arXiv subjects

Linda Doyle

Publications and source records attributed to Linda Doyle.

16 recordsLinked to original sources

arXiv2020

Towards scalable user-deployed ultra-dense networks: Blockchain-enabled small cells as a service

Neutral Host Small Cell Providers (SCP) represent a key element of the 5G vision of ultra-dense mobile networks. However, current business models mostly focus on multi-year agreements for large venues, such as stadiums and hotel chains. These business agreements are regulated through binding Service Level Agreements (SLAs), which tend to be too cumbersome and costly for smaller scale SCPs. As a result, the neutral host model does not scale up to its full potential. In this paper, we propose a framework to enable the participation of small- to medium-sized players in the cellular market as providers offering network resources to Mobile Network Operators (MNOs). To this purpose, we review the current and emerging spectrum and technology opportunities that SCPs can use for neutral host deployments. We also propose the use of blockchain-enabled smart contracts as a simple and cost-efficient alternative to traditional SLAs for small-scale SCPs. To demonstrate this, we describe a proof of concept implementation of an Ethereum-based smart contract platform for best-effort service between an SCP and an MNO. Our simulations on potential smart contract-based deployments in city centre Dublin show that the received signal strength in the considered area will increase by an average of $10$ percent.

cs.NI↗

arXiv2020

Blind Channel Estimation for Massive MIMO: A Deep Learning Assisted Approach

Large scale multiple-input multiple-output (MIMO) or Massive MIMO is one of the pivotal technologies for future wireless networks. However, the performance of massive MIMO systems heavily relies on accurate channel estimation. While the acquisition of channel state information (CSI) in such systems requires an increasingly large amount of training overhead as the number of users grows. To tackle this issue, in this paper, we propose a deep learning assisted blind channel estimation technique for orthogonal frequency division multiplexing (OFDM) based massive MIMO systems. We prove that by exploiting the asymptotic orthogonality of the massive MIMO channels, the channel distortion can be averaged out without the prior knowledge of channel impulse responses, and after some mathematical manipulation, different users transmitted data symbols can be extracted. Thus, by deploying a denoising convolutional neural network algorithm (DnCNN), we mitigate a remaining channel and noise effect to accurately detect the transmitted data symbols at the channel sounding stage. Using the detected data symbols as virtual pilots, we estimate the CSI of all the users at each BS antennas. Our simulation results testify the efficacy of our proposed technique and demonstrate that it can provide a mean square error (MSE) performance which coincides with that of the data-aided channel estimation technique.

eess.SP↗

arXiv2019

QuTiBench: Benchmarking Neural Networks on Heterogeneous Hardware

Neural Networks have become one of the most successful universal machine learning algorithms. They play a key role in enabling machine vision and speech recognition for example. Their computational complexity is enormous and comes along with equally challenging memory requirements, which limits deployment in particular within energy constrained, embedded environments. In order to address these implementation challenges, a broad spectrum of new customized and heterogeneous hardware architectures have emerged, often accompanied with co-designed algorithms to extract maximum benefit out of the hardware. Furthermore, numerous optimization techniques are being explored for neural networks to reduce compute and memory requirements while maintaining accuracy. This results in an abundance of algorithmic and architectural choices, some of which fit specific use cases better than others. For system level designers, there is currently no good way to compare the variety of hardware, algorithm and optimization options. While there are many benchmarking efforts in this field, they cover only subsections of the embedded design space. None of the existing benchmarks support essential algorithmic optimizations such as quantization, an important technique to stay on chip, or specialized heterogeneous hardware architectures. We propose a novel benchmark suite, QuTiBench, that addresses this need. QuTiBench is a novel multi-tiered benchmarking methodology that supports algorithmic optimizations such as quantization and helps system developers understand the benefits and limitations of these novel compute architectures in regard to specific neural networks and will help drive future innovation. We invite the community to contribute to QuTiBench in order to support the full spectrum of choices in implementing machine learning systems.

