The　high　penetration　of　renewable　energy　sources　and　the　consequent　integration　of　such　distributed　energy　generation　systems　into　the　grid　have　significantly　increased　the　interaction　between　power　sources,　which　can　result　in　low　frequency　oscillations.　Different　control　techniques　are　proposed　in　literature　to　suppress　low　frequency　oscillations,　but　the　first　step　to　attenuate　such　oscillations　is　to　characterize　them　in　terms　of　their　dominant　modes.　This　paper　combines　different　results　from　literature　to　propose　a　unified　two-step　methodology　for　the　mode　characterization　of　low　frequency　oscillations　based　on　the　signals　acquired　by　wide　area　measurement　systems.　Since　the　measured　signals　typically　contain　noise,　the　first　stage　of　the　proposed　method　uses　the　basis　pursuit　denoising　method　combined　with　a　Tunable　Q-factor　wavelet　transform　to　increase　the　signal-to-noise　ratio　and　improve　the　identification　results.　In　a　second　stage,　an　improved　version　of　the　matrix　pencil　algorithm,　as　proposed　in　this　paper,　is　used　to　identify　the　parameters　of　low　frequency　oscillation　dominant　modes.　Both　simulation　and　experimental　results　show　that　the　proposed　method　has　better　characterization　accuracy　than　traditional　methods,　especially　when　Gaussian　noise　is　considered　in　the　measurements.　In　addition,　the　processing　time　has　proven　to　be　reasonable　for　online　identification　and　characterization　of　low　frequency　oscillations.