Intelligent Communications Research in the Electronic Engineering Department

1
Intelligent Communications Research in the
Electronic Engineering Department

• Applying advanced information communication and
  network management techniques to real world
  applications.
• Interdisciplinary research
• Maths, CS, EE, Eng
• Nov 5th meeting

2
We are unique! - we combine

• knowledge of intelligent systems
• worked on AI and agent systems for many years
• knowledge of how communication systems work
• know how IP, wireless networks, electrical
  systems work
• Normally tackle large-scale industrial or
  commercial problems where autonomous components,
  communication and distributed intelligence, give
  leverage
• information management, telecoms network
  management, tourism, environmental data
  management, security, e- commerce..

3
What is going on?

• NM, e.g.distributing intelligence through
  heterogeneous networks to support QoS and QoE

• Real time cooperative coverage in mobile and WLAN
  networks

• Location aware services

• Security and survivability of critical
  infrastructures

• Trust

• Semantic information retrieval

4
Projects

• Thematic network on electronic trust management
• Funded by EU

• customisable, component-based road maps to aid
  the early discovery of security vulnerabilities,
  inconsistencies and redundancies
• experimentation in the fields of e-commerce and
  tele-medicine.
• Funded by EU

• Security and survivability in large scale
  critical infrastructures
• Distributed intelligent components for anomaly
  detection and correlation of information
• Electricity telecom networks
• Funded by EU

• ADAMANT

quote-driven automated market as a means of
facilitating trading of real-time
products. Validate in in the telecommunications
domain for trading bandwidth EPSRC

• using a common interface to data mine the most
  complete set of information sources independent
  of physical or logical location of the databases,
  the database languages used, or of the specific
  nomenclature used in a particular database
• without knowing which database or databases
  contain the relevant information
• Funded by EU

In operation at Athens airport for the
Olympics Location based services over wireless
LANS Cooperative shaping of WLAN radiation
patterns for priority management in emergencies
and classes of service Funded by EU

• ONR

QoS in heterogeneous military networks Using MCDM
techniques to rank calls based on priorities and
context combat, logistics, Applies to
transport of multimedia in commercial networks
Funded by Office of Naval Research Washington

• TORRENT

QoS in residential access to multi-service
networks mapping service characteristics to
network performance parameters Funded by EU

• SAFEGUARD

• ITRUST

• CoRAS

• Quote-Driven Automated Market

• EDEN

5
People

• 5.5 Academics
• John Bigham Eliane Bodanese Rachel Bourne Laurie
  Cuthbert Stefan Poslad Eric Sharfe
• 9 RAs
• 1 Post-Doc visitor
• 12 RSs

6
Potential Synergies Cooperation

• Much of our work on agents in network management
  is a form of distributed decision support (CS and
  Maths?)
• in dynamic, non stationary environments
• e.g. utility can be lost in some data connections
• Relating conventional QoS to QoE (CS?)
• human factor considerations, multiple sensors,
  feedback control to guide transmission management
  , use of content context
• Using stored data, e.g. recommendation systems,
  on a large scale in a very dynamic context, to
  find and guide
• Work on trust, security on survivability (CS and
  Maths?)
• Uncertain reasoning with temporal dependencies,
  non monotonic logics
• Formalising notions of trust
• Novel anomaly detection mechanisms invariants
  and quasi invariants