Skip to content
Events

Communication networks beyond the capacity crunch - further discussion

Scientific meeting

Starts:

May
132015

09:00

Ends:

May
142015

17:00

Location

Kavli Royal Society Centre, Chicheley Hall, Newport Pagnell, Buckinghamshire, MK16 9JJ

Overview

Satellite meeting organised by Professor Andrew Ellis, Professor Sir David Payne CBE FREng FRS and Professor David Saad

Copyright: Artida/Shutterstock

Event details

This satellite meeting followed the discussion meeting Communication networks beyond the capacity crunch and was a more focused discussion of the issues raised in the previous meeting, and possible solutions.

Talks at this meeting were predominantly given by early-career researchers selected following a call for proposals. Audio recordings will be available on this webpage shortly.

Download the meeting programme

Event organisers

Select an organiser for more information

Schedule of talks

Session 1: Overcoming fundamental obstacles

7 talks Show detail Hide detail

Optical communication systems with soft-decision FEC: replacing the FEC limit paradigm

Dr Alex Alvarado, University College London, UK

Abstract

The so-called FEC limit paradigm is the prevalent practice for designing optical communication systems to attain a certain bit-error rate (BER) without forward error correction (FEC). Under this paradigm, the existence of an FEC code that brings down the BER after decoding to the desired target level is assumed. In this presentation, we challenge this paradigm and show that the concept of a channel-independent FEC limit is invalid for modern soft-decision bit-wise decoding and low to medium code rates. A better predictor is the generalized mutual information, which is shown to give consistent post-FEC BER predictions across different channel conditions and modulation formats. Extensive optical full-field simulations and experiments were carried out in both the linear and nonlinear regimes to confirm the theoretical analysis. The decoding of modern soft-decision FEC codes is computationally intensive, and thus, this GMI limit paradigm becomes a powerful design tool that allows an accurate prediction of the BER after decoding without actually simulating the soft-decision FEC.

Show speakers

Beyond the crunch with specialty fibres

Dr Behrad Gholipour, Nanyang Technological University, Singapore / University of Southampton, UK

Abstract

Optical fibres provide a mature mass-manufacturable technology that has given rise to complex networks of interconnected computing nodes transferring information around the planet. With ever increasing demand for information, new approaches need to be taken to avoid a capacity crunch. Our research involves three layers of activity, novel fabrication techniques enabling multi-material nano-composite embedded fibres, all-optical fibre based devices for in-line information processing, culminating in the exploration of alternative network protocols that may drastically increase the efficiency and capacity of existing networks. In this pursuit, we demonstrate a fabrication method taking advantage of planar nanofabrication in conjunction with fibre drawing. Through this, different mutlimaterial nanowire embedded fibres are demonstrated, providing a platform for in-line nanowire fibre embedded light sources and novel photodetectors, removing electro-optic bottlenecks in current networks. Furthermore, we utilize photomodulation effects within semiconducting fibres to realise optical axons and photonic synapses based on the alloy gallium lanthanum oxysulphide. Finally, by combining this alloy with silicate in hybrid core/clad fibres, optical fibre analogues of electronic transistors are demonstrated, essential for realizing next generation all optical data processing networks. Through this approach, communication networks can be pushed beyond the capacity crunch into a realm of multimaterial all-optical cognitive photonic networks.

Show speakers

An asymptotically zero increase in capacity: why compensating nonlinearity will never meet capacity demands

Dr Domaniç Lavery, University College London, UK

Abstract

Current research efforts are focussed on overcoming the apparent limits of communication in single mode optical fibre resulting from distortion due to fibre nonlinearity. Recently, it was experimentally demonstrated that this so-called 'nonlinear Shannon limit' is not a fundamental limit; thus it is pertinent to review where the fundamental limits of communication lie, and direct future research on this basis. This presentation reviews the intrinsic limits of optical communication over standard single mode optical fibre (SMF) and shows that bandwidth, and not optical launch power, is the key to unlocking the capacity of SMF. The empirical limits of silica fibre homogeneity, and practical limits of transceiver design both introduce a weak upper bound to the capacity of communication using SMF, on the order of 1 Pbit/s. Transmission rates in excess of 1 Pbit/s are shown to be possible, however, with currently available optical fibres, attempts to transmit beyond this rate will lead to an asymptotically zero increase in capacity. Crucially, a rate of 1 Pbit/s is not achievable when transmitting only in the C- and L-bands.

