System and method for symbol boundary detection in orthogonal frequency divison multiplexing based data communication
View Patent ↗A method for determining a symbol boundary of a data packet of a received signal, where the data packet includes a first training field, a guard interval, and a second training field. The received signal is sampled to obtain multiple samples. A first symbol boundary estimate is determined using one or more block auto-correlation values. Thereafter, a second symbol boundary estimate is determined based on the first symbol boundary estimate and using one or more cross-correlation values. The second symbol boundary estimate then is shifted using moving average auto-correlation values for the samples in the vicinity of the second symbol boundary estimate to obtain an accurate symbol boundary estimate.
1. A method for determining a symbol boundary in a data packet belonging to a received signal, the data packet comprising a first training field, a guard interval, and a second training field, the method comprising:
sampling the received signal at a predetermined sampling frequency to obtain a plurality of samples of the received signal;
determining a first symbol boundary estimate using one or more block auto-correlation values for the plurality of samples with reference to a previously obtained plurality of samples, wherein the first symbol boundary estimate is an estimate of a boundary between the first training field and the guard interval;
selecting a first set of samples using the first symbol boundary estimate;
determining a second symbol boundary estimate using the first set of samples, wherein the second symbol boundary estimate is an estimate of a boundary between the guard interval and the second training field;
selecting a second set of samples in a vicinity of the second symbol boundary estimate;
calculating one or more moving average auto-correlation values for the second set of samples, wherein each of the one or more moving average auto-correlation values corresponds to a sample belonging to the second set of samples; and
shifting the second symbol boundary estimate using the one or more moving average auto-correlation values.
2. The method for determining a symbol boundary in the data packet of claim 1 , wherein the first symbol boundary estimate is estimated based on a maximum block auto-correlation value of the one or more block auto-correlation values.
3. The method for determining a symbol boundary in the data packet of claim 1 , wherein the determination of the second symbol boundary estimate comprises determining one or more inverted half auto-correlation values for the first set of samples based on the previously obtained plurality of samples.
4. The method for determining a symbol boundary in the data packet of claim 3 , wherein the second symbol boundary estimate is estimated based on a maximum inverted half auto-correlation value of the one or more inverted half auto-correlation values.
5. The method for determining a symbol boundary in the data packet of claim 1 , wherein the determination of the second symbol boundary estimate comprises determining one or more cross-correlation values for the first set of samples with reference to a set of stored samples.
6. The method for determining a symbol boundary in the data packet of claim 5 , wherein the second symbol boundary estimate is selected based on a maximum cross-correlation value of the one or more cross-correlation values.
7. The method for determining a symbol boundary in the data packet of claim 1 , wherein the second set of samples comprises a first predetermined number of samples received prior to the second symbol boundary estimate and a second predetermined number of samples received subsequent to the second symbol boundary estimate.
8. A method for determining a symbol boundary in a data packet belonging to a received signal, the data packet comprising a first training field, a guard interval, and a second training field, the method comprising:
sampling the received signal at a predetermined sampling frequency to obtain a plurality of samples of the received signal;
determining a first symbol boundary estimate using one or more block auto-correlation values for the plurality of samples based on a previously obtained plurality of samples, wherein the first symbol boundary estimate is an estimate of a boundary between the first training field and the guard interval;
selecting a first set of samples using the first symbol boundary estimate;
determining a second symbol boundary estimate using the first set of samples, wherein the second symbol boundary estimate is an estimate of a boundary between the guard interval and the second training field;
determining a first predetermined number of the one or more block auto-correlation values for a second set of samples obtained subsequent to the second symbol boundary estimate with reference to the previously obtained plurality of samples;
shifting the second symbol boundary estimate using the first predetermined number of the one or more block auto-correlation values to obtain a third symbol boundary estimate;
selecting a third set of samples in a vicinity of the third boundary estimate;
calculating one or more moving average auto-correlation values for the third set of samples, wherein each of the one or more moving average auto-correlation values corresponds to a sample belonging to the third set of samples; and
shifting the third symbol boundary estimate using the one or more moving average auto-correlation values.
9. The method for determining a symbol boundary in the data packet of claim 8 , wherein the first symbol boundary estimate is estimated based on a maximum block auto-correlation value of the one or more block auto-correlation values.
10. The method for determining a symbol boundary in the data packet of claim 8 , wherein the determination of the second symbol boundary estimate comprises, determining one or more inverted half auto-correlation values for the first set of samples based on the previously obtained plurality of samples.
11. The method for determining a symbol boundary in the data packet of claim 10 , wherein the second symbol boundary estimate is estimated based on a maximum inverted half auto-correlation value of the one or more inverted half auto-correlation values.
12. The method for determining a symbol boundary in the data packet of claim 8 , wherein the determination of the second symbol boundary estimate comprises, determining one or more cross-correlation values for the first set of samples based on a set of stored samples.
13. The method for determining a symbol boundary in the data packet of claim 12 , wherein the second symbol boundary estimate is selected based on a maximum cross-correlation value of the one or more cross-correlation values.
14. The method for determining a symbol boundary in the data packet of claim 8 , wherein the second symbol boundary estimate is shifted based on comparison of the first predetermined number of the one or more block auto-correlation values with a maximum block auto-correlation value.
15. The method for determining a symbol boundary in the data packet of claim 8 , wherein the third set of samples comprises a first predetermined number of samples received prior to the third symbol boundary estimate and a second predetermined number of samples received subsequent to the third symbol boundary estimate.
16. A system for determining a symbol boundary in a data packet belonging to a received signal, the data packet comprising a first training field, a guard interval, and a second training field, the system comprising:
a sampling unit for sampling the received signal at a predetermined sampling frequency to obtain a plurality of samples of the received signal;
a memory, connected to the sampling unit, for storing at least one of the plurality of samples, one or more block auto-correlation values, one or more cross-correlation values, and one or more moving average auto-correlation values;
an auto-correlation unit, connected to the sampling unit and the memory, for determining the one or more block auto-correlation values for the plurality of samples based on a previously obtained plurality of samples;
a cross-correlation unit, connected to the sampling unit and the memory, for determining the one or more cross-correlation values for a first set of samples based on one or more of the stored plurality of samples;
a moving average auto-correlation unit, connected to the memory, for calculating the one or more moving average auto-correlation values for a second set of samples, wherein each of the one or more moving average auto-correlation values corresponds to a sample belonging to the second set of samples; and
a processor, connected to the moving average auto-correlation unit, the memory, and the cross-correlation unit, for processing at least one of the one or more block auto-correlation values, the cross-correlation values, and the moving average auto-correlation values to determine at least one symbol boundary estimate.
17. The system for determining a symbol boundary in the data packet of claim 16 , wherein the at least one symbol boundary estimate is one of an estimate of a boundary between the first training field and the guard interval, and a boundary between the guard interval and the second training field.
18. The system for determining a symbol boundary in the data packet of claim 16 , further comprising a radio-frequency unit, coupled to the sampling unit, for receiving an Orthogonal Frequency Division Multiplexing (OFDM) radio signal and providing the OFDM radio signal to the sampling unit.
19. The system for determining a symbol boundary in the data packet of claim 16 , wherein the system is part of an IEEE 802.11n Wireless Local Area Network (WLAN) communication system.