Autonomous and Intelligent Healthcare System (SYSIASS) Activity 2 Progress June 2011 Part-financed by the European Regional Development Fund.

Autonomous and Intelligent Healthcare System (SYSIASS)

Activity 2 ProgressJune 2011

Part-financed by the European Regional Development Fund

Reminder of Activity 2 priorities

• Technology to address issues preventing remote healthcare

• Two major strands FSS: Utilise wireless communications technology via

Frequency Selective Surfaces ICmetrics: Secure data via encryption based directly

on system properties

Sub actions

Current Progress

Sub-action 1 commenced.

•ICmetrics• Collaborative visit Essex-Kent-ISEN 13/6/11• Paper submitted to EST 2011:Security Infrastructure for Autonomous and Intelligent Healthcare System

•FSS• General principles established• One invited publication to date

ICmetric Encryption

Encryptionmessage cipher

Key

Conventional encryption system

Encryptionmessage cipher

Key

Key Generator

ICmetricICmetric encryption system

Advantages

• No reference/template storage

• “No back door” : Another ICmetric sample required or break cipher

• System compromise does not compromise associated systems

• Circuit or software tampering (e.g. malware) implicitly prevented as ICmetric modified

• No encryption key storage

Goals for this action

Student Welfare Presentation – Department of ElectronicsPage 7

• Development of hardware and software instrumentation for feature collection

• Building an experimental platform for identification of candidate features

• Analysis of measured feature data

• Encryption key generation from normalised feature data

• Developing evaluation and calibration platform

Calibration Phase

• (applied once only per application domain)

• For each sample circuit: record a set of desired measurements associated with the circuit known generically as features.

• Generate feature distributions for each feature tabulating the frequency of each occurrence of each discrete value within the given value scale for each sample circuit.

• Normalise the feature distributions generating normalisation maps for each feature.

Operation Phase

• (applied each time an encryption key is desired for a given circuit)

• Measure features for the given circuit for which an encryption key is desired.

• Apply the normalisation maps to generate values suitable for key generation.

• Apply the key generation algorithm.

Unseen Devices

• Assumptions:- Distribution clusters identified in calibration sets Behaviours of unseen circuits combinations of

behaviours of seen circuits

• Analogy: Can read someone’s handwriting if unseen

previously but not if they write using different letter set such as Greek or Arabic

Multi-Cluster Distributions present

Unseen Devices

No of Samples Percentage Accuracy5 55.6%10 58%30 62%50 67.6%70 74%100 79.6%200 80.6%300 81%400 82.6%500 84.6%600 87%700 89.6%800 91%1000 92.6%

Summary

• Ability to employ ICmetrics in medical environments has potential Need to employ on unseen circuits essential The ability to offer secure communications

with no template or key storage

• Offering Generalised use of wireless technology Up-to-minute medical data at point of use Addressing security concerns