Show speakers

High capacity enablers - science and technology

Dr Fatima Gunning, Department of Physics and Tyndall National Institute, University College Cork

Abstract

Capacity crunch has been discussed extensively in the last few years, with many suggestions on how to avoid it or on how to overcome it, if ever possible. In this talk we will discuss the interpretations for capacity crunch, the possible solutions to increase capacity, or at least increase spectral efficiency, if standard single mode fibres are to remain the main transmission medium, and alternative solutions for high demand links.

Show speakers

Regeneration limit of classical Shannon capacity

Dr Mariia Sorokina, Aston University, UK

Abstract

Since Shannon derived the seminal formula for the capacity of the additive linear white Gaussian noise channel, it has commonly been interpreted as the ultimate limit of error-free information transmission rate. However, the capacity above the corresponding linear channel limit can be achieved when noise is suppressed using nonlinear elements; that is, the regenerative function not available in linear systems. Regeneration is a fundamental concept that extends from biology to optical communications. All-optical regeneration of coherent signal has attracted particular attention. Surprisingly, the quantitative impact of regeneration on the Shannon capacity has remained unstudied. Here we propose a new method of designing regenerative transmission systems with capacity that is higher than the corresponding linear channel, and illustrate it by proposing application of the Fourier transform for efficient regeneration of multilevel multidimensional signals. The regenerative Shannon limit—the upper bound of regeneration efficiency—is derived.

Show speakers

Communication beyond the nonlinear-Shannon limit

Mr Son T. Le, Aston University, UK

Abstract

Continuing demand from the growing number of bandwidth-consuming applications and on-line services (such as cloud computing, HD video streams etc.) is pushing the required communication systems capacity close to the theoretical limit of a standard single-mode fibre, which is, in turn, imposed by the inherent fibre nonlinearity (Kerr effect) and often referred as the nonlinear-Shannon limit. In recent years, extensive efforts in attempting to surpass the Kerr nonlinearity limit have been made through several nonlinearity compensation techniques including digital back propagation (DBP), mid-link optical phase conjugation (OPC) and phase conjugated twin wave (PCTW). However, many challenges in applying the aforementioned nonlinear compensation methods remain because the transmission technologies utilized in optical fibre communication systems were originally developed for linear communication channels. Alternatively, several nonlinear transmission schemes, in which the fibre nonlinearity is taken into account in a ‘constructive way’, have also been proposed as promising techniques to combat the fibre nonlinearity. These include eigenvalue communications and nonlinear inverse synthesis (NIS). In this talk, we discuss the advantages, challenges and provide a comprehensive comparison of the achievable performance gains of DBP, mid-link OPC and PCTW. We also discuss in details the design, performance and challenges of eigenvalue communications and NIS as an emerging transmission technology for communicating beyond the nonlinear-Shannon limit.

Show speakers

High capacity optical fiber transmission experiments using multiple modes

Dr Vincent Sleiffer, Yacht, The Netherlands

Abstract

Space-division multiplexing (SDM) is considered as one of the possibilities to overcome or at least delay the much expected capacity crunch. The MODEGAP project was one of the groups in the world, trying to develop and demonstrate SDM technology. This presentation will show some key achievements of the system level experiments, which were conducted in a laboratory environment and during a field trial. The focus will be on mode-division multiplexing as a means to increase the capacity of a single-core fibre system, and will show straight-line and re-circulating loop experiments using few-mode fibre and hollow-core photonic band-gap fibre. Finally a potential gradual upgrade of legacy single-mode technology towards MDM technology is demonstrated.

Show speakers

Session 2: Enhancement to the physical layer

7 talks Show detail Hide detail

High density space division multiplexed transmission systems

Dr Chigo Okonkwo, Eindhoven University of Technology, The Netherlands

Abstract

In recent years, society’s requirement to remain connected 24hrs per day continues unabated as corroborated by major internet exchanges in Europe (e.g. Amsterdam and London) which offer peering services connecting Internet service providers to end users and businesses. The demand from Internet based services continues at an exponential rate requiring optical transmission systems primarily based on single mode transmission fibre for point to-point and/or point-to multipoint networks.  The capacity in single mode fibres (SMF) is defined by the non-linear Shannon limit. With this fundamental limit in SMF, a capacity ‘crunch’ is looming and methods to address this issue is being investigated. One of the methods proposed is space division multiplexing. Although, more SMF can be deployed, employing additional multiplexing layer such as multiple fibre cores, multiple modes or a combination of both methods in a single fibre can provide a cost effective scaling in capacity. These systems require new low loss components and fibres to access the capacity. This talk will discuss the state of the art few-mode and multimode transmission systems currently demonstrating more than 6-spatial mode transmission in a single fibre. The challenges of multiplexing into the fibres and the digital signal processing complexity will be discussed.

