Frequency-domain AoA estimation and beamforming with wideband hybrid arrays

Xiaojing Huang*, Y. Jay Guo

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

44 Citations (Scopus)


High speed long range millimetre-wave (mm-wave) links can be achieved by using wideband hybrid antenna arrays of sub-arrays. Due to the array architecture difference, conventional wideband angle-of-arrival (AoA) estimation and beamforming techniques are not applicable to such wideband hybrid arrays. Targeted at point-to-point line-of-sight wireless transmission in the 70/80 GHz E bands, a unified frequency-domain AoA estimation and beamforming algorithm suitable for large scale wideband hybrid arrays of both interleaved and side-by-side sub-arrays is proposed in this paper. The AoA estimation performance is analyzed by deriving a recursive modified Cramr-Rao bound (MCRB). The effect of mutual coupling among antenna elements on the estimation performance is also considered for the hybrid array of side-by-side sub-arrays. The analytical results can be used to determine system parameters according to required system specifications. Simulation results show that the proposed AoA estimation algorithm is robust against practical impairments, and the frequency dependency of the array pattern is significantly reduced after digital beamforming. Simulated mean square errors of AoA estimation are also compared with the analytical bounds, showing that the derived recursive MCRB provides a meaningful indication to the AoA estimation performance.

Original languageEnglish
Article number5934684
Pages (from-to)2543-2553
Number of pages11
JournalIEEE Transactions on Wireless Communications
Issue number8
Publication statusPublished - Aug 2011
Externally publishedYes


  • and beamforming
  • angle-of-arrival estimation
  • hybrid antenna array
  • Millimetre-wave (mm-wave) communications
  • sub-array


Dive into the research topics of 'Frequency-domain AoA estimation and beamforming with wideband hybrid arrays'. Together they form a unique fingerprint.

Cite this