A Cross-layer Dual Queue Approach for Improving TCP Fairness in Infrastructure WLANs

Fairness is one of the most important performance measures in IEEE 802.11 Wireless Local Area Networks (WLANs), where channel is accessed through competition. In this paper, we focus on the fairness problem between TCP uplink and downlink flows in infrastructure WLANs from the cross-layer perspective. First, we show that there exists a notable discrepancy between throughput of uplink flow and that of downlink flow, and discuss its root cause from the standpoint of different responses to TCP data packet drop and TCP ACK packet drop at the access point (AP) buffer. In order to mitigate this unfairness, we propose a dual queue scheme, which works in a cross-layer manner. It employs two separate queues at the AP, one for the data packets of downlink TCP flows and another for the ACK packets of uplink TCP flows, and selects these queues with appropriate probabilities so that TCP per-flow fairness is improved. Moreover, we analyze the behavior of the dual queue scheme and derive throughputs of uplink and downlink flows. Based on this analysis, we obtain the optimal queue selection probabilities for fairness. Extensive simulation results confirm that the proposed scheme is effective and useful in resolving the TCP unfairness problem without deteriorating overall utilization.     

Measuring and modelling sewer pipes from video

This article presents a system for the automatic measurement and modelling of sewer pipes. The system recovers the interior shape of a sewer pipe from a video sequence which is acquired by a fish-eye lens camera moving inside the pipe. The approach is based on tracking interest points across successive video frames and posing the general structure-from-motion problem. It is shown that the tracked points can be reliably reconstructed despite the forward motion of the camera. This is achieved by utilizing a fish-eye lens with a wide field of view. The standard techniques for robust estimation of the two- and three-view geometry are modified so that they can be used for calibrated fish-eye lens cameras with a field of view less than 180°. The tubular arrangement of the reconstructed points allows pipe shape estimation by surface fitting. Hence, a method for modelling such surfaces with a locally cylindrical model is proposed. The system is demonstrated with a real sewer video and an error analysis for the recovered structure is presented.     

A generic camera model and calibration method for conventional, wide-angle, and fish-eye lenses

Fish-eye lenses are convenient in such applications where a very wide angle of view is needed, but their use for measurement purposes has been limited by the lack of an accurate, generic, and easy-to-use calibration procedure. We hence propose a generic camera model, which is suitable for fish-eye lens cameras as well as for conventional and wide-angle lens cameras, and a calibration method for estimating the parameters of the model. The achieved level of calibration accuracy is comparable to the previously reported state-of-the-art.     

A picture is worth a thousand words: an empirical study on the influence of content visibility on diffusion processes within a virtual world

ABSTRACT

Studying information diffusion and the spread of goods in the real world and in many digital services can be extremely difficult since information about the information flows is challenging to accurately track. How information spreads has commonly been analysed from the perspective of homophily, social influence, and initial seed selection. However, in virtual worlds and virtual economies, the movements of information and goods can be precisely tracked. Therefore, these environments create laboratories for the accurate study of information diffusion characteristics that have been difficult to study in prior research. In this paper, we study how content visibility as well as sender and receiver characteristics, the relationship between them, and the types of multilayer social network layers affect content absorption and diffusion in virtual world. The results show that prior visibility of distributed content is the strongest predictor of content adoption and its further spread across networks. Among other analysed factors, the mechanics of diffusion, content quality, and content adoption by users’ neighbours on the social activity layer had very strong influences on the adoption of new content.     

SERA: a secure energy reliability aware data gathering for sensor networks

Wireless sensor networks are used in many applications in military, ecology, health, and other areas. These applications often include the monitoring of sensitive information making the security issue one of the most important aspects to consider in this field. However, most of protocols optimize for the limited capabilities of sensor nodes and the application specific nature of the networks, but they are vulnerable to serious attacks. In this paper, a Secure Energy and Reliability Aware data gathering protocol (SERA) is proposed, which provides energy efficiency and data delivery reliability as well as security. The proposed protocol’s security was confirmed by a formal verification carried out using the AVISPA tool and analysis of the most common network layer attacks such as selective forwarding, sinkhole, Sybil, wormhole, HELLO flood, and acknowledgment spoofing attacks. Additionally, a visual simulation environment was developed to evaluate the performance of the proposed protocol.     

