Search

Service provisioning systems assign users to service providers according to allocation criteria that strike an optimal trade-off between users Quality of Experience (QoE) and the operation cost endured by providers. These systems have been leveraging Smart Contracts (SCs) to add trust and transparency to their criteria. However, deploying fixed allocation criteria in SCs does not necessarily lead to the best performance over time since the blockchain participants join and leave flexibly, and their load varies with time, making the original allocation suboptimal. Furthermore, updating the criteria manually at every variation in the blockchain jeopardizes the autonomous and independent execution promised by SCs. Thus, we propose a set of light-weight agents for SCs that are capable of optimizing the performance. We also propose using online learning SCs, empowered by Deep Reinforcement Learning (DRL) agent, that leverage the chained data to continuously self-tune its allocation criteria. We show that the proposed learning-assisted method achieves superior performance on the combinatorial multi-stage allocation problem while still being executable in real-time. We also compare the proposed approach with standard heuristics as well as planning methods. Results show a significant performance advantage over heuristics and better adaptability to the dynamic nature of blockchain networks. IEEE Scopus

Distributed learning algorithms aim to leverage distributed and diverse data stored at users' devices to learn a global phenomena by performing training amongst participating devices and periodically aggregating their local models' parameters into a global model. Federated learning is a promising paradigm that allows for extending local training among the participant devices before aggregating the parameters, offering better communication efficiency. However, in the cases where the participants' data are strongly skewed (i.e., non-IID), the local models can overfit local data, leading to low performing global model. In this paper, we first show that a major cause of the performance drop is the weighted distance between the distribution over classes on users' devices and the global distribution. Then, to face this challenge, we leverage the edge computing paradigm to design a hierarchical learning system that performs Federated Gradient Descent on the user-edge layer and Federated Averaging on the edge-cloud layer. In this hierarchical architecture, we formalize and optimize this user-edge assignment problem such that edge-level data distributions turn to be similar (i.e., close to IID), which enhances the Federated Averaging performance. Our experiments on multiple real-world datasets show that the proposed optimized assignment is tractable and leads to faster convergence of models towards a better accuracy value. 2013 IEEE. Scopus

Artificial intelligence (AI) has witnessed a substantial breakthrough in a variety of Internet of Things (IoT) applications and services, spanning from recommendation systems to robotics control and military surveillance. This is driven by the easier access to sensory data and the enormous scale of pervasive/ubiquitous devices that generate zettabytes (ZB) of real-time data streams. Designing accurate models using such data streams, to predict future insights and revolutionize the decision-taking process, inaugurates pervasive systems as a worthy paradigm for a better quality-of-life. The confluence of pervasive computing and artificial intelligence, Pervasive AI, expanded the role of ubiquitous IoT systems from mainly data collection to executing distributed computations with a promising alternative to centralized learning, presenting various challenges. In this context, a wise cooperation and resource scheduling should be envisaged among IoT devices (e.g., smartphones, smart vehicles) and infrastructure (e.g. edge nodes, and base stations) to avoid communication and computation overheads and ensure maximum performance. In this paper, we conduct a comprehensive survey of the recent techniques developed to overcome these resource challenges in pervasive AI systems. Specifically, we first present an overview of the pervasive computing, its architecture, and its intersection with artificial intelligence. We then review the background, applications and performance metrics of AI, particularly Deep Learning (DL) and online learning, running in a ubiquitous system. Next, we provide a deep literature review of communication-efficient techniques, from both algorithmic and system perspectives, of distributed inference, training and online learning tasks across the combination of IoT devices, edge devices and cloud servers. Finally, we discuss our future vision and research challenges., Survey paper accepted in IEEE Communications Surveys & Tutorials

This deliverable presents DAEMON���s initial view on the problem of real-time control and network functions intelligence. The work presented in this document tackles two main points: ��� Understanding what the main challenges from the architectural perspective are and proposing initial solutions to overcome them. This activity has a strong relationship with the work carried out in WP2 in terms of requirements and architectural design. Namely, we identified additional requirements that are specifically related to the fast timescale NI applications. ��� The design of specific NI algorithms that focus on the inclusion of NI very close to the user plane. A lot of attention was paid to edge network functions, such as those related to the access network (including Reconfigurable Intelligent Surfaces). In total, we describe 11 algorithms. These algorithms can be generally categorized into two main classes: ��� Algorithms that jointly optimize radio and compute resources for intelligent NFV in vRANs. These algorithms are mainly presented in section 4 and deploy novel mechanisms and techniques for solving such difficult and coupled decision-making problems. ��� Algorithms that are capable of running on different timescales. These algorithms are of specific interest to the work package as they directly tackle the highly heterogeneous nature of edge and fog domains and are mainly presented in section 5. Each proposed algorithm specifically mentions the targeted KPI so as to create a clear, propaedeutic environment for the follow-up WP5 activities., This project has received funding from the European Union's Horizon 2020 Research and Innovation Programme under grant agreement No. 101017109.