Show speakers

Emerging Technical Approaches to Scaling Optical Fibre Networks

Professor David Richardson FREng, University of Southampton, UK

Abstract

Researchers are within a factor of 2 or so from realizing the maximum practical transmission capacity of conventional single mode fibre transmission technology operating at 1550nm and new approaches offering the potential for more cost effective scaling of network capacity than simply installing more and more conventional single mode systems in parallel need to be explored. In this talk I will review potential physical layer options that can be considered to address this requirement including the use of space division multiplexed fibre technology, ultra-broadband fibre amplifiers and the potential for reduction in both fibre loss and nonlinearity.

Show speakers

Visible light communication using Gallium Nitride micro-LEDs

Mr Hyunchae Chun, University of Oxford, UK

Abstract

Visible light communications (VLC) is a potential candidate in dealing with the capacity crunch, providing 100s of THz of unregulated bandwidth. The adoption of solid-state lighting is creating a new illumination and communications infrastructure that can assist in alleviating the wireless communications spectrum ‘drought’. Novel Gallium Nitride LEDs, and integrated transmitter and receiver subsystems are being developed under a UK EPSRC funded programme on ultra-parallel visible light communications (UP-VLC). This has led to data rates of 3G bit/s being transmitted with a single LED, using orthogonal frequency division multiplexing techniques, a current world record. In addition we have developed and demonstrated integrated multi-input multi-output (MIMO) subsystems operating at Gbit/s. These will form a foundation for future systems, with the potential to combine multi-colour transmission and illumination. Such systems have the potential to provide extremely high rate ‘visibly secure’ wireless communications. We will present recent results from this work, and detail how VLC might fit into the future landscape of wireless and fibre communications.

Show speakers

Capacity and complexity Increase in MIMO-SDM systems

Dr Kai Shi, University College London, UK

Abstract

Space-division multiplexing (SDM) has been widely viewed as the only option to further scale the capacity of optical networks. The upper bound of the capacity C in any multiple-input multiple-output (MIMO) system scales by CS × M, where M is the number of modes that are supported by the fibre and addressed by the transponders, and CS is the single-mode capacity. However, as the capacity increases linearly through the use of more spatial modes, the digital signal processing (DSP) complexity can increase at different rates, dependent on the different algorithms being used and the differential group delays (DGDs) between the spatial modes. In this presentation, we studied the complexity of both time domain equalizers (TDEs) and frequency domain equalizers (FDEs) at a specific transmission distance, to reveal situations where the increase of the capacity offered by the SDM outperforms the increase in complexity. Three different SDM fibre types is examined: a conventional graded index multimode fibre (GI-MMF), a ring core fibre (RCF), and a graded index few mode fibre (GI-FMF). From the study, we find that the most promising scheme is to use the delayed TDE in GI-FMF SDM systems with low DGD between mode groups and no DGD between degenerate modes in each mode group.

Show speakers

Telecommunications beyond silicon: one atom thick chalcogenide photodiodes

Dr Kevin Chung-Che Huang, University of Southampton, UK

Abstract

One atom thick, two dimensional materials exhibit remarkable properties and are strong contenders to lead the way to the post silicon era. Graphene, the first two dimensional material   has  carrier mobility up to 200,000 cm2/V-1s-1.  This makes it a perfect candidate for high frequency electronic devices. However the lack of a bandgap limits it use for future optoelectronic fast devices.  Two dimensional transition metal chalcogenides (TMDC’s) are semiconductors of the type MX2  where M is a transition metal and X a chalcogen atom, such as sulphur, selenium or tellurium. In electronic applications, this group of 2D materials has the potential to outperform silicon and more importantly even graphene.  When going from bulk to two dimensions, TMDC’s also exhibit a direct band gap allowing for switching current ratios up to 108.  Remarkably, this band gap can be fine-tuned simply by changing the number of the atomic layers giving unprecedented device versatility. We demonstrate promising results on atomically thin, one layer and multi-layer MoS2 (molybdenum disulphide) based photo transistors with high responsivity and high on / off current ratio.