cs.AR↗

arXiv2018

Spectrum Matching in Licensed Spectrum Sharing

Spectrum sharing is one of the promising solutions to meet the spectrum demand in 5G networks that results from the emerging services like machine to machine and vehicle to infrastructure communication. The idea is to allow a set of entities access the spectrum whenever and wherever it is unused by the licensed users. In the proposed framework, different spectrum provider (SP) networks with surplus spectrum available may rank the operators requiring the spectrum, called spectrum users (SUs) hereafter, differently in terms of their preference to lease spectrum, based for example on target business market considerations of the SUs. Similarly, SUs rank SPs depending on a number of criteria, for example based on coverage and availability in a service area. Ideally, both SPs and SUs prefer to provide/get spectrum to/from the operator of their first choice, but this is not necessarily always possible due to conflicting preferences. We apply matching theory algorithms with the aim to resolve the conflicting preferences of the SPs and SUs and quantify the effect of the proposed matching theory approach on establishing preferred (spectrum) provider-user network pairs. We discuss both one-to-one and many-to-one spectrum sharing scenarios and evaluate the performance using Monte Carlo simulations. The results show that comprehensive gains in terms of preferred matching of the provider-user network pairs can be achieved by applying matching theory for spectrum sharing as compared to uncoordinated spectrum allocation of the available spectrum to the SUs.

cs.IT↗

arXiv2017

Complex Systems Science meets 5G and IoT

We propose a new paradigm for telecommunications, and develop a framework drawing on concepts from information (i.e., different metrics of complexity) and computational (i.e., agent based modeling) theory, adapted from complex system science. We proceed in a systematic fashion by dividing network complexity understanding and analysis into different layers. Modelling layer forms the foundation of the proposed framework, supporting analysis and tuning layers. The modelling layer aims at capturing the significant attributes of networks and the interactions that shape them, through the application of tools such as agent-based modelling and graph theoretical abstractions, to derive new metrics that holistically describe a network. The analysis phase completes the core functionality of the framework by linking our new metrics to the overall network performance. The tuning layer augments this core with algorithms that aim at automatically guiding networks toward desired conditions. In order to maximize the impact of our ideas, the proposed approach is rooted in relevant, near-future architectures and use cases in 5G networks, i.e., Internet of Things (IoT) and self-organizing cellular networks.

cs.NI↗

arXiv2017

Spectrum Monitoring for Radar Bands using Deep Convolutional Neural Networks

In this paper, we present a spectrum monitoring framework for the detection of radar signals in spectrum sharing scenarios. The core of our framework is a deep convolutional neural network (CNN) model that enables Measurement Capable Devices to identify the presence of radar signals in the radio spectrum, even when these signals are overlapped with other sources of interference, such as commercial LTE and WLAN. We collected a large dataset of RF measurements, which include the transmissions of multiple radar pulse waveforms, downlink LTE, WLAN, and thermal noise. We propose a pre-processing data representation that leverages the amplitude and phase shifts of the collected samples. This representation allows our CNN model to achieve a classification accuracy of 99.6% on our testing dataset. The trained CNN model is then tested under various SNR values, outperforming other models, such as spectrogram-based CNN models.

cs.NI↗

arXiv2017

Smart Contract SLAs for Dense Small-Cell-as-a-Service

The disruptive power of blockchain technologies represents a great opportunity to re-imagine standard practices of telecommunication networks and to identify critical areas that can benefit from brand new approaches. As a starting point for this debate, we look at the current limits of infrastructure sharing, and specifically at the Small-Cell-as-a-Service trend, asking ourselves how we could push it to its natural extreme: a scenario in which any individual home or business user can become a service provider for mobile network operators, freed from all the scalability and legal constraints that are inherent to the current modus operandi. We propose the adoption of smart contracts to implement simple but effective Service Level Agreements (SLAs) between small cell providers and mobile operators, and present an example contract template based on the Ethereum blockchain.

cs.NI↗

arXiv2017

5G Wireless and Wired Convergence in a Passive Optical Network using UF-OFDM and GFDM

The provision of both wireless and wired services in the optical access domain will be an important function for future passive optical networks (PON). With the emergence of 5th generation (5G) mobile communications, a move toward a dense deployment of small cell antenna sites, in conjunction with a cloud radio access network (C-RAN) architecture, is foreseen. This type of network architecture greatly increases the requirement for high capacity mobile fronthaul and backhaul links. An efficient way of achieving such connectivity is to make use of wavelength division multiplexed (WDM) PON infrastructure where wireless and wired services may be converged for distribution. In this work, for the first time, the convergence of 5G wireless candidate waveforms with a single-carrier wired signal is demonstrated in a PON. Three bands of universally filtered orthogonal frequency division multiplexing (UF-OFDM) and generalized frequency division multiplexing (GFDM), are transmitted at an intermediate frequency in conjunction with a digital 10Gb/s pulse amplitude modulation (PAM) signal in the downlink direction. Orthogonal frequency division multiplexing (OFDM) is also evaluated as a benchmark. Results show, for each waveform, how performance varies due to the 5G channel spacing - indicating UF-OFDM's superiority in terms of PON convergence. Successful transmission over 25km of fibre is also demonstrated for all waveforms.