Smart Contracts (SCs), which are software programs that run on blockchain platforms, provide appealing security guarantees characterized by decentralized, autonomous, and verifiable execution. On the other hand, Service Provisioning (SP) systems (i.e., systems that assign users to service providers in a way that maximizes the global utility) have been leveraging SCs to provide trust and transparency features. Such features are obtained by deploying the SP's assignment criteria as an SC on the blockchain. However, deploying optimal assignment criteria as SCs does not guarantee the best performance over time since the blockchain participants join and leave flexibly, and their load varies with time, potentially deeming the initial assignment sub-optimal. Furthermore, modifying the criteria manually by a third party at every variation in the blockchain jeopardizes the autonomous and independent execution promised by SCs. Thus, in this paper, we propose the use of online learning SCs that leverage the chained data to continuously self-tune their assignment criteria and maintain maximum utility. We show that the proposed data-driven method can achieve high performance on the multi-stage assignment problem. We also compare the proposed approach to multiple assignment algorithms as well as planning methods. Results show a significant performance advantage over heuristics and better adaptability to the dynamic nature of blockchain networks compared to planning techniques. 2021 IEEE. Scopus

With the increase in Electric Vehicles (EVs) penetration, their charging needs form an additional burden on the grid. Thus, charging coordination is necessary for safe and efficient EV use. The scheduling of EVs is especially essential in Vehicle-to-Grid schemes, where EVs energy cost can be reduced by intelligently managing the charging and discharging according to real-time prices and owners' needs. Further, utilities can perform load shifting by price-based demand-response programs. However, the scheduling problem is challenging in the presence of multiple unknown variables, such as real-time prices, commuting behavior, and energy needs. Most of the scheduling techniques attempt to model the system uncertainties (e.g., forecast prices) and plan accordingly. In this paper, we propose the use of a model-free Reinforcement Learning (RL) technique that can deduce a charging/discharging strategy without requiring any model of the system. We evaluate the performance of the proposed system with real-world data and investigate the learned charging/discharging pattern to show the effectiveness of the proposed method.

Traceability and autonomous business logic execution are highly desirable features in IoT monitoring applications. Traceability enables verifying signals history for security or analytical purposes. On the other hand, the autonomous execution of pre-defined rules establishes trust between parties involved in such applications and improves the efficiency of their workflow. Smart Contracts (SCs) firmly guarantee these two requirements due to the blockchain's immutable distributed ledger and secure cryptographic consensus rules. Thus, SCs emerged as an appealing technology for monitoring applications. However, the cost of using public blockchains to harvest these guarantees can be prohibitive, especially with the considerable fluctuation of coin prices and different use case requirements. In this paper, we introduce a general SC-based IoT monitoring framework that can leverage the security features of public blockchains while minimizing the corresponding monetary cost. The framework contains a reinforcement learning agent that adapts to users needs and acts in real-time to smartly set the data submission rate of IoT sensors. Results based on the Ethereum protocol show significant potential cost saving depending on users preferences. Qatar University Scopus

Modern day robotics is fast developing and evolving in various areas of multi sensor applications like body sensor networks and mine detection. A flexible platform that can be easily integrated into variety of sensors is very essential in any complex environment. A reconfigurable hardware provides a suitable platform to identify and improve existing models used in metal detection industry. In this paper, we propose and implement a metal detection module using Terasic Spider Robot, planned to be used in landmine detection operations. A hardware circuit model to detect metal was designed for the metal detection module and embedded on the TSR. The on board control system was implemented using a reconfigurable DE0 Nano System on Chip platform that can further process the information from the metal detector using efficient algorithm. The movement of the TSR was controlled remotely by Bluetooth using a smartphone app designed specifically for this application. The design also intimates the user with a message on detecting the metal. The design was implemented successfully and the metal detection module detected buried metals at a depth of maximum 7cm.