Show speakers

Hollow core photonic bandgap fibres for ultra-high capacity data transmission

Dr Natalie Wheeler, University of Southampton, UK

Abstract

Hollow core photonic bandgap fibres (HC-PBGFs) enable low loss light guidance in an air core and represent one of the most exciting recent developments in fibre technology. HC-PBGFs possess outstanding qualities which are highly relevant for applications in high capacity data transmission such as ultimate low latency, ultra-low non linearity and the promise of lower attenuation than current state-of-the-art conventional solid fibres. Interestingly, this lowest attenuation is predicted to be achieved in the 2 µm spectral region, where the development of new components, such as the wide bandwidth thulium-doped fibre amplifier, is attracting increasing interest and opening up the potential of a new transmission band. Furthermore, HC-PBGFs typically support many guided modes which can be exploited to achieve capacity enhancement through space division multiplexing.

We report extensive progress in both fabrication and characterisation of HC-PBGFs, including demonstration of fibres with record combinations of low loss and wide bandwidth, record fibre lengths, first demonstrations of high capacity single mode transmission at 1.55 and 2 µm and multimode transmission at 1.55 µm, as well as the development of a portfolio of tailored characterisation techniques to provide extensive data on modal qualities and fibre uniformity. We discuss future challenges for implementation.

Show speakers

Optical amplifiers for space division multiplexed transmission

Dr Yongmin Jung, University of Southampton, UK

Abstract

Space division multiplexing utilizing few-mode fibres supporting multiple spatial modes, is currently under intense investigation as an efficient approach to overcome the current capacity limitations of high-speed long-haul transmission systems based on single mode optical fibre. In order to realize the potential energy and cost savings offered by SDM systems, the individual guided modes should be simultaneously amplified within a few-mode erbium doped fibre amplifier (FM-EDFA). In this presentation, a review of recent progress on the implementation of FM-EDFAs for SDM transmission is provided. Especially, the design and construction of fully fiberised few-mode EDFAs will be discussed.

Show speakers

Session 3: Future networks

7 talks Show detail Hide detail

Towards terabit wide-coverage indoor optical wireless communications

Mr Ariel Gomez Diaz

Abstract

Wireless transmission capacity has always lagged that of optical fibres, usually by many orders of magnitude. High-frequency RF wireless offers rates closer to that of fibres, but at the carrier frequencies used and allowed power levels only narrow line of sight channels are available. This capability comes at the cost of complex electronics and components. Our approach is to dispense with the RF channel, and to build a fibre-(optical wireless)-fibre link. Holographic beamsteering is used to direct light exiting a fibre through free space to a terminal where a similar holographic beamsteerer is used to couple the light back into the fibre. This leads to a transparent optical end-to-end architecture that operates bi-directionally. We have demonstrated an eye-safe point-to-point, bidirectional, 112 Gigabit/s wide field of view link that uses coherent optical transmission, with a coverage area of 35 m2 in an indoor room-like environment. This system has the potential to provide Terabit/s aggregate data rates to nomadic users. Performance compares extremely favourably with high-frequency radio frequency line-of-sight approaches, at much reduced complexity, intrinsic transparency and bi-directional operation. Such an approach has the potential to create an all optical network, spanning both wireless and fibre domains, with near term application in broadcast and data centres, and longer term applications in ultra-high speed indoor systems.

Show speakers

Towards ultimate convervence of all networks

Professor Dimitra Simeonidou, University of Bristol, UK

Abstract

Global demand for broadband communications continues to increase substantially every year. A major factor contributing to this demand is the growing number of fixed and mobile broadband users, data-hungry applications like video as well as an ever-increasing number of network-connected everyday objects and machines. These trends pose entirely new challenges related to data volume, granularity, end-to-end connectivity and reach as well as increasing heterogeneity in network technologies (i.e. wireless and wired), networked-connected devices (i.e. sensors, mobile phones, computers, TVs, Data Centres) and services (i.e. Tbps data transfer for e-science, ultra-low latency financial transaction, real-time media streaming, kbps for sensor-based monitoring). Addressing these challenges necessitates radically new network models supporting convergence of traditionally separate network technology domains and offering high flexibility and adaptability in data granularity and throughput. This talk will discuss novel approaches to achieve ultimate network convergence enabled by radically new technology agnostic architectures targeting a wide range of applications and end users. The aim is to facilitate optimal interconnection of any network technology domains, networked devices and data sets with high flexibility, resource and energy efficiency, and to satisfy the full range of Quality of Service (QoS) and Quality of Experience (QoE) requirements.