Averaging in vitro cardiac field potential recordings obtained with microelectrode arrays

Extracellular field potential (FP) recordings with microelectrode arrays (MEAs) from cardiomyocyte cultures offer a non-invasive way of studying the electrophysiological properties of these cells at the population level. Several studies have examined the FP properties of cardiomyocytes of various origins, including stem cell-derived cardiomyocytes. This focus reflects growing importance and interest in the field of MEA. High-quality cardiac FP signals are often difficult to obtain, especially from stem cell-derived cardiomyocyte cultures, which represent an important new field in cardiac electrophysiology. One way to improve the quality of these recordings is to average the cardiac FP signals. To date, however, no studies have examined the effect of averaging on cardiac FP signals. We report here that cardiac FP averaging can yield higher-quality signals than original individual FPs, and therefore promise more accurate detection of different phases and analysis of the cardiac FP signal. Averaged signals improved the signal-to-noise ratio (SNR), and obtaining reliable averages required approximately 50 cardiac cycles. We therefore propose that routine cardiac FP averaging can serve as a tool to compare the effects of different experimental conditions or stimuli on the properties of cardiac FPs.     

Efficient Multisplitting Revisited: Optima-Preserving Elimination of Partition Candidates

We consider multisplitting of numerical value ranges, a task that is encountered as a discretization step preceding induction and also embedded into learning algorithms. We are interested in finding the partition that optimizes the value of a given attribute evaluation function. For most commonly used evaluation functions this task takes quadratic time in the number of potential cut points in the numerical range. Hence, it is a potential bottleneck in data mining algorithms.We present two techniques that speed up the optimal multisplitting task. The first one aims at discarding cut point candidates in a quick linear-time preprocessing scan before embarking on the actual search. We generalize the definition of boundary points by Fayyad and Irani to allow us to merge adjacent example blocks that have the same relative class distribution. We prove for several commonly used evaluation functions that this processing removes only suboptimal cut points. Hence, the algorithm does not lose optimality.Our second technique tackles the quadratic-time dynamic programming algorithm, which is the best schema for optimizing many well-known evaluation functions. We present a technique that dynamically—i.e., during the search—prunes partitions of prefixes of the sorted data from the search space of the algorithm. The method works for all convex and cumulative evaluation functions.Together the use of these two techniques speeds up the multisplitting process considerably. Compared to the baseline dynamic programming algorithm the speed-up is around 50 percent on the average and up to 90 percent in some cases. We conclude that optimal multisplitting is fully feasible on all benchmark data sets we have encountered.     

Forecasting Financial Time Series Using Robust Deep Adaptive Input Normalization

Deep Learning provided powerful tools for forecasting financial time series data. However, despite the success of these approaches on many challenging financial forecasting tasks, it is not always straightforward to employ DL-based approaches for highly volatile and non-stationary time financial series. To this end, in this paper, an adaptive input normalization layer that can learn to identify the distribution from which the input data were generated and then apply the most appropriate normalization scheme is proposed. This allows for promptly adapting the input to the subsequent DL model, which can be especially important, given recent findings that hint at the existence of critical learning periods in neural networks. Furthermore, the proposed method operates on a sliding window over the time series allowing for overcoming non-stationary issues that often arise. It is worth noting that the main difference with existing approaches is that the proposed method does not just learn to perform static normalization, e.g., using a fixed set of parameters, but instead it adaptively calculates the most appropriate normalization parameters, significantly improving the robustness of the proposed approach when distribution shifts occur. The effectiveness of the proposed formulation is verified using extensive experiments on three challenging financial time-series datasets.

     

A decision-theoretic approach for model interpretability in Bayesian framework

Machine Learning - A salient approach to interpretable machine learning is to restrict modeling to simple models. In the Bayesian framework, this can be pursued by restricting the model structure...