Automatic Inference of Cloud Application Performance Degradation

In order for a cloud operator to make effective decisions with respect to horizontal scaling, migration, and root cause analysis, current cloud orchestration tools often need to monitor application or service-level metrics, such as latency (response time) or throughput. The interpretation of these metrics often requires significant experience (and expertise) ...

An Efficient Model for Mobile Network Slice Embedding under Resource Uncertainty

The fifth generation (5G) of mobile networks will support several new use cases, like the Internet of Things (IoT), massive Machine Type Communication (mMTC) and Ultra- Reliable and Low Latency Communication (URLLC) as well as significant improvements of the conventional Mobile Broadband (MBB) use case. End-to-end network slicing is a ...

Path Planning Problems with Side Observations- When Colonels Play Hide-and-Seek

Resource allocation games such as the famous Colonel Blotto (CB) and Hide-and-Seek (HS) games are often used to model a large variety of practical problems, but only in their one-shot versions. Indeed, due to their extremely large strategy space, it remains an open question how one can efficiently learn in ...

Combinatorics of nondeterministic walks of the Dyck and Motzkin type

This paper introduces nondeterministic walks, a new variant of one-dimensional discrete walks. At each step, a nondeterministic walk draws a random set of steps from a predefined set of sets and explores all possible extensions in parallel. We introduce our new model on Dyck steps with the nondeterministic step set ...

Symbolic Method and Directed Graph Enumeration

We introduce the arrow product, a new generating function technique for directed graph enumeration. It provides new short proofs for previous results of Gessel on the number of directed acyclic graphs and of Liskovets, Robinson and Wright on the number of strongly connected directed graphs. We also obtain new enumerative ...

On Hyper-local Conversational Agents in Urban Settings

In the last decade, a large number of virtual assistants have been offered to the market. These conversational agents use vocal commands from the users to perform various personalized tasks that are constrained by the applications a user installs on her devices. We consider a brand new conversational agent in ...

An Early Characterisation of Wearing Variability onMotion Signals for Wearables

We explore a new variability observed in motion signals acquired from modern wearables. Wearing Variability refers to the variations of the device orientation and placement across wearing events. We collect the accelerometer data on a smartwatch and an earbud and analyse how motion signals change due to the wearing variability. ...

RSRP Fingerprinting and Deep Learning Assisted UE Positioning in 5G

In this work, we investigate user equipment (UE) positioning assisted by deep learning (DL) in 5G and beyond networks. As compared to state of the art positioning algorithms used in today's networks, radio signal fingerprinting and machine learning (ML) assisted positioning requires smaller additional feedback overhead; and the positioning estimates ...

Benefits of software testing using the Scrum framework. How to avoid the Cargo Cult?

Short analysis of diffrent approaches to the small project/ team management.

A Systematic Study of Unsupervised Domain Adaptation for Robust Human-Activity Recognition

Wearable sensors are increasingly becoming the primary interface for monitoring human activities. However, in order to scale human activity recognition (HAR) using wearable sensors to million of users and devices, it is imperative that HAR computational models are robust against real-world heterogeneity in inertial sensor data. In this paper, we ...

Machine Learning Data Sets for Cross-Layer and End-to-End Networks

The emergence of cloud computing has had a profound impact on our lives, enabling for example, scalable compute and storage for enterprises, and cloud-based personal digital assistants based on artificial intelligence (AI) in homes. While considerable investment has gone into creating on-demand cloud-compute resources, the flexibility of the underlying network, ...

Detection-Localisation-Identification of Vibrations over Long Distance SSMF using Coherent Phase-OTDR

We propose a novel interrogation technique for Distributed Acoustic Sensing (DAS) and demonstrate the detection and recognition of multiple vibration events over 50km of SSMF. A differential-phase optical time-domain reflectometry (diff-Phi OTDR) approach is used with the transmission of a continuous laser modulated with polarization-multiplexed orthogonal binary sequences and the ...

DCI Field Trial demonstrating 1.3-Tb/s Single-Channel and 50.8-Tb/s WDM Transmission Capacity

We demonstrate, to the best of our knowledge, the highest reported single- and multi-channel capacities over a field-deployed data center interconnection (DCI) system operating over 93 km of standard single-mode fiber using only Erbium-doped fiber amplification. Using off-the-shelf hardware components, and in particular using only a single digital-to-analog converter per ...

A 112-GS/s 1-to-4 ADC front-end with more than 36-dBc SFDR and 28-dB SNDR up to 43-GHz in 130-nm SiGe BiCMOS

Voltage-mode sampling circuits operating at mm-wave frequen-cies require clocks with root-mean-squared (RMS) jitter in the two-digit femtosecond range for high SNR performance. The generation of such clean clocks, however, presents a serious de-sign challenge. To mitigate this problem, an alternative sampling approach based on charge sampling is proposed. For the ...

Pre- and post-transmission digital signal processing using neural networks applied to fiber-optic communication systems relying on power modulation and power detection : Advantages and drawbacks

For ML Photonica 2019: https://ml-photonica2019.astonphotonics.uk/ In this presentation we explain the nonlinearities and complexity of transceivers relying on power modulation and detection. We detail the challenges for reliable communication through this type of channel when trying to deliver high bitrates over a few tens of kilometers.

VIBRATION IDENTIFICATION OVER 50KM SSMF WITH POL-MUX CODED PHASE-OTDR

We demonstrate multi-detection and recognition of mechanical events over 50km of SSMF with coherent homodyne phase-OTDR using novel Pol-Mux orthogonal interrogation codes. Pattern identification of an engine noise is highlighted, paving the way for numerous smart-city/industrial monitoring applications over already deployed telecom fibers.

Beyond 100 GBaud Transmission Supported by a 120 GSa/s CMOS Digital to Analog Converter

We demonstrate for the first time a 105 GBaud transmission using a CMOS digital to analog converter. By variable spectral efficiency by probabilistically shaped constellations we assess the maximal achievable information rate of 4 bit per symbol and demonstrate successful transmission of a 750 Gb/s channel over xx km.

SUPPLY-POWER-CONSTRAINED CABLE CAPACITY MAXIMIZATION USING DEEP NEURAL NETWORKS

We experimentally demonstrate a 19% capacity gain per Watt of electrical supply power in a 12-span link by eliminating gain flattening filters and optimizing the launch power profile using machine learning by deep neural networks in a massively parallel fiber context.

Efficient Multi-Event localization from Rayleigh backscatter in phase sensitive OTDR

We introduce a phase-OTDR multi-resolution approach to detect and localize mechanical events in standard telecom fibers. Perturbations are first roughly localized from a subset of high intensity backscatters, followed by a low complexity localization refinement

1.11 Tb/s/lambda and 18.9 Tb/s WDM transmission over DCI distances supported by CMOS DACs

We report on the optimization of a CMOS DAC transmitter based coherent system supporting bitrates beyond 1 Tb/s. We demonstrate WDM transmission over DCI distances up to 118 km at spectral efficiencies of 9.9 bit/s/Hz