Show speakers

Orders of magnitude higher throughput in cellular communications by optical attocells

Dr Dushyantha Basnayaka, University of Edinburgh, UK

Abstract

Global mobile voice and data traffic reached 2.5 exabytes (2.5 billions GB) per month at the end of 2014, up from 1.5 exabytes per month at the end of 2013, and monthly global mobile traffic will surpass 24.3 exabytes by 2019. Clearly this exponential growth cannot be handled by radio frequency (RF) cellular networks alone. Major telecommunication operators worldwide already offload some if not majority of mobile traffic onto Wi-Fi networks. As a consequence, about 80 percent of all data traffic on smartphones currently travels through Wi-Fi networks. However, the radio spectrum is limited. In contrast, the visible light spectrum is 10,000 times larger than the entire 30 GHz of RF spectrum which currently hosts all commercial RF systems. This has led to the introduction of Li-Fi which defines high-speed, bidirectional, networked and mobile wireless communications using light. Li-Fi opens up the opportunity to introduce a new small cell layer to the currently deployed heterogeneous wireless networks. A network of Li-Fi access points forms the optical attocellular layer which does not create any interference to existing RF networks. The Li-Fi attocell layer exploits the existing lighting infrastructure, and off-the-shelf LED light bulbs. The recent research efforts dedicated to Li-Fi have resulted in a number of advancements. They include the new generation of faster LEDs and modulation schemes which enable multi gigabyte indoor communication links, and multiuser Li-Fi systems for both uplink and downlink, and self-sufficient optical energy harvesting VLC access points. The current theoretical and experimental research at University of Edinburgh considers networking challenges of Li-Fi systems such as optimum attocell deployment scenarios, resource allocation schemes, and load balancing schemes to enhance the average and outage user rate performance. Moreover, our research shows that hybrid Wi-Fi and Li-Fi systems can significantly improve capacity, coverage and user experience which is an important finding in the context of 5G requirements such as to support 100 times more devices to enable the Internet of Things as well as to carry 10,000 times increased wireless traffic. This talks summarises the key research findings on the introduction of a new optical attocell layer in heterogeneous wireless networks aimed at future 5G networks.

Show speakers

High fidelity optical carrier dissemination using coherent communication techniques: HIPERFREQ Project

Dr Ehsan Sooudi, Tyndall National Institute and Department of Physics, University College Cork, Ireland

Abstract

Dissemination of optical clocks enables wider usage of such precise frequencies in fundamental physics tests, comparison of frequency references and definition of a second. Current dissemination techniques involve a bi-directional link and amplifier to stabilize the delivered optical carrier which is all-optical and usually requires a dedicated (dark) fibre link. This scheme is not generally deployable within the fibre-based telecommunication networks in which optical/electrical/optical conversions are occurring and the transmission is uni-directional. Emerging applications such as novel GPS systems and time synchronization and frequency transfer through optical networks require dissemination schemes deployable in optical networks. In this presentation, we discuss the HIPERFREQ project with European Space Agency which investigates a uni-directional dissemination approach based on coherent communication techniques, compatible with fibre optic networks. For the first time, a 10 Gb/s BPSK signal from an ultra-narrow linewidth laser was transmitted through a field-installed optical fibre with round-trip length of 124 km from Cork to Clonakilty using self-homodyne detection and optical injection-locking to recover and re-generate the optical carrier. The beating linewidth of the original carrier and the transmitted one was measured to be ~ 2.8 kHz with long-term fractional stability (true Allan deviation) of 3.3 x10-14 for 1 s averaging time.