cs.IT↗

arXiv2016

Deep Activity Recognition Models with Triaxial Accelerometers

Despite the widespread installation of accelerometers in almost all mobile phones and wearable devices, activity recognition using accelerometers is still immature due to the poor recognition accuracy of existing recognition methods and the scarcity of labeled training data. We consider the problem of human activity recognition using triaxial accelerometers and deep learning paradigms. This paper shows that deep activity recognition models (a) provide better recognition accuracy of human activities, (b) avoid the expensive design of handcrafted features in existing systems, and (c) utilize the massive unlabeled acceleration samples for unsupervised feature extraction. Moreover, a hybrid approach of deep learning and hidden Markov models (DL-HMM) is presented for sequential activity recognition. This hybrid approach integrates the hierarchical representations of deep activity recognition models with the stochastic modeling of temporal sequences in the hidden Markov models. We show substantial recognition improvement on real world datasets over state-of-the-art methods of human activity recognition using triaxial accelerometers.

cs.LG↗

arXiv2016

LSA-Advanced and C-RAN: A (5G) Romance of Many Dimensions

We examine the dimensional limitations of Licensed Shared Access (LSA) regulatory framework for spectrum sharing and propose a solution for its more dynamic implementation. We consider an additional dimension for sharing, beyond those of time, space and frequency, i.e. the sharing of the infrastructure. We explain why we believe that the emerging regulations and use-case scenarios are limiting the potential of LSA, and with this in mind make a set of recommendations to unlock it fully. As a specific case study, we present our architectural work around Cloud-based Radio Access Network (C-RAN) and LSA to demonstrate the possibility of a dynamic implementation of multiple incumbents and multiple LSA licensees engaged in sharing. Our case study highlights multiple ways for LSA framework to evolve into a global and flexible sharing model, and prove that LSA and C-RAN are a good match to rethink how we look at networks targeting a more flexible, efficient and profitable way of operating them.

cs.CY↗

arXiv2016

Census Tract License Areas: Disincentive for Sharing the 3.5GHz band?

Flexible licensing model is a necessary enabler of the technical and procedural complexities of Spectrum Access System (SAS)-based sharing framework. The purpose of this study is to explore the effectiveness of 3.5GHz Licensing Framework - based on census tracts as area units, areas whose main characteristic is population. As such, the boundary of census tract does not follow the edge of wireless network coverage. We demonstrate why census tracts are not suitable for small cell networks licensing, by (1) gathering and analysing the official census data, (2) exploring the boundaries of census tracts which are in the shape of nonconvex polygons and (3) giving a measure of effectiveness of the licensing scheme through metrics of area loss and the number of people per census tract with access to spectrum. Results show that census tracts severely impact the effectiveness of the licensing framework since almost entire strategically important cities in the U.S. will not avail from spectrum use in 3.5GHz band. Our paper does not seek to challenge the core notion of geographic licensing concept, but seeks a corrective that addresses the way the license is issued for a certain area of operation. The effects that inappropriate size of the license has on spectrum assignments lead to spectrum being simply wasted in geography, time and frequency or not being assigned in a fair manner. The corrective is necessary since the main goal of promoting innovative sharing in 3.5 GHz band is to put spectrum to more efficient use.

cs.CY↗

arXiv2016

Substitutability of Spectrum and Cloud-based Antennas in Virtualised Wireless Networks

Some of the new trends emerging in future wireless networks enable a vastly increased fluidity in accessing a wide range of resources, thus supporting flexible network composition and dynamic allocation of resources to virtual network operators (VNOs). In this work we study a new resource allocation opportunity that is enabled by the cloud radio access network architecture. In particular, we investigate the relationship between the cloud-based antennas and spectrum as two important resources in virtualized wireless networks. We analyze the interplay between spectrum and antennas in the context of an auction-based allocation mechanism through which VNOs can bid for a combination of the two types of resources. Our analysis shows that the complementarity and partial substitutability of the two resources significantly impact the results of the allocation of those resources and uncovers the possibility of divergent interests between the spectrum and the infrastructure providers.

