Quantifying Model Drift in Machine Learning for Estimating Wireless Link Quality

In this paper we consider the impact of quality of wireless sensor links on the overall detection performance of a large sensor network system. Independent and identical sensors gather observations regarding the presence or the absence of a phenomenon of interest and then transmit their binary decisions to a fusion center over parallel, non-interfering but slow Ralyeigh fading, wireless links. We derive asymptotic error exponents of the probability of false alarm and the probability of miss at the fusion center for the following cases: (a) BPSK modulation and (i) maximal ratio combining (ii) equal gain combining (iii) decision fusion (b) BFSK modulation and (i) square law combining (ii) decision fusion. In the case of BPSK, the EGC performs the best for low and moderate SNR, with the DF achieving the next best performance. The DF scheme performs the best for large SNR values, whereas the MRC performs the best for very low SNR values. Similar relative performance results were obtained by others for the case of a finite number of sensors. In the case of BFSK, square law combining outperforms DF, except for large SNR values. Finally, we show how the false alarm and the detection probabilities of the decision of a sensor, as seen at the fusion center, are altered by changes to the threshold of the matched filter receiver.

With the success of wireless sensor networks (WSNs), traditional engineering and infrastructure industries are starting to develop solutions using WSN technologies. One of the main challenges of designing and developing WSNs for industrial monitoring and control is satisfying their strict reliability requirements. In this paper, we present a network-level reliability model, namely, end-to-end data delivery reliability (E2E-DDR), for estimating and optimizing the reliability performance of WSNs. In the E2E-DDR model, a framework is presented for capturing the mapping function between the packet reception ratio, background noise, and received signal strength (RSS). We use an alpha-stable distribution to accurately represent the background noise and a modified log-normal path loss model to more realistically describe the RSS. We also report a comprehensive performance evaluation performed by applying the E2E-DDR model in a real-world case study to estimate the network-level reliability and optimize the WSN deployment parameters. Note to Practitioners —The goal of this paper is to improve the estimation and optimization of the network-level reliability performance of industrial wireless sensor networks (WSNs) to satisfy their strict control requirements. In harsh industrial application scenarios, many factors affect radio link quality, such as RF transmit power, communication distance, and random background noise. Existing link quality estimation approaches mainly focus on providing a smoothed estimation of radio link quality without any solution for optimizing communication reliability to satisfy certain requirements. This paper suggests a new approach in which WSN nodes are used to measure and estimate the parameters of the scenario in which the nodes will be deployed and then to estimate and optimize the worst case reliability (which is a lower-bound value rather than a smoothed value) to ensure that the network is qualified. Through a real-world case study, we demonstrate how to estimate the lower bound on reliability and how to optimize the reliability by computing the maximum deployment distance between nodes as an example. The experiments suggest that this approach is feasible.

Our research center has developed a cost-effective marine communication network architecture to provide affordable Internet access to marine fishermen. This solution uses a multi-level infrastructure network based on long range Wi-Fi and TDMA for backhaul. Directional antennas are used to increase the range. Standard Wi-Fi connectivity is provided on the boats to which the fishermen connect using their smartphones. The various angular and linear movements of the boat due to the oceanic waves will impact the alignment of the directional antenna to the base station thereby affecting the link quality. In this work, the extent of such movements is modeled theoretically, and the values are compared with the actual results obtained from field trials for sea states 4 and 5. It is also shown that the signal strength fluctuations are higher when the sea is rougher. Finally, future work planned for further analysis based on experimental modeling is also described.

Commercial cloud services are growing rapidly, and system architecture that can process large volumes of large-scale data for mobile devices and sensing devices is expected to emerge in the future. Mobile devices will serve as both client terminals and virtual machines in such architecture. However, end-to-end TCP sessions are affected by frequent changes in wireless link quality and movement of virtual machines in the cloud data center. In this paper, we propose a method to improve communication performance by dynamically reselecting and retuning congestion control algorithms and parameters according to changes in link quality and attributes of applications.

Among the wide range of wireless sensor network applications, many require reliable data communications such that data packets can be delivered to the destination without loss. This thesis is focused on one such application, viz. metal-fill monitoring system for the lost foam casting process implemented using capacitive sensors and Wireless Sensor Networks. Experiments were carried out in the TTU foundry to measure Radio Signal Strength and Link Quality and characterize data packet losses. The problem of packet losses associated with wireless transmission of data using wireless sensor networks with emphasis on the foundry environment has been explored and two solutions were investigated. The first is based on Delayed Retransmission and is suitable for offline monitoring. The second is for online monitoring and is based on Erasure Coding using Cauchy Reed - Solomon codes. The two solutions have been tested and evaluated for the metal-fill monitoring application. The entire system from sensing to data transmission provides a flexible and cost effective method to monitor the metal fill process.

In the resource constraint Wireless Sensor Network(WSN), link quality and node residual energy are the key factors that need to be considered in the routing problem. This paper proposes a routing algorithm based on game theory. Firs it sets up the game model based on the Quality of Service(QoS) and node residual energy. Then it shows that the game has a unique Nash equilibrium solution, and analyses the feasibility of the algorithm. The simulation results show that the proposed model could optimize the QoS, reduce the energy consumption of nodes and prolong the network life time.

