Module Title:   Networks Performance Modelling and Analysis

Module Credit:   20

Module Code:   CM-0602D

Academic Year:   2015/6

Teaching Period:   PG Computing Summer Semester

Module Occurrence:   A

Module Level:   FHEQ Level 7

Module Type:   Standard module

Provider:   Computer Science

Related Department/Subject Area:   SCIM (Dept of Computer Science)

Principal Co-ordinator:   Prof D D Kouvatsos

Additional Tutor(s):   Prof. M. E. Woodward

Prerequisite(s):   None

Corequisite(s):   None

Aims:
To define basic performance measures of computer systems and communication systems (e.g. throughput and response time). To describe performance laws and determine operational relationships. To introduce quantitative techniques for solving queuing network models (QNMs) and develop algorithms.

Learning Teaching & Assessment Strategy:
The delivery of the module will consist of lectures and tutorials and will be assessed by coursework and examinations.

Lectures:   24.00          Directed Study:   157.00           
Seminars/Tutorials:   16.00          Other:   0.00           
Laboratory/Practical:   0.00          Formal Exams:   3.00          Total:   200.00

On successful completion of this module you will be able to...

Mastery of quantitative techniques for the design, development and capacity planning of computing and communication systems.

On successful completion of this module you will be able to...

Model and evaluate system/network performance and predict the effects of changes in operational parameters on system/network behaviour.

On successful completion of this module you will be able to...

Ability to analyse performance and plan capacity in computer systems.

  Coursework   20%
 
  Practical problems on queuing theory and systems performance (equivalent to 1000 words)
  Examination - closed book 3.00 80%
 
  Examination
  Examination - closed book 3.00 100%
 
  Supplementary examination

Outline Syllabus:
Stochastic Processes: Poisson Process, Birth & Death Process, Markov Chains. Queueung Theory and Building Blocks: Birth & Death queuing models (M/M/1, M/M/c, M/M/1/N, M/M/c/N, M/M/1/N/N). Maximum Entropy and polling systems. Finite capacity queues: Two stage cyclic queuing models. The central server model. Maximum entropy algorithms for the analysis of QNMs with multiple job cases and blocking mechanisms. Applications: Performance modelling and evaluation of multiprogramming systems, multiprocessor systems, concurrency control of database systems, complex Input/Output (I/O) subsystems, scheduling algorithms of operating systems, end-to-end delay analysis and congestion control in high speed telecommunication networks.

Version No:  3