Method for noise energy estimation
AssigneeNokia Corporation
InventorAki Happonen
Life of this patent
Estimating the term…
Abstract
A signal processing method is described for processing a signal transmitted via a transmission channel. The method comprising the steps of: receiving (S11) a plurality of samples of the signal; analyzing (S12) the amplitude of a subset of the received samples; and estimating (S13) the variance of the signal based on the amplitude of the subset of the received samples. A corresponding signal processing device is also described. Thus, with the present invention it is possible to estimate the variance on the basis of the amplitude of the received signal. In consequence, no information about channel impulse response or training sequence is required.
Claims12
A signal processing method for processing a signal having been transmitted via a transmission channel, the method comprising the steps of:
- receiving (S11) a plurality of samples of said signal;
- analyzing (S12) the amplitude of a subset of said plurality of received samples;
- and
- estimating (S13) the variance of said signal based on the amplitude of said subset of said plurality of received samples,
- wherein said subset of said plurality of samples is defined as a number (N) of samples within an interval ((x:y)) of the plurality of received signal samples, with x being a first sample of said subset and y being a N sample of said subset,
- wherein said step of analyzing (S12) the amplitude comprises
- a first detecting step (S21) of detecting an absolute maximum (refi) among imaginary parts of said number (N) of signal samples, and
- a second detecting step (S22) of detecting an absolute maximum (refq) among real parts of said number (N) of signal samples,
- wherein in said first detecting step (S121), said absolute maximum (refi) among imaginary parts is detected according to a relation refi=max(|Idata(x:y)|), and
- in said second detecting step (S122), said absolute maximum (refq) among real parts is detected according to a relation refq=max(|Qdata(x:y)|), wherein Idata represents an imaginary part of a received sample, and Qdata represents a real part of a received sample, and
- wherein in said estimating step (S13) said variance (var) is estimated according to a relation
- 5 var = A * var p = A * 1 / N * ∑ k = 1 N ( ( refi - [LeftBracketingBar] Idata ( x :
- y ) [RightBracketingBar] ) 2 + ( refq - [LeftBracketingBar] Qdata ( x :
- y ) [RightBracketingBar] ) 2 )
- where Idata(x:y) and Qdata(x:y), respectively, denote a respective imaginary/real component of an kth sample within the subset of N samples in which x represents the first sample and y represents the Nth sample, and varp represents a preliminary variance to be scaled by a scaling factor A.
A signal processing device (30) for processing a signal having been transmitted via a transmission channel (31), the device comprising:
- receiving means (32,33) adapted to receive a plurality of samples of said signal;
- analyzing means (34, 34a, 34b) adapted to analyze the amplitude of a subset of said plurality of received samples;
- and
- estimating means (35, 35a) adapted to estimate the variance of said signal based on the amplitude of said subset of said plurality of received samples,
- wherein said subset of said plurality of samples is defined as a number (N) of samples within an interval ((x:y)) of the plurality of received signal samples, with x being a first sample of said subset and y being a Nth sample of said subset, said subset being buffered in a buffer means (33),
- wherein said analyzing means (34) further comprises
- a first analyzing element (34a) adapted to detect an absolute maximum (refi) among imaginary parts of said number (N) of signal samples, and
- a second analyzing element (34b) adapted to detect an absolute maximum (refq) among real parts of said number (N) of signal samples,
- wherein said first analyzing element (34a) is adapted to detect said absolute maximum (refi) among imaginary parts according to a relation refi=max(|Idata(x:y)|), and
- said second analyzing element (34b)is adapted to detect said absolute maximum (refq) among real parts according to a relation refq=max(|Qdata(x:y)|), wherein Idata represents an imaginary part of a received sample, and Qdata represents a real part of a received sample, and
- wherein said estimation means (35) further comprises calculation elements (35a), said calculation element being configured such that said variance (var) is estimated according to a relation
- 6 var = A * var p = A * 1 / N * ∑ k = 1 N ( ( refi - [LeftBracketingBar] Idata ( x :
- y ) [RightBracketingBar] ) 2 + ( refq - [LeftBracketingBar] Qdata ( x :
- y ) [RightBracketingBar] ) 2 )
- where Idata(x:y) and Qdata(x:y), respectively, denote a respective imaginary/real component of an kth sample within the subset of N samples in which x represents the first sample and y represents the Nth sample, and varp represents a preliminary variance to be scaled by a scaling factor A.
Dependencies are inferred from the claim text.