While usable water on earth is sparse and costly to treat, statistics show excessive use worldwide. Penalties are issued to limit the excessive consumption; however, their impact is marginal as they do not identify the reason for the exorbitant consumption. Therefore, giving users the full awareness about their water consumption will greatly help in minimizing the water waste. This article aims to present a smart self-powered water monitoring system that leverages IoT and cloud computing. The system consists of an IoT device that can be installed at any water source, a cloud application to receive the data from the devices, and a mobile app to visualize the water consumption at every monitored source. This system gives the user the ability to identify the location and time of the excessive usage and leaks on the mobile phone. Experiment results confirm the self-powered premise of the solution. A prototype of mobile application and backend have been developed and tested. This work was partially supported by Microsoft for being a winner project in Microsoft Imagine Cup 2017 Competition - Middle East & Africa. Scopus

Redundant Array of Independent Disks (RAID) storage architectures provide protection of digital infrastructure against potential disks failures. For example, RAID-5 and RAID-6 architectures provide protection against one and two disk failures, respectively. Recently, the data generation has significantly increased due to the emergence of new technologies. Thus, the size of storage systems is also growing rapidly to accommodate such large data sizes, which increases the probability for disks failures. This necessitates a new RAID architecture that can tolerate up to three disk failures. RAID architectures implement coding techniques. The code specifies how data is stored among multiple disks and how lost data can be recovered from surviving disks. This abstract introduces a novel coding scheme for new RAID-7 architectures that can tolerate up to three disks failures. The code is an improved version of the existing BP-XOR code and is called “Almost BP-XOR”.There are multiple codes that can be used for RAID-7 architectures. However, [5,2] BP-XOR codes have significantly lower encoding and decoding complexities than most common codes [1]. Regardless of this fact, this code does not achieve the fastest data decoding and reconstruction speeds due to its relatively low efficiency of 0.4. Furthermore, the existence of MDS [6,3] bx6 BP-XOR codes, b>2 (which achieves efficiency of 0.5) is still an open research question. This work proposes [6,3] 2 x 6 Almost BP-XOR codes. These codes largely utilize the simple and fast BP-XOR decoder while achieving an efficiency of 0.5, leading to the fastest recovery from disk failures among other state-of-the-art codes. An algorithm to generate a [6, 3] 2 x 6 Almost BP-XOR code has been developed and an example code is provided in Table 1. The [6, 3] 2 x 6 Almost BP-XOR codes are constructed in a way that any three-column-erasure pattern will result in one of the following two main scenarios. First: At least one of the surviving degree-three encoding symbols contains two known information symbols. This scenario occurs in 70% of three-column erasure cases (i.e. 14 out of the 20 possible cases). The recovery process in such scenario is identical to that of the BP-XOR codes; Knowing any two information symbols in a degree-three encoding symbol is sufficient to know the third information symbol through performing a simple XOR operation. Second: None of the surviving degree-three encoding symbols contains two known information symbols. This scenario occurs in the remaining 30% of three-column erasure cases (i.e., 6 out if the possible 20). The BP-XOR decoder fails in such a scenario. However, due to the construction of the codes, at least one surviving degree-three encoding symbol contains a known information symbol. Thus, knowing one of the reaming two information symbols in such a degree-three encoding symbol will initiate the BP-XOR decoder again.Table 2 shows these erasure patterns along with an expression for one of the missing information symbols. these expressions can be stored in buffers and used whenever the corresponding erasure pattern occurs. Solutions in Table 2 are derived from the inverse of a 6x6 submatrix that results from a generator matrix G by deleting columns from G corresponding to erased code columns. The read complexity of almost BP-XOR codes is 1. On the other hand. The decoding of almost BP-XOR codes require just 6 XOR operations when for a given three-column-erasure pattern BP-XOR decoding succeeds. However, when the BP- XOR decoder fails, it will require up to 15 XOR operations in total. The normalized repairing complexity is 15/6 = 2.5.Experimentally, Fig.1 shows that the proposed Almost BP-XOR codes require the least amount of time to decode and reconstruct erased columns. Thus, it is concluded that the [6, 3] 2x6 almost BP-XOR codes are best suited for RAID-7 system that requires storage efficiency of 0.5. References [1] Y. Wang. “Array BP-XOR codes for reliable cloud storage systems,” in Proc. of the 2013 IEEE International Symposium on Information Theory (ISIT), pp. 326–330, Istanbul, Turkey, July 2013.NoteFigures, Tables, and more details are provided in the complete attached file titled Abstract-ARC18.pdf, (respecting the same word count restriction).

