Prognostic Health Management of a RF Transceiver Chain

Abstract

The impact that modern wireless communications plays in our everyday lives is huge. A great deal of our day to day operation, be it commercial, civil or defence relies on the ability to deploy these technologies reliably, with the minimum of failure. This paper presents a system based Prognostic Health Management (PHM) methodology for health monitoring and prognostics within a RF transceiver chain, with the ability to provide impending detection of individual system component failure. The primary focus of this paper is towards military assets such as radar and electronic warfare (EW) systems, but it is envisaged that it will have other connotations in helping to provide quality of service (QoS) in wireless communications as well as civilian radar systems. Methods of failure modes and degradation in RF devices are examined before suggesting a methodology where parametric data may be extracted for health monitoring by non-linear techniques. This data is then compared to an ideal healthy transceiver by the use of a metric employing the Mahalanobis distance. A prognostic is applied in the form a sequential Monte-Carlo method by the use of a particle filter to give an approximation of remaining useful life (RUL).