|
|
|
|
|
(This page is aimed at non-technical people who don't have access to the intranet where Gray works.)
This is Gray's work web page. If you wanted personal information you'd need his home web page.
Gray has recently taken up a post at GlobespanVirata based in the Science Park on the outskirts of Cambridge.
Prior to that Gray worked at the only research laboratory for AT&T (the American telegraph and telecommunications company) that operates outside the USA. It was called AT&T Laboratories Cambridge Ltd and was found just opposite the Fitzwilliam Museum in the heart of Cambridge (UK).
Sadly AT&T moved to close this laboratory down on 24 April 2002.This is what you can find below:
As part of the Consumer Broadband Project he worked, in liaison with a small group in AT&T Research, Florham Park , to provide an operating system for the cheap network "switches" that are needed to demonstrate the feasibility of connecting the internet to your home with a high enough speed and quality of service to provide your television that way.
He was responsible, in particular, for making the operating system " Linux" work on a MIPS processor chip nicknamed a Banyan which, as well as executing programs, can use "Asynchronous Transmission Mode" ( ATM ) connections. Such connections are popular inside telecommunications networks because they are fast, reliable and basically simple. He wrote the part of Linux that is needed to use the two or three ATM interfaces on a Banyan.
ATM technology is not quite as pervasive as the Ethernet, which almost all office networks use. Lately most of Gray's time using some software from Columbia University, called Netbind, which enables some of the latest Intel "network processor " chips (in particular the IXP 1200) to be reconfigured to support different kinds of fast Ethernet connections quickly.
He also helps supervise and examine students from time to time and is currently Andy Hopper 's proxy supervisor for Alastair Berisford , who is undertaking a security-related Ph.D. in the Laboratory for Communications Engineering in the Department of Engineering of Cambridge University.
Most recently Gray was involved in the, now quiescent, Prototype Embedded Networks ( PEN ) project which sought to provide a platform of programs and tiny radio-computers that could be used to assemble small networks to sense and control their environment. The radios developed were very low power and ideally (although that ideal was never met with the working hardware) would operate for years, for example embedded in a washing machine or a wall. He designed and built a small operating system (the Embedded Environment Kernel, EEK ) that allowed small modules of program code to be easily collected together for one of the radio-computer nodes. Although unsuitable for the current type of node, he also specified and implemented a programming language "Quiver" that could be written by people or by the PEN computers so that they could send programs to each other. Quiver would allow battery power to be saved by sending programs to nodes nearer power sources (such as those plugged in to a wall socket). He also helped devise the communication rules (protocol) used by radios which allowed them to conserve their batteries whilst they talked to each other.
A forerunner to EEK was used in the Radio ATM project, which has "spun-out" to become first Active Broadband Ltd and subsequently Cambridge Broadband Ltd .
Expertise in ATM was exported in a previous spin-out, from the ATMos project, to a company at first called ATM Ltd and then Virata Ltd . Following their spin-out some further development of the operating system ("ATMos") used in the project's equipment was necessary. Gray remade the support ATMos had for the communication rules that characterize the Internet (the protocols called TCP/IP) and extended the support so that they would work over connections that were not ATM based. He implemented some other related protocols to help the computer find communication routes to other computers and made it operate faster. ATMos went back to Virata (and has now evolved into an operating system called ISOS) and the protocols (developed further by them) are an important part of their product line.
One of the first things Virata made was a Network Interface Card (NIC), which could be inserted into a PC to allow it to use ATM connections. Gray wrote the program that allowed this NIC to be used from the Linux operating system.
As part of the Medusa project many small computers using ATM were connected together: each providing one microphone, one video display, one loud speaker, or some other part of a distributed "multimedia" system. For some time Gray was responsible for maintaining this network of small ATM computers. He built a system that would let people "debug" programs they ran even though the users themselves were using a PC in a different part of the building.
You're obviously a very inquisitive reader to have got this far...
Clearly, Gray is older than he looks because he graduated from the department of computing at, what was then, the Imperial College of Science and Technology in 1978. His thesis at the Cambridge University Computer Laboratory was all about how to prove who you are, when you use all kinds of different services provided by computers on a network, without giving your password out every time, and without using encipherment. One of his claims to fame from this period was the adoption of his ideas by the X consortium who have use "cookies" in their ubiquitous windowing system after a meeting with Paul Karger, a security contributor to X from MIT.
At Acorn, who made the BBC microprocessor a household brand in the UK and the then Acorn RISC Machine (ARM) well-known more globally, Gray worked on a before-its-time (and doomed) project to use the BBC micro for a new style of email that would have carried pictures and speech as easily as words of text.
For seven years following that Gray worked as part of a small computer network security group at another of the "Cambridge Phenomenon" companies, Topexpress, which changed hands to end its days as a part of the acquisitive SAIC group of companies, while the security group, itself, was acquired by Perihelion Software. The work there was almost entirely for a single client, the Communications and Electronics Security Group at the Government Communications Headquarters ( GCHQ) "British Intelligence". It involved a wide variety of research from developing mathematical models of confidentiality, providing a framework and a formal method for constructing networks of computers that you might be able to "prove" were secure, and investigating how easy it is for spys to use a network to pass messages to each other covertly. (His paper about the latter is one of the few referred to by a publication describing the security criteria to rate the quality of computer networks, then called the "Red Book", which was issued by the forerunner of the current US National Security Agency).
During this period Gray had an increasing involvement in international standards, convening the British Standard Institute's (BSI's) network security architecture group and representing the UK at the relevant working group of the right sub-committee of the joint technical committee of the International Standards Organization (ISO) and the International Telecommunication Union (ITU)! The resulting standards can now be read in ISO 7498-2 , the security part of the "seven layer model" used throughout the standards community to position communication standards, and in ISO 1018 , the "security frameworks", a seven part standard providing fixed names and ideas relevant to six important security services. He was also involved to a lesser degree in many other standards, e.g. the UK and European versions of the "Red Book".
Much of the above standardization started by considering (only) two computers that were communicating with each other. A newer area of communications standardization focussing on many computers interacting in order to accomplish some distributed processing task was started in the mid-80s and is called Open Distributed Processing (ODP). Most of the ISO standards in this area derive directly from work in the UK Advanced Network System Architecture (ANSA) project which solidified into a company, Architecture Projects Management (APM) which Gray joined in 1992. APM had both a "business" and "research" side.
In the business part of the company he contributed to the ODP security standards but also became involved in some European funded projects and a Defence Research Agency (DRA) analysis of existing security architectures. He contributing to and edited a Department of Trade and Industry (DTI) report about testing conformance to ODP standards. The Object Management Group (OMG) provide a specification for most popular ODP system: The Common Object Request Broker Architecture (CORBA). Another of Gray's tasks was to mediate the contribution of OMG's director, Richard Soley to the DTI report.
In the research part of the company, as part of work towards making it easier to generate a federation of computers, he was involved in the design of a new kind of "trader". Traders are components in ODP systems used to find other computers that will do the kind of work you want done. This one, "the property repository", would have used a database to enable components to be located by their attributes in more general ways than previous traders allowed.
4 Stansfield Gardens
Fulbourn
Cambridge
CB1 5HX
UK
Tel: +44 1223 561817
Email: cgg@cantab.net