cs.NI↗

arXiv2015

Effect of LOS/NLOS Propagation on Ultra-Dense Networks

This paper aims at investigating the achievable performance and the issues that arise in ultra-dense networks (UDNs), when the signal propagation includes both the Line-of-Sight (LOS) and Non-Line-Of-Sight (NLOS) components. Backed by an analytical stochastic geometry-based model, we study the coverage, the Area Spectral Efficiency (ASE) and the energy efficiency of UDNs with LOS/NLOS propagation. We show that when LOS/NLOS propagation components are accounted for, the network suffers from low coverage and the ASE gain is lower than linear at high base station densities. However, this performance drop can partially be attenuated by means of frequency reuse, which is shown to improve the ASE vs coverage trade-off of cell densification, provided that we have a degree of freedom on the density of cells. In addition, from an energy efficiency standpoint, cell densification is shown to be inefficient when both LOS and NLOS components are taken into account. Overall, based on the findings of our work that assumes a more advanced system model compared to the current state-of-the-art, we claim that highly crowded environments of users represent the worst case scenario for ultra-dense networks. Namely, these are likely to face serious issues in terms of limited coverage.

cs.IT↗

arXiv2015

A Stochastic Geometry Framework for LOS/NLOS Propagation in Dense Small Cell Networks

The need to carry out analytical studies of wireless systems often motivates the usage of simplified models which, despite their tractability, can easily lead to an overestimation of the achievable performance. In the case of dense small cells networks, the standard single slope path-loss model has been shown to provide interesting, but supposedly too optimistic, properties such as the invariance of the outage/coverage probability and of the spectral efficiency to the base station density. This paper seeks to explore the performance of dense small cells networks when a more accurate path-loss model is taken into account. We first propose a stochastic geometry based framework for small cell networks where the signal propagation accounts for both the Line-of-Sight (LOS) and Non-Line-Of-Sight (NLOS) components, such as the model provided by the 3GPP for evaluation of pico-cells in Heterogeneous Networks. We then study the performance of these networks and we show the dependency of some metrics such as the outage/coverage probability, the spectral efficiency and Area Spectral Efficiency (ASE) on the base station density and on the LOS likelihood of the propagation environment. Specifically, we show that, with LOS/NLOS propagation, dense networks still achieve large ASE gain but, at the same time, suffer from high outage probability.

cs.IT↗

arXiv2014

Effect of LOS/NLOS Propagation on Area Spectral Efficiency and Energy Efficiency of Small-Cells

In this paper we investigate the effect of Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS) propagation on the Area Spectral Efficiency (ASE) and on the energy efficiency of dense small-cell networks. We show that including both LOS and NLOS propagation in the path-loss model provides a completely different picture of the behaviours of ASE and energy efficiency than what would be observed in case of either LOS or NLOS propagation only. In particular, with combined LOS/NLOS path-loss, the ASE exhibits superlinear and sublinear behaviour at low and high cell densities, respectively. In addition, the energy efficiency as a function of the cell density has a global maximum and is not a monotonically increasing function like in case of LOS or NLOS propagation only. Based on our findings, we claim that Line-of-Sight (LOS) and Non-Line-of-Sight (NLOS) propagations play an important role in studying the performance of extremely dense small-cell networks.

cs.NI↗

arXiv2011

A Combinatorial Optimisation Approach to Designing Dual-Parented Long-Reach Passive Optical Networks

We present an application focused on the design of resilient long-reach passive optical networks. We specifically consider dual-parented networks whereby each customer must be connected to two metro sites via local exchange sites. An important property of such a placement is resilience to single metro node failure. The objective of the application is to determine the optimal position of a set of metro nodes such that the total optical fibre length is minimized. We prove that this problem is NP-Complete. We present two alternative combinatorial optimisation approaches to finding an optimal metro node placement using: a mixed integer linear programming (MIP) formulation of the problem; and, a hybrid approach that uses clustering as a preprocessing step. We consider a detailed case-study based on a network for Ireland. The hybrid approach scales well and finds solutions that are close to optimal, with a runtime that is two orders-of-magnitude better than the MIP model.

cs.AI↗