In this paper, we propose an improved sectorization scheme, called Narrow Beam Quad-sector Cell (NBQC), in which each cell is divided into 4 sectors and each sector is covered by a 60° antenna. The NBQC structure allows to implement the concept of interleaved channel assignment (ICA), which can take full advantage of antenna directivity. With ICA, the NBQC system can enhance the system performance from several perspectives. First, the NBQC system with the ICA significantly improves the system capacity and quality of radio links. We demonstrate that in a typical radio environment, the NBQC system can achieve a reuse factor N = 2 with the signal-to-interference ratio (SIR) as high as 11 dB in 90 percent of the cell area. Comparing to the Global System for Mobile telecommunications (GSM) that requires N = 4 to achieve the SIR requirement of 9 dB, our proposed system thus improves the system capacity by a factor of two. Second, the proposed NBQC system can also improves the coverage performance. Third, the proposed system has better handoff performance because it is less sensitive to the impact of handoff hysteresis.

We propose a rate-adaptive multicast (RAM) routing protocol for mobile ad-hoc networks (MANETs) that is multirate-aware. During the process of path discovery, the quality of wireless links is estimated to suggest optimal transmission rates, which are then used to calculate the total transmission time incurred by the mobile nodes on a path. Among several considered paths from a source to a destination, RAM selects the path with the lowest total transmission time. Simulation results have shown that RAM outperforms single-rate multicast in terms of packet delivery ratio, packet end-to-end delay, and throughput of the multicast group. In this paper, we apply RAM to effectively support video multicast in MANETs

Fractures Robots (FracBots) are underground IoT (internet of things) wireless sensor nodes designed to be deployed in oil/gas reservoirs to map hydraulic fractures (HF) and monitor in real-time main reservoir parameters including pressure and temperature in real-time. A significant amount of FracBots will be injected inside hydraulic fractures immediately after conducting the fracturing job to gain the required data. They will launch the network connectivity in a fashion of FracBot-to-FracBot to collect, exchange data and transmit it to the base station installed in the wellbore. FracBots technology employs the same concept of the wireless underground sensor network (WUSN), a very capable technology for the applications in hydraulic fractures network. However, developing such technology is exceptionally challenging due to many environmental complications. The main challenges are severe environment, size, and energy limitations which contribute heavily to destroying the quality of the underground wireless communication link. FracBots technology comprises a gateway, a base station (BS), and FracBots. The main role of the BS is to transfer power to FracBots for charging purpose, gather data from FracBots, and finally transmit data to a gateway that send them to a centralized computer for further processing. The function of FracBot nodes is to sense the required data, send it wirelessly among them in multi-hop fashion and then communicate data to BS. This paper discusses the available wireless communication techniques including electromagnetic waves (EM) and magnetic induction (MI) wireless communication that might be suitable candidates for hydraulic fractures environment. They are compared in terms of the path loss model, path loss parameters, and the advantages and disadvantages of each technique. Moreover, we provide several simulations showing the importance of omnidirectional antenna and the effect of an alignment angle (between the axes of transmitter and receiver) on the path loss of the MI network. Also, we analyze the environmental effects on the path loss inside hydraulic fractures. After studying both EM and MI techniques in fractures environment. We found that the heterogenous environment stiffly influences the EM and MI communications inside hydraulic fractures in oil reservoirs. A hybrid solution that combines both EM and MI communication technologies will be very suitable for such an environment.

The quality of wireless communication links that employ code-division multiple access (CDMA) mainly suffers from user capacity limit caused by multiple access interference (MAI). To reduce the MAI, we propose parallel interference cancellation assisted channel estimation (PICACE) especially for the indoor reverse link and present theoretical results. We use the signal processing worksystem (SPW) simulator for evaluating 2-D linear 1st order Gaussian interpolation. Because each user's signal on the reverse link may experience different channel fading, we propose reference aided data structure to distinguish each user's own signal. The theoretical results illustrate that PICACE can extend the user capacity of CDMA.

This paper gives an overview on technical issues for future mobile networks. the QoS control for multimedia services, security function, expansion of network capacity, expansion of address space, multicast function, radio link quality and mobility control of terminals are taken up as technical issues. Then, a novel DLC scheme in wireless link named EREP-ARQ (Estimation and Report of Error Pattern ARQ) which can reduce the retransmission ratio is proposed. In this scheme, error pattern generated in the radio link is estimated at the receiver side and the estimated error pattern is returned to the transmission side. If the estimation is correct, no retransmission is needed. Compared to conventional SR-ARQ, throughput can be improved by 17% and Cell Loss Ratio (or Packet Loss Rate) can be reduced to less than 1/10 by the proposed EREP-ARQ.

