Development of an intelligent system for operator support during wireless infrastructure system testing

摘要

Intense competition and the requirement to continually drive down costswithin a mature mobile telephone infrastructure market calls for new andinnovative solutions to process improvement. One particular challenge isto improve the quality and reliability of the diagnostic process for systemstesting of GSM and UMTS products. In this thesis, we concentrate on aparticularly important equipment type – the Base Transceiver Station(BTS). The BTS manages the radio channels and transfers signallinginformation to and from mobile stations (i.e. mobile phones). Most of thediagnostic processes are manually operated and rely heavily on individualoperators and technicians' knowledge for their performance. Hence, thereis a high cost associated with trouble-shooting in terms of time andmanpower. To address this issue, we employ Bayesian networks (BNs) tomodel the domain knowledge that comprises the operations of the SystemUnder Test (SUT), Automated Test Equipment (ATE) and the diagnosticskill of experienced engineers, in an attempt to enhance the efficiency andreliability of the diagnostic process. The proposed automated diagnostictool (known as Wisdom) consists of several modules. An intelligent user interface will provide possible solutions to test operators/ technicians; to capture their responses, and to activate the automatedtest programme. Server and client software architecture will be used tointegrate Wisdom with the ATE seamlessly and to maintain Wisdom as anindependent module. A local area network will provide the infrastructurefor managing and deploying the multimedia and text information in realtime. We describe how a diagnostic model can be developed andimplemented using a Bayesian network approach. We also describe howthe resulting process of diagnosis following failure, advice generation andsubsequent actions by the operator are handled interactively by theprototype system.