Estimation of Integer Carrier Frequency Offset in OFDM Systems Based on the Maximum Likelihood Principle

Toumpakaris, Dimitris; Lee, Jung Won; Lou, Hui‐Ling

doi:10.1109/tbc.2009.2013754

Cited by 30 publications

(29 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We first briefly describe three sets of existing pilot-aided estimators (PAEs) and blind estimators (BEs), namely, T. Schmidl's (TS) PAE [7] and BE [21], M. Morelli's (MM) PAE and BE [8] and D. Toumpakaris's (DT) PAE and BE [22], and then study the impact of the accurate signal model on them.…”

Section: B Effects On Existing Icfo Estimatorsmentioning

confidence: 99%

“…Their estimators use the following correlation between two received OFDM symbols: (14) where 1) Existing PAE: Three existing PAEs will be briefly described. DT's PAE in [22] reads as (15) where is the pilot-based metric calculated as (16) and is the data-based metric calculated as (17) where and mean the real part and the imaginary part of the enclosed quantity, respectively. In practical OFDM systems (e.g., [23], [24]), the power of a pilot subcarrier is set times larger than the average power of a data subcarrier in order to aid reliable synchronization.…”

Section: B Effects On Existing Icfo Estimatorsmentioning

confidence: 99%

“…, (18) and the DT's PAE is modified to (19) where the energy correction factor equals (20) When signal to noise power ratio (SNR) is large enough and the system uses TYPIST modulation, [22] simplifies as…”

Section: B Effects On Existing Icfo Estimatorsmentioning

confidence: 99%

See 2 more Smart Citations

New Integer Normalized Carrier Frequency Offset Estimators

Zhan¹,

Minn²

2015

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

Carrier frequency offset (CFO) introduces spectrum misalignment of the transmit and receive filters, causing energy loss and distortion of received signal. A recent study presents an accurate signal model that incorporates these characteristics. However, CFO estimators in the existing literature were developed based on the signal models without these characteristics. This paper generalizes the accurate signal model with CFO to include timing offset and sampling offset and then studies what effects the accurate signal model bring to the CFO estimation and how to address them. We investigate different existing CFO estimators and find some of them lose their optimality under the accurate signal model. We also develop two new integer normalized CFO (ICFO) estimators under the accurate signal model using only one OFDM symbol; one of them is a pilot-aided estimator using both pilot and data and the other one is a blind estimator using data only. Compared with the existing approaches, both of the new estimators make better use of channel and energy loss information, thus yielding more accurate estimation. Furthermore, the proposed methods overcome the existing approaches' limitation to phase shift keying modulation. Simulation results corroborate the advantages of the new estimators. Analytical performance results are also provided for the proposed estimators.

show abstract

Section: B Effects On Existing Icfo Estimatorsmentioning

confidence: 99%

Section: B Effects On Existing Icfo Estimatorsmentioning

confidence: 99%

See 1 more Smart Citation

New Integer Normalized Carrier Frequency Offset Estimators

Zhan¹,

Minn²

2015

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

show abstract

“…However, it is sensitive to receiver synchronisation errors such as carrier frequency offset (CFO) which causes inter-carrier interference (ICI), and which is exacerbated by issues such as motion-induced Doppler shifts and local oscillator instability [1]. Large CFO is split into fractional (FFO) and integer (IFO) offsets for estimation [2], [3], [4], [5], [6]. IFO causes a circular shift of the subcarrier in the frequency domain while FFO results in ICI due to lost orthogonality between subcarriers.…”

Section: Introduction and Related Workmentioning

confidence: 99%

“…The technique is simple, but its performance degrades significantly in frequency selective channels. Maximum Likelihood Estimation (MLE) has been used to estimate IFO [5], [6], based upon the observation of pilots from two consecutive OFDM symbols. A reduced complexity alternative is to only compute over one preamble OFDM symbol, using differential encoding among adjacent subcarriers and correlation estimation [2], [3].…”

Section: Introduction and Related Workmentioning

confidence: 99%