Show speakers

Optimisation technique for optical OFDM systems

Miss Funmilayo Ogunkoya, Glasgow Caledonian University, UK

Abstract

The convergence of technology, embedded computing devices, telecommunication and networking of systems, devices and services has led to the development of integrated applications and increase in data traffic over transmission media. Optical wireless communication (OWC) systems have been demonstrated as a transmission technique that can meet the ever-increasing data rate demand of these applications. However, one of the modulation techniques in OWC – orthogonal frequency division multiplexing (OFDM) – capable of supporting high data rate suffer from high peak-to-average power ratio (PAPR) problem. Transmission of optical OFDM signal with high peaks results into bit-error rate (BER) degradation. In this work, we present performance evaluation of the pilot-assisted OFDM PAPR reduction technique proposed for optical communication systems using empirical and theoretical approach. The PAPR reduction is achieved by rotating the phase of OFDM data symbols with P-iterations of randomly generated pilot symbol. Conventional OFDM system utilises pilot signal for channel estimation. In our work, we use the pilot signal to serve a dual purpose of PAPR reduction at the transmitter and channel estimation at the receiver to enhance OFDM. Our results show that the PAPR values of the pilot-assisted OFDM are lower than that of basic OFDM where no PAPR reduction technique is applied and there is no BER penalty in implementing the pilot-assisted technique.

Show speakers

Engineering emergence for large-scale cyber physical systems through software

Dr Mariam Kiran, University of Bradford, UK

Abstract

The emergence of technologies, with ‘Internet of things concept’ of connected devices has given rise to a world controlled by software. This software needs to be intelligent to control network formations, perform load balancing, for dynamic and manageable architecture formation to allow a cost effective emergent network. Agile methodologies and on demand ‘everything as a service’ services through cloud computing, need to come together to understand the lower level software interactions to make intelligent formations emerge at higher levels.

My research involves investigating issues on how lower lever interactions, using software, can aid in formation of intelligent emergent architectures for the Internet of things or cyber physical systems. I am using principles of complex systems such as insect colonies to allow individual units to interact in manners to allow large scale society behaviours to emerge. These concepts are adopted in software based agents which interact with other similar agents to evolve behaviour of large scale networks. This research also involves investigating issues of risk, security, network resilience and data transfer to build reliable decentralised cloud networks which act as complex systems for achieving efficient data management for Smart City applications, mimicking how natural systems build complex reliable structures.

Show speakers

Seamless and adaptive interfaces for wired and multi-technology wireless for future converged networks

Dr Paul Anthony Haigh, University of Bristol, UK

Abstract

In this talk, an overview of the recently funded EPSRC TOUCAN project; which aims to provide major innovation in network convergence. As opposed to modern day, where the majority of access networks are based on Wi-Fi, in the future it is foreseen that access networks will be based on multi-technology components such as LTE, Li-Fi, Bluetooth and Wi-Fi, amongst others. Each of these technologies has its own standards, protocols and throughput requirements. This presents a significant challenge to network designers at the access/intermediate network edge who must provide a transparent interface as well as efficient aggregation of the data flows, making best use of resources. Here, this interface will be discussed including current progress and future directions.

Show speakers

Session 4: Efficient methods to utilise communication networks

6 talks Show detail Hide detail

Solving the bigger problem: point-point transmission in the broader network context

Professor Andrew Lord, BT

Abstract

This talk will overview BT's optical network structures and discuss the main themes and challenges faced by a large operator in this space – challenges ranging from cost effective capacity, energy and space constraints, to bandwidth dynamicity and end to end network orchestration and control. The talk will seek to put the basic point-to-point bandwidth transmission objective into a broader context, covering spectrum and bit rate flexibility. It will review recent research into flexgrid and elastic network design, examining the potential for novel bitrate variable transponder technology, using results from the Idealist FP7 collaborative project. Finally it will seek to highlight key and interesting research challenges for the coming years.

Show speakers

Optimising the MIMO Wi-Fi performance

Dr Evangelos Mellios, University of Bristol, UK

Abstract

More than 70% of all the data traffic today is over Wi-Fi, while the demand for high data-rate demanding applications such as high-definition video streaming is increasing constantly. Modern WiFi standards (namely 802.11ac) try to meet these demands for fast, reliable and robust communications with the employment of advanced technologies, such as MIMO antenna systems and beam-forming transmission techniques. However, our experience from testing numerous (in excess of 40) state-of-the-art Wi-Fi access-points shows that due to poor MIMO antenna design, the benefits offered by this technology are limited. Furthermore, current international regulations for EIRP reduce significantly the advantages of MIMO techniques such as beam-forming. This paper will initially present an innovative, rigorous and repeatable MIMO Wi-Fi testing procedure, which combines 3D antenna radiation pattern measurements in an anechoic chamber, with state-of-theart propagation modelling and an accurate but time-efficient system-level simulator. Conductive testing of real devices using a channel emulator (hardware-on-the-loop) will also be discussed. We will then try to answer the difficult question “what is a good MIMO antenna array?” by quantifying the benefits of various MIMO antenna configurations. Finally, we will present an innovative technique for maximising the EIRP while still remaining within the limits imposed by the regulations.

