Posters Galore: 'Artificial Intelligence and Quantum computing in Wireless Networks: Within and Beyond'

Rishi Raj, Teoco

Rishi Raj, Teoco

Email

Wireless networks, such as 5G, is advancing to provide faster throughput and minimizing latency. Networks are getting more complex and it needs smarter ways to deliver an optimum capability of the network in reality. 

Therefore, it is critical to understand the need for Artificial Intelligence (AI) and Quantum computing (QC) within the Network Architecture. This is to provide the required capability for beam management, resource management, and interference management. 

The subsequent wireless network needs to be self-predictive and proactive to handle futuristic applications like holographic communication, haptic feedback, or any latency dependent application. Radio frequency itself is a dynamically changing component, therefore rather having fixed rules it should prefer to have a dynamically adaptable network. 

Today, the main challenge in the modern wireless network is to have constant knowledge of the radio environment - including multipath propagation of MIMO channels - especially with the introduction of Narrow beams in higher frequency (tracking of beams will be a challenge). But with an increasing number of beams and different kind of users - apart from mobile users like vehicle to vehicle communication, holographic communication, and haptic feedback - it will be very difficult just with the CSI (Channel State Information) and SRS (Sounding Reference Signal) reports constantly from the devices with changing radio environment. 

View the poster here
View more posters and videos

Posters Galore: 'GAIUS: Powering hyperlocal content ad ecosystems for communities'

Arjuna Sathiaseelan
GAIUS Networks

Arjuna Sathiaseelan, GAIUS Networks 
Email

The World Wide Web (WWW) is facing two critical issues:  
1. increasingly centralised with a few organisations controlling majority of users' data and hence global ad revenue.  
2. Centralised cloud infrastructures induces significant latency and performance issues for users in rural/remote areas with challenged backhaul - a misnomer for UK's 5G story.  

In this poster, I want to showcase the potential of building Decentralised Hyperlocal Web for communities in the UK using our GAIUS platform - a decentralised content ad exchange platform that can fully function at any edge. Each GAIUS server sits at the edge of a community (e.g. on the MEC) serving highly localised content and ads specific to a particular community. Users in a community can create, interact and transact with hyperlocal content through their mobile device. Businesses and organisations (e.g. schools, local government bodies etc) in a community can promote their product offerings and services within their own community. A major focal point of the GAIUS platform is that the data is governed and owned by the community.  

The GAIUS platform will demonstrate a new usecase for MEC: how operators can empower communities to create, interact and transact with hyperlocal content and ads at the operator's edge especially in rural areas. Apart from creating a new usecase (& potential revenue stream), GAIUS empowers operators to potentially fulfil their CSR requirements by supporting local community needs and facilitating local community engagement and social cohesion as well as addressing the major concern of data privacy 

View the poster here
View more posters and videos

Posters Galore: 'Deploying and scaling models in Data science'

Panagiotis Kourouklidis
University of York 

Panagiotis Kourouklidis, University of York
Email

In recent years, commoditisation of computing and digital storage resources has allowed businesses to extract insights from the data generated by their operations at a scale not witnessed before. Additionally, thanks to advances in the field of AI, a number of tasks that previously required human labour can now be automated. Of course, human labour is still needed in order to develop the system that is going to complete the aforementioned tasks.

The development of such a system starts with a data scientist producing a model, usually by using machine learning techniques and a training dataset. Unfortunately, the model produced by the data scientist is not the only component needed for a production-ready system that can run reliably. Usually, after the model has been produced it is passed on to a team of software engineers so that they build the rest of the system around it. Components of this kind of system include data gathering, cleaning and transformation as well as model deployment and performance monitoring.

The ongoing goal of the presented research project is to develop a framework that can streamline the development of the above components and can be used by a person not necessarily well versed in software engineering. That will allow data scientists to develop systems that are closer to being production-ready while reducing their reliance on software engineering teams.

