Process for detecting the nature of data-falsifying interference
AssigneeSiemens Aktiengesellschaft
InventorsHans-Werner Arweiler, Andreas Wolf
Life of this patent
Estimating the term…
Abstract
Test data are inserted on the data source side into a data stream (DS) transmitted to a data receiver (5) from a data source (1) via at least one transmission section (4). A bit error function (e(n)) is obtained from the test data received. The bit error function (e(n)) is evaluated in a part by part manner to discover its periodicity, the length of the partial section evaluated at the same time being measured in such a way that it contains a plurality of bit errors (BF). The bit error function (e(n) or err(n)) is subjected to auto-correlation. The maxima (M, M1, . . . Mi) of the auto-correlation functions (AKF(m)) thus obtained are used to determine the nature of the interference (10, 11) with reference to its stochastic or deterministic nature.
Claims14
In a system having
- a data source, providing a data stream,
- at least one transmission path, providing a path for transmission of the data stream, and
- a data receiver, receiving the data stream, a method for determining the type of a data corrupting interference, which causes bit errors in the data stream, comprising steps of:
- a) emitting test data from the data source;
- b) determining bit errors in data received;
- c) deriving a bit error function for the bit errors determined in step (b);
- d) dividing the bit error function into sections;
- e) performing a section by section autocorrelation of the bit error function to obtain an autocorrelation function, a length of an autocorrelation section at any one time being rated to contain a plurality of bit errors;
- f) checking the periodicity of the bit-error function based on the autocorrelation performed in step (e);
- g) determining the type of interference to be stochastic whenever the autocorrelation function obtained in step (e) exhibits a significant maximum without further pronounced maxima, and determining the type of interference to be deterministic whenever the autocorrelation function obtained in step (e) exhibits several maxima of an equidistant distribution.
In a system having
- a data source, providing a data stream, the data stream including successive data cells, each data cell including a cell head and a useful cell field,
- a scrambler scrambling the data stream in accordance with a specified instruction by a known period length,
- at least one transmission path, providing a transmission path for the scrambled data stream,
- a descrambler, descrambling the scrambled data in accordance with an instruction corresponding to the specified instruction, and
- a data receiver, receiving the descrambled data stream, a method for determining the type of a data corrupting interference, which causes bit errors in the data stream, comprising steps of:
- a) writing test data into the data cells at the data source to create test cells;
- b) determining bit errors in data received;
- c) deriving a bit error function for the bit errors determined in step (b);
- d) dividing the bit error function into sections;
- e) performing a section by section autocorrelation of the bit error function to obtain an autocorrelation function, a length of an autocorrelation section at any one time is selected to be at least one decade longer than the known period length;
- f) checking the periodicity of the bit-error function based on the autocorrelation performed in step (e);
- g) determining the type of interference to be stochastic whenever the autocorrelation function obtained in step (e) exhibits a significant maximum without further pronounced maxima, and determining the type of interference to be deterministic whenever the autocorrelation function obtained in step (e) exhibits several maxima of an equidistant distribution.
Dependencies are inferred from the claim text.