Show speakers

Tackling the capacity crunch: an attempt to rationalisation and mitigation

Dr Marco Ruffini, University of Dublin, Ireland

Abstract

This talk will first debate the prospect of a future capacity crunch and propose its rationalisation by discussing the relationship between user-generated data and overall network capacity. It will then propose scalable network architectures, discussing results from the European project DISCUS, to tackle the capacity requirements for the coming future.

Show speakers

Ultimate classical communication rates of quantum optical channels

Dr Raúl García-Patrón Sánchez, Université Libre de Bruxelles, Belgium

Abstract

The capacity of Additive White Gaussian Noise channels plays a pervasive role in optical-fiber communication, as it is a key tool in the estimation of the capacity of the physical communication channel. In this talk we will present a quantum generalization of the AWGN channel capacity and discuss its different regimes: from convergence to the classical AWGN to high divergence, depending on the value of the signal and noise. Our work solves a longstanding open problem in the field of quantum optics, raised in 1962 by J. P. Gordon on his seminal paper ‘Quantum effects in communications systems’. We will conclude with a discussion of potential implications and application of our result.

Show speakers

From 4G LTE-A to 5G massive MIMO – theory, simulation and testbed

Miss Siming Zhang, University of Bristol, UK

Abstract

Wireless data traffic demands in cellular networks are increasing dramatically and significant improvements are required in capacity. Some solutions for the next-generation cellular networks include the deployment of heterogeneous networks, where lower power picocell base-stations will complement the traditional macrocell base-stations; Multi-User MIMO techniques, which are expected to enhance spectral efficiency significantly; and Massive MIMO technologies, which promise drastic spectral and energy enhancements by deploying large antenna arrays. This paper will evaluate, quantify, and compare the performance of an up to 16x8 Single-User MIMO solution, a Multi-User MIMO scenario, and a Massive MIMO system (with up to 128 antenna elements at the base-station) in realistic urban heterogeneous environments while single- and dual-antenna cases are considered at the UE. A 3D laser-scanned city database (Bristol, UK) was used as input to a ray-tracing channel propagation model along with measured patterns for the individual base-station and user-equipment antenna elements. Hundreds of thousands of links were evaluated to ensure statistical relevance, whereas High Performance Computing is crucial to performing such computation-intensive tasks. Finally, we will introduce a novel 128-element Massive MIMO testbed which is being developed by the University of Bristol, in collaboration with National Instruments, Bristol City Council and Lund University, which allows antenna segments to be co-located or distributed when deployed.

Show speakers

Practical considerations on discrete multi-tone transmission for cost-effective access networks

Dr Zhixin Liu, University of Southampton, UK

Abstract

Short distance optical communications such as inter data center connections and passive optical networks require low cost and high capacity optical transmission links, typically covering a span of 10 km to 50 km. Direct-modulation direct-detection (DM-DD) schemes require minimum optical hardware, minimizing the cost. Recent research demonstrated up to 117 Gb/s DM-DD single-channel transmission at 1310 nm and 2×50 Gb/s transmission at 1550 nm, both using discrete multi-tone (DMT) modulation. This is enabled by novel lasers, digital signal processing (DSP), and most importantly the state-of-the-art high-speed electronics. However, the achievable capacity does not scale linearly with the available electronic bandwidth. Our recent study shows it may be more sensible to operate the system at a slightly lower capacity in the presence of dispersion. This could make significant savings as lower bandwidth electronics would be used, reducing cost, power consumption and simplifying the heat dissipation management, resulting in lower cost-per-bit. Additionally, we studied the impact of the commonly used waveform pre-emphasis and showed that it is of limited value in directly modulated DMT. We believe that the results of our analysis will allow optimization of the cost-per-bit without significantly compromising the performance of the system.

Show speakers
Communication networks beyond the capacity crunch - further discussion Kavli Royal Society Centre, Chicheley Hall Newport Pagnell Buckinghamshire MK16 9JJ