View the poster here
View more posters and videos

Posters Galore: 'Ensemble-Based Data Relationship Discovery Framework in Supply Chain Business Processes'

Akinola Ogunsemi
Robert Gordon University

Akinola Ogunsemi, Robert Gordon University
Email

This project represents the initial phase of developing a holistic, data driven formulation of service and supply chain optimisation problem. This is a prime prerequisite into understanding links between billions of data records generated from supply chain business processes.

We performed an investigation of how eight data relationship discovery algorithms can be combined to identify more comprehensive links between database tables. This could be used in providing database users/analysts with the opportunity to understand data relationships and the ability to extract insights from data for commercial advantages like reducing time-intensive analysis and exploration of data by domain experts.

We performed comparative analysis to rank our individual discovery algorithms, the combination framework and selected state-of-the-art algorithms to evaluate performance and suitability. Preliminary results are indicative that using an appropriate combined strategy of discovery techniques can best provide a comprehensive and more supported discovery of data linkages in supply chain business processes.

View the poster here
View more posters and videos

Posters Galore: 'Improved detection of nuisance calls in-network using machine learning techniques'

Matthew Middlehurst
University of East Anglia

Matthew Middlehurst, University of East Anglia
Email

Nuisance marketing, nuisance sales and fraud calls - generalised as nuisance calls - are a problem which has rapidly expanded in recent years. Yearly, tens of thousands of people lose money to telephone scams and a significantly larger amount deal with the annoyance of spam from call centres.

Previous studies into automated detection of these calls have largely neglected the presence of legitimate call centres. These cases have similar characteristics to nuisance call centres and can cause great harm if incorrectly blocked. 

For a range of calling line identities used by legitimate and nuisance call centres, we track the outgoing call volume for a single day. In recent years, time series classification (TSC) has had a rapid advance in predictive power. Using this time series data we attempt to classify cases as nuisance or legitimate using a variety of TSC techniques. 

We show that the shapelet transform classifier and the time series meta-ensemble HIVE-COTE can differentiate between these centres reasonably well, but may not reach the specificity required for real world application on their own. A further look into cases misclassified as nuisance shows a lot of cases to have patterns resembling robotic callers. This leads us to doubt the certainty of the provided labels, and suspect the classifiers performed better than initially presented. Further analysing the results, we describe the most informative parts of the series for discriminating between legitimate and nuisance using the inherent interpretability provided by shapelets.

View the poster here
View more posters and videos

Posters Galore: 'Detecting Emergent Phenomena within Streaming Data'

Edward Austin
Lancaster University

Edward Austin, Lancaster University

Email

Throughput data is a measure of the volume of internet traffic passing through points on the BT network. Due to consumer behaviour over the course of a day, this data takes an expected shape; put simply, less data is streamed in the early hours of the morning, than in the middle of the working day. Deviations from this expected shape, however, are indicators of faults, outages, or unexpected customer behaviour taking place on the network. The rapid detection of these deviations is critical as it allows maintenance to be performed to fix the problem. This benefits the customer, and in turn enhanced BT revenue and reputation. 

This work explores how mathematical tools can be used to model daily throughput data with smooth curves, and then how differential equations can be used to model the expected shape the throughput data takes. A test is then presented for the detection of deviations from this shape in real time. 

View the poster here
View more posters and videos

Thought Leadership Event: Dr Fumiya Iida on "Agri-Food Robot Revolution"

Thought Leadership Event
Speaker: Dr Fumiya Iida, Bio-inspired Robotics Laboratory, University of Cambridge
Title: Agri-Food Robot Revolution

Agriculture and food supply is the largest industry in the UK, but it faces significant challenges of transformation needs because of wage increase, geopolitical uncertainties (Brexit) and health and safety concerns (COVID-19). 

Despite the increasing demands, the automation of processes that are currently being performed by human labourers is not trivial for many reasons, not only the cost and the efforts, but also the fundamental lack of technologies.

In this talk, FumiyaI introduced the team’s efforts to analyse the landscape of challenges in the agri-food sectors from the technologist’s viewpoint and discussed what the problems that we can solve today are, as well as the challenges for the future.