. However, with the emergence of Edge Computing, another convenient computing paradigm has been presented for executing such applications. Edge computing introduces the concept of using the underutilised potential of a large number of computing enabled connected devices such as IoT, located outside the cloud. In order to develop optimal strategies to utilise this vast number of potential resources, a realistic test bed is required. However, due to the overwhelming scale and heterogeneity of the edge computing device deployments, the amount of effort and investment required to set up such an environment is high. Therefore, a realistic simulation environment that can accurately predict the behaviour and performance of a large-scale, real deployment is extremely valuable. While the state-of-the-art simulation tools consider different aspects of executing applications on edge or cloud computing environments, we found that no simulator considers all the key characteristics to perform a realistic simulation of the execution of DSP applications on edge and cloud computing environments. To the best of our knowledge, the publicly available simulators lack being verified against real world experimental measurements, i.e. for calibration and to obtain accurate estimates of e.g. latency and power consumption. In this paper, we present our ECSNeT++ simulation toolkit which has been verified using real world experimental measurements for executing DSP applications on edge and cloud computing environments. ECSNeT++ models deployment and processing of DSP applications on edge-cloud environments and is built on top of OMNeT++/INET. By using multiple configurations of two real DSP applications, we show that ECSNeT++ is able to model a real deployment, with proper calibration. We believe that with the public availability of ECSNeT++ as an open source framework, and the verified accuracy of our results, ECSNeT++ can be used effectively for predicting the behaviour and performance of DSP applications running on large scale, heterogeneous edge and cloud computing device deployments.

On the practical implementation of propagation delay and clock skew compensated high-precision time synchronization schemes with resource-constrained sensor nodes in multi-hop wireless sensor networks

In wireless sensor networks (WSNs), implementing an energy-efficient, high-precision time synchronization scheme on resource-constrained sensor nodes is challenging but essential to long-running applications for accurate environment monitoring and detection of important events, often covering a large area in the field. Our investigation of the practical implementation on a real testbed of the state-of-the-art WSN time synchronization scheme based on the asynchronous source clock frequency recovery and the reverse two-way message exchange, which can compensate for both propagation delay and clock skew for higher precision and energy efficiency, reveals that its performance on battery-powered, low-complexity sensor nodes is not up to that predicted from simulation experiments due to the limited precision floating-point arithmetic of sensor nodes. Noting the lower computational capability of typical sensor nodes and its impact on time synchronization, we propose an asymmetric high-precision time synchronization scheme that can provide high-precision time synchronization even with resource-constrained sensor nodes in multi-hop WSNs. In the proposed scheme, all synchronization-related computations are done at the head node equipped with abundant computing and power resources, while the sensor nodes are responsible for timestamping only to further release their computational burdens. Experimental results with a testbed based on TelosB motes running TinyOS demonstrate that the proposed time synchronization scheme can avoid time synchronization errors resulting from the single-precision floating-point arithmetic of the resource-constrained sensor nodes and achieve microsecond-level time synchronization accuracy in multi-hop WSNs.

Closed-Loop Identification for Model Predictive Control of HVAC Systems: From Input Design to Controller Synthesis

Heating, ventilation, and air conditioning (HVAC) systems are responsible for maintaining occupants’ thermal comfort and share a large portion of the overall building energy use. Hence, it is of great interest to improve the performance of HVAC control systems and thus the building energy efficiency. Model predictive control (MPC) has been proved to be a promising control strategy to be employed in this field. However, MPC implementation relies on the model of the system, and inaccurate models can deteriorate the control performance, while overly complicated models can lead to the prohibitive computational burden. Because of this, existing models do not usually allow the MPC controller to adjust multiple set points (e.g., both temperature and flow rates) and do not include the dynamics of the heating and ventilation subsystems with their local controllers. In this paper, we address the challenge of developing more reliable HVAC models for MPC controllers based on the experimental data. Data are obtained from an experiment designed using a graph theoretical technique, which guarantees maximum information content in the data. The resulting models are employed to design local controllers of the heating and ventilation subsystems, which are experimentally tested in a real HVAC test bed. A supervisory MPC controller that incorporates the closed-loop models of the heating and ventilation subsystems is then developed. This can lead to a control strategy able to more effectively adapt key HVAC set points based on weather conditions, occupancy, and actual thermal comfort, as shown by a numerical study based on the data from the HVAC test bed.

Challenges and opportunities for the simulation of calcium waves on modern multi-core and many-core parallel computing platforms.

State-of-the-art distributed-memory computer clusters contain multicore CPUs with 16 and more cores. The second generation of the Intel Xeon Phi many-core processor has more than 60cores with 16GB of high-performance on-chip memory. We contrast the performance of the second-generation Intel Xeon Phi, code-named Knights Landing (KNL), with 68computational cores to the latest multicore CPU Intel Skylake with 18cores. A special-purpose code solving a system of nonlinear reaction-diffusion partial differential equations with several thousands of point sources modeled mathematically by Dirac delta distributions serves as realistic test bed. The system is discretized in space by the finite volume method and advanced by fully implicit time-stepping, with a matrix-free implementation that allows the complex model to have an extremely small memory footprint. The sample application is a seven variable model of calcium-induced calcium release (CICR) that models the interplay between electrical excitation, calcium signaling, and mechanical contraction in a heart cell. The results demonstrate that excellent parallel scalability is possible on both hardware platforms, but that modern multicore CPUs outperform the specialized many-core Intel Xeon Phi KNL architecture for a large class of problems such as systems of parabolic partial differential equations.

On the Realistic Radio and Network Planning of IoT Sensor Networks.

Planning and deploying a functional large scale Wireless Sensor Network (WSN) or a Network of Internet of Things (IoTs) is a challenging task, especially in complex urban environments. A main network design bottleneck is the existence and/or correct usage of appropriate cross layer simulators that can generate realistic results for the scenario of interest. Existing network simulators tend to overlook the complexity of the physical radio propagation layer and consequently do not realistically simulate the main radio propagation conditions that take place in urban or suburban environments, thus passing inaccurate results between Open Systems Interconnection (OSI) layers. This work demonstrates through simulations and measurements that, by correctly passing physical information to higher layers, the overall simulation process produces more accurate results at the network layer. It is demonstrated that the resulting simulation methodology can be utilized to accomplish realistic wireless planning and performance analysis of the deployed nodes, with results that are very close to those of real test-beds, or actual WSN deployments.