Department of Mathematics & Computer Science

MC397 Programming Secure and Distributed Systems

Explanation of Pre-requisites

A significant aspect of the module will be the reinforcement of material delivered in lectures with practicals involving students implementing distributed systems in Java. Hence a basic knowledge of Java, as provided in MC103 and MC104, will be essential. A basic grounding in discrete mathematics will assumed during the lectures on security.

Although not essential, students will benefit from prior knowledge of SQL as obtained from MC115 when integrating databases into distributed systems. Similarly, software design as found in MC206 will aid the design of distributed systems while the close relationship between distributed computing and concurrency means that knowledge of MC307 will also be of benefit.

Course Description

The internet has caused a revolution in the way we use and think about computers. One of the key technologies underlying the internet is distributed computing which allows individual computing agents to be located, or distributed, on a network of computers but nevertheless work together on cooperative tasks. There are many motivations for distributing applications including:

• Inherent Parallelism: In distributed systems, many computers can work together at the solution of a problem simultaneously. Although this can be done using parallel computers, it is often cheaper to distribute applications over cheaper networks of PCs.
• Inherent Distribution: In many applications, distribution arises naturally. For example, in a traffic control system, data may be gathered at different measuring facilities, analysed at a central location and, depending upon the results of this analysis, commands sent back to improve the flow of traffic.
• Scalability: Distributed systems are easier to scale than centralised systems. For example, increasing computing power of a centralised system to cope with increasing demand is usually much more expensive than increasing the number of computer nodes in a distributed system.
• Data Access: It is often difficult and expensive to move large data sets between different locations. Instead, it can be more efficient to store data in locally and allow other users to remote access using a data server.
• Resource Sharing: In distributed systems, expensive resources such as printers can be used by devices distributed across a network. Again, this sharing can produce significant cost savings.

The development and programming of distributed systems is considerably more complex than that of local applications. Distributed applications usually run on spatially separated computers, consist of different components which need to communicate with each other and must ensure a consistent management of shared resources. In addition, of increasing importance is the issue of security in distributed systems, eg how can a component of a system detect security breaches elsewhere in the system. This course will introduce these and other fundamental problems in distributed computing, explain some of the solutions available and cover their implementation in the Java programming language. Students will thus gain insight into distributed computing as well as practical skills of immediate use.

Aims

Objectives

• To build simple distributed applications using Java's networking and remote method invocation capabilities.
• To build concurrent distributed applications using multiple threads.
• To build distributed applications with security enhancements using Java's security and cryptographic extensions.

Transferable Skills

• To understand the limitations of the basic model of computation as provided in elementary programming courses and the consequent development of more sophisticated programming paradigms such as distributed, concurrent, and secure programming.
• To use the theoretical perspectives offered by this module to implement distributed applications in other programming languages.
• To further develop design and implementational skills as taught during the degree.
• Students who master the material offered will be prepared not only to write distributed applications or to read the research literature, but also to evaluate systems, algorithms and languages from a broad perspective.

Syllabus

• Distributed Programming: Networking in Java using sockets and streams, multi-threaded computation, liveness, safety, deadlock, livelock, mutual exclusion, communication protocols, semaphores and monitors, synchronous and asynchronous message passing in Java, client-server architectures.
• Security: Security issues and concerns. Key management including generation, translation, agreement protocols and management paradigms. Authentication including message digests, MAC's, signatures, certificates. Symmeteric and antisymmetric ciphers, eg DES, IDEA, RSA.
• Java Support: Java security, Java Cryptography Architecture (JCA), Java Cryptography Extension (JCE), Database integration using Java Database Connectivity (JDBC) and SQL, Java remote method invocation (RMI).

Reading list

Essential:

J. Farley, Java Distributed Computing, O'Reilly. 1998..

S. Hartley, Concurrent Programming, The Java Programming Language, OUP. 1998..

D. Lea, Concurrent Programming in Java: Design Priciples and Patterns, Addison-Wesley. 1997..

Recommended:

S. Oaks, Java Security, O'Reilly. 1999..

M. Boger, Java in Distributed Systems, Wiley. 2001..

Background:

Details of Assessment

The written May/June examination contains six questions, and candidates can obtain full marks from four questions.

Author: S. J. Ambler, tel: +44 (0)116 252 3884
Last updated: 2001-09-20
MCS Web Maintainer
This document has been approved by the Head of Department.
© University of Leicester.