Wireless LANs, CDPD, and microcellular data networks are fast becoming the preferred method of access to the Internet. Wireless links supporting end-user mobility, however, exhibit several characteristics that are different from traditional wired networks. These include: dynamically changing bandwidth due to mobile host movement in and out of cells where bandwidth is shared, high rates of packet corruption and subsequent loss, and frequent and lengthy disconnections due to obstacles, fading, and movement between cells. In addition, these effects are short-lived and difficult to reproduce, leading to a need for adequate measurement and analysis tools and for feedback utilities to allow the mobile user to respond to changes in wireless link quality.

The smart grid will play an important role in the future city to support the diversified energy supply. Wireless communication, the most cost-effective alternative to the traditional wire-lines, promises to provide ubiquitous bi-direction information channel for smart grid devices. However, due to the complex environment that smart grid devices located in, the wireless link is easily been interfered with and therefore appears strong stochastic features. Considering different smart grid application traffics have different and strict reliability requirements, the confidence interval lower boundary is more suitable to represent the worst-case reliability of the stochastic wireless link quality and trustworthy for judging whether the link quality is qualified for the next transmission. In this paper, we propose a Long-Short-Term-Memory (LSTM) based link quality confidence interval lower boundary prediction for the smart grid. According to the analysis of the characteristics of the wireless link, we employ the wavelet denoising algorithm to decompose the signal-to-noise ratio time series into the deterministic part and the stochastic part for training two LSTM neural networks. Then, the deterministic part and the variance of the stochastic part are predicted respectively. Lastly the confidence interval boundary is calculated. To verify the performance of the proposed LQP method, real-world experiments are carried out and the results show that our method is more accurate and trustworthy in comparison with other link quality prediction methods.

Machine learning (ML) is being used to improve intrusion detection mechanisms and identification in cyber security. Network data volume scaling (with the help of Machine learning) — Automated analysis and pattern recognition for large amounts of network-data, thereby detection of anomalies / potentially malicious activities that escape current rule-based techniques. By training ML models on historical data, these models can learn benign network behavior as well the anomalies in them that may result from malicious activities. The purpose of this essay/report is to begin taking a look under the hood at how ML can be used for security threat detection and analysis in-networking on an ongoing basis. This paper covers the automation of ML algorithms to enhance network security. Given the current state of electronic threats and their evolution, traditional security methods are typically insufficient. ML can analyze large volumes of data, learn patterns from it and this makes it suitable to complement network defense mechanisms. It then discusses different ML applications in network cybersecurity: intrusion detection, anomaly detection, spam and malware analysis; which the paper characterizes. It analyzes the potential, benefits and constrains of major ML methods in network security like supervised learning; unsupervised learning and reinforcement-learning. Finally, this paper represents recent progress in the use and impact of ML techniques along with case studies.The paper discusses the existing difficulties in the field such as the necessity for datasets and the vulnerability of machine learning models to adversarial attacks.T he paper also highlights avenues for exploration by focusing on developing scalable security solutions based on machine learning that are resilient and flexible.The goal of this examination is to offer both researchers and industry professionals valuable perspectives into the opportunities and obstacles linked to utilizing machine learning, in the domain of network security. ML methods have potential to improve network security by addressing the challenges posed by the increasing cyber threats that traditional security measures struggle to combat effectively over time. One key strength of ML lies in its capacity to analyze datasets and identify intricate patterns efficiently. The research paper delves into applications of ML in enhancing network security. The list covers security tools like Intrusion Detection Systems (IDS) examining malware and phishing attempts as well as anomalies in network activity and user behavior analysis (UEBA). The study explores both supervised and unsupervised learning methods. How they are used for quick threat detection and response in real time scenarios.You will find case studies and recent developments that showcase the implementation and effectiveness of these strategies.In addition the article delves into the obstacles linked to using machine learning techniques in network security including the necessity, for datasets, The paper's goal is to give an enlightening summary to scholars and practitioners about how machine learning can be applied to network security in order to provide solutions that are robust, adaptive, and scalable. To this end, it touches on several relevant aspects. One is the threat posed by adversarial attacks on the sorts of models that are likely to be used in this context. Another is the imperative, deriving from both adversarial threat and model drift, that models needed in this context be available in a form usable for continuous update. Keywords: Network Security, machine learning (ML), Intrusion Detection Systems (IDS), entity behavior analytics (UEBA).

Machine learning (ML) has been used to develop increasingly accurate link quality estimators for wireless networks. However, more in-depth questions regarding the most suitable class of models, most suitable metrics and model performance on imbalanced datasets remain open. In this paper, we propose a new tree-based link quality classifier that meets high performance and fairly classifies the minority class and, at the same time, incurs low training cost. We compare the tree-based model, to a multilayer perceptron (MLP) non-linear model and two linear models, namely logistic regression (LR) and SVM, on a selected imbalanced dataset and evaluate their results using five different performance metrics. Our study shows that 1) non-linear models perform slightly better than linear models in general, 2) the proposed non-linear tree-based model yields the best performance trade-off considering F1, training time and fairness, 3) single metric aggregated evaluations based only on accuracy can hide poor, unfair performance especially on minority classes, and 4) it is possible to improve the performance on minority classes, by over 40% through feature selection and by over 20% through resampling, therefore leading to fairer classification results.