No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Comparison of the long-term forecasting method of RSSI by machine learning
Abstract
In order to improve the efficiency of spectrum use, systems that share spectrum while avoiding interference between different systems are being investigated. In the millimeter-wave band, which is expected to be utilized in the future, the received power fluctuates due to quasi-static obstructions such as people and vehicles, but such temporal variations have not been taken into account in conventional methods. In this paper, we use a variety of machine learning algorithms for comparative evaluation to forecast the temporal fluctuations of radio propagation due to changes in the number of people and vehicles in order to achieve more dynamic spectrum access.
Scikit-learn is a Python module integrating a wide range of state-of-the-art
machine learning algorithms for medium-scale supervised and unsupervised
problems. This package focuses on bringing machine learning to non-specialists
using a general-purpose high-level language. Emphasis is put on ease of use,
performance, documentation, and API consistency. It has minimal dependencies
and is distributed under the simplified BSD license, encouraging its use in
both academic and commercial settings. Source code, binaries, and documentation
can be downloaded from http://scikit-learn.sourceforge.net.
Learning to store information over extended time intervals by recurrent backpropagation takes a very long time, mostly because of insufficient, decaying error backflow. We briefly review Hochreiter's (1991) analysis of this problem, then address it by introducing a novel, efficient, gradient-based method called long short-term memory (LSTM). Truncating the gradient where this does not do harm, LSTM can learn to bridge minimal time lags in excess of 1000 discrete-time steps by enforcing constant error flow through constant error carousels within special units. Multiplicative gate units learn to open and close access to the constant error flow. LSTM is local in space and time; its computational complexity per time step and weight is O(1). Our experiments with artificial data involve local, distributed, real-valued, and noisy pattern representations. In comparisons with real-time recurrent learning, back propagation through time, recurrent cascade correlation, Elman nets, and neural sequence chunking, LSTM leads to many more successful runs, and learns much faster. LSTM also solves complex, artificial long-time-lag tasks that have never been solved by previous recurrent network algorithms.
Tree boosting is a highly effective and widely used machine learning method. In this paper, we describe a scalable end-to-end tree boosting system called XGBoost, which is used widely by data scientists to achieve state-of-the-art results on many machine learning challenges. We propose a novel sparsity-aware algorithm for sparse data and weighted quantile sketch for approximate tree learning. More importantly, we provide insights on cache access patterns, data compression and sharding to build a scalable tree boosting system. By combining these insights, XGBoost scales beyond billions of examples using far fewer resources than existing systems.
Tree boosting is a highly effective and widely used machine learning method. In this paper, we describe a scalable end-to-end tree boosting system called XGBoost, which is used widely by data scientists to achieve state-of-the-art results on many machine learning challenges. We propose a novel sparsity-aware algorithm for sparse data and weighted quantile sketch for approximate tree learning. More importantly, we provide insights on cache access patterns, data compression and sharding to build a scalable tree boosting system. By combining these insights, XGBoost scales beyond billions of examples using far fewer resources than existing systems.
Spectrum occupancy models are very useful in cognitive radio designs. They can be used to increase spectrum sensing accuracy for more reliable operation, to remove spectrum sensing for higher resource usage efficiency or to select channels for better opportunistic access, among other applications. In this survey, various spectrum occupancy models from measurement campaigns taken around the world are investigated. These models extract different statistical properties of the spectrum occupancy from the measured data. In addition to these models, spectrum occupancy prediction is also discussed, where the auto-regressive and/or moving-average models are used to predict the channel status at future time instants. After comparing these different methods and models, several challenges are also summarized based on this survey.
As a part of the global effort to address the overwhelming demand for wireless broadband capacity, the wireless communities in the United States have undertaken innovative initiatives such as tiered-access to shared spectrum. The Federal Communications Commission has proposed a dynamic spectrum management framework for a Citizen Broadband Radio Service (CBRS) governed by a spectrum access system (SAS). The implementation of a SAS capable of dynamic frequency assignment and interference management is critical for the success of the proposed framework. In this paper we present the efforts toward a spectrum sharing system in the U.S. context by summarizing different interest groups??? standpoint on the FCC proposed framework. We also present an example SAS architecture that accommodates the tiered access to shared spectrum and example approaches to achieve important SAS capabilities.
Spectrum sharing between an existing incumbent spectrum user and an LTE/LTEAdvanced network with conditions that resemble exclusive licensing have become an appealing solution for mobile network operators to respond to the growing traffic and spectrum demand in a timely manner. While traditional exclusive licensing continues to be the preferred option for MNOs, the new sharing- based licensed shared access (LSA) concept is receiving growing interest in research, regulation and standardization. When applied to mobile communications, the LSA concept would allow an MNO to share spectrum from another type of incumbent spectrum user under a regulator??s supervision with predetermined rules and conditions that guarantee operational certainty for both MNO and incumbent. This article reviews different types of spectrum bands for LTE/LTE-Advanced and beyond networks, and focuses on the LSA concept as a spectrally efficient solution for spectrum access in the future. The article identifies the key stakeholders, including the incumbent spectrum user, the MNO, and the regulator, and their roles in the LSA concept. The key elements in the LSA concept are reviewed, and a work flow for the life cycle of the LSA concept is proposed, consisting of LSA preparation, licensing, deployment, and release phases. The tasks of the key stakeholders in the different phases of the LSA work flow are discussed. The LSA concept can offer a complementary approach to traditional exclusive licensing and license-exempt operations with features that benefit all involved stakeholders. It can be realized with reasonable modifications to the existing network infrastructure and regulatory framework with two new elements for managing the varying spectrum availability: the LSA Repository and LSA Controller.
Guidelines for evaluation of radio interface technologies for IMT-2020
- M Itu-R Report
ITU-R Report M.2412-0, "Guidelines for evaluation of radio interface technologies for IMT-2020," 2017.