Background and Objectives One of the key problems in mobile robotics is the ability to understand and analyze the surrounding environment in a useful way. This is especially important in dangerous applications where human involvement should be avoided. A clear example of employing the robots in dangerous applications is mine detection which is mostly done through metal detection techniques. Among the various types of walking robots, Hexapod walking robots offer a good static stability margin and faster movement especially in rough terrain applications [1] Thus, the “Hexapod Terasic Spider Robot” is a suitable platform for the metal detection purpose especially that it is equipped with Altera DE0-Nano field programmable gate arrays (FPGA) SoC which allows for extremely high performance and accuracy. This work introduces a novel implementation of a metal detection module on the Terasic Spider Robot; the metal detection module is designed and interfaced with the robot in order to perform the metal detection. The user can control the robot and receive the feedback through a Bluetooth-enabled android phone. In addition, a general-purpose design flow that can be used to implement other applications on this platform is proposed. This proves the versatility of the platform as well.Method The designed metal detection module (MDM) is mainly based on an oscillator and a coil, its operation principle is that when the coil approaches a metal, the frequency of the oscillator will change [2]. This frequency change can be accurately monitored in real time using the FPGA SoC board. Thus, the module can be used for detecting metals. The metal detection module is interfaced with DE0-Nano SoC board where the detection algorithm is implemented. The development of the algorithm is carried out on the board available on this robot. The board includes a FPGA, which provides a high-performance and real-time implementation of parts of the algorithm, and a hard processor system (HPS) running Linux OS which can be used to easily interface the board with other computer systems and peripherals such as mobile phones and cameras[3]. As shown in Fig. 1, the detection algorithm is based on hardware/software co-design; the output of the MDM is provided to the FPGA part of the board in order to achieve an accurate and real-time monitoring. Upon detection, the FPGA sends a detection signal through the shared memory interface to the HPS part of the board. The HPS is then responsible for sending a warning to the mobile through multi-threaded communication application that is running on the HPS. Figure 1 General architecture of the metal detection system In order to implement the metal detection algorithm on the Terasic Spider Robot, it was necessary to formulate and follow the design flow provided in Fig. 2. This design flow can be used to implement other applications that can utilize the hardware/software co-design approach for better performance. Figure 2 General purpose design flow for the Altera Terasic Spider Robot Platform. Results and discussion Due to the coil specification and the circuit design. The frequency captured at normal situations is (no metal presence) is 2155 ± 20 Hz. The frequency increases Inversely proportional to the distance of the metal from the coil. In other words, the frequency increases when the distance between the metal and the coil decrease. When a metal whose size is at least the same size as the coil is present at 7 cm distance from the detection coil, the frequency will exceed 2200 Hz Regardless of the medium. The tested medium is wood. However, similar results were obtained with air medium. These numbers are specific to the proposed system. Changing the circuit parameters will increase the detection distance if desired. For example, having more coil turns and bigger diameter as well as faster oscillation will increase the detection distance. To avoid any interference between the robot body and the metal detection circuit readings, a 15 inches plastic arm is used to connect the metal detection module to the body of the robot. The electronics components is attached to this arm to the nearest possible point to the coil. The metal detection module attached to a plastic arm and then to the robot. the metal detection module and the spider robot is shown in Fig. 3 and 4 respectively. Figure 3 The Metal Detection Circuit Combined with the Arm Fig. 4 MDM Connected to the Terasic Spider Robot The robot is then controlled through a mobile application, the mobile application is modified so that the robot can send feedback (detection warning) to the mobile phone. Figure 5 shows an example of the notification message «Metal Detected» whenever a metal is detected. Figure. 5. Metal Detection Message for Mobile Application Interface Summary and Conclusion This abstract includes a general description of research project that aims to utilize the Terasic Spider Robot platform to perform accurate and real-time metal detection. This is an important application that helps humans avoid involvement in dangerous operations like mine detection. Nonetheless, a general-purpose design flow is proposed for the benefit of the research community and anyone who intends to implement an application on this platform in the future. Acknowledgment This project was funded by Qatar University Internal Grants program. References [1] Y. Zhu, B. Jin, Y. Wu, T. Guo and X. Zhao, «Trajectory Correction and Locomotion Analysis of a Hexapod Walking Robot with Semi-Round Rigid Feet», Sensors, vol. 16, no. 9, p. 1392, 2016. [2] T. Alauddin, M. T. Islam and H. U. Zaman, «Efficient design of a metal detector equipped remote-controlled robotic vehicle,» 2016 International Conference on Microelectronics, Computing and Communications (MicroCom), Durgapur, 2016, pp. 1-5 [3] «Cyclone V Device Overview», Altera, 2016. [Online]. Available: https://www.altera.com/en_US/pdfs/literature/hb/cyclone-v/cv_51001.pdf. [Accessed: 16- Oct- 2017]