Device and method to process OFDM-based symbols in wireless network

1. A device to process OFDM-based symbols comprising a base band input data and a base band output data, and comprising at least two programmable execution units connected to at least one working memory, this device being characterized in that, the programmable execution units are connected to the memory through a memory management unit, said programmable execution units having an input memory range and an output memory range, the device further comprising a state scheduler connected to the execution units and to the memory management unit, said scheduler having means to

load a related program into the execution units thus defining the task and the input and output memory ranges of said programmable execution units,

assign the input and output memory range of said working memory relative to the programmable execution units with the programmable execution units input and output memory ranges through the memory management unit at each pipeline period,

define a processing data flow of the symbols by defining the number of programmable execution units and the order in which the programmable execution units are involved.

2. The device of claim 1 , characterized in that, the scheduler has means to set the base band input and base band output data priority for each execution unit defining the priority to access the working memory.

3. The device of claim 1 comprising at least three execution units, characterized in that the scheduler has means to map, through the memory management unit, the input memory range of the last execution unit to the output memory range of the first execution unit during a first pipeline period and to map the input memory range of the last execution unit to the output memory range of the second execution unit during a second pipeline period.

4. The device of claim 1 comprising at least three execution units, characterized in that the scheduler has means to map, through the memory management unit, the output memory range of the first execution unit to the input memory range of the second execution unit during a first pipeline period and to map the output memory range of the first execution unit to the input memory range of the third execution unit during a second pipeline period.

5. The device of claim 1 , characterized in that, the scheduler has means to deactivate the execution units not currently involved in the processing thus saving power supply.

6. The device of claim 5 , characterized in that, the deactivation of an execution unit is achieved by freezing the clock supplied.

7. The device of claim 4 , characterized in that, the deactivation of an execution unit is achieved by switching off the power supply.

8. The device of claim 1 , characterized in that, the scheduler comprises several set of parameters, each set defining the program to be loaded in the execution units, the execution units involved to process the data flow between the base band input and the base band output data, and the mapping between the working memory and the input and output memory range for each execution unit relative to the pipeline period.

9. The device of claim 8 , characterized in that, a different set of parameters is loaded according to different protocols such as DAB, DVB, WLAN or WiMAX.

10. A method of processing OFDM-based symbols in a device comprising at least two programmable execution units connected to at least one working memory through a memory management unit a scheduler allowing the execution units to be mapped to the working memory, this method comprising the steps of:

loading a related program into the execution units thus defining an execution unit function,

loading into the memory management unit the input and output memory ranges of said programmable execution units allowing access by the execution units to the input and output data of said execution unit,

assigning the input and output memory range of said working memory relative to the programmable execution units with the programmable execution unit input and output memory ranges through the memory management unit at each pipeline period,

defining a processing data flow of the symbols by defining the number of programmable execution units and the order in which the programmable execution units are involved.

11. The method of claim 10 , characterized in that, it further comprises the step of setting the base band input and base band output data priority for each execution unit defining the priority to access the working memory.

12. The method of claim 10 , involving at least three execution units, characterized in that it further comprises the steps of mapping, through the memory management unit, the input memory range of the last execution unit to the output memory range of the first execution unit during a first pipeline period and mapping the input memory range of the last execution unit to the output memory range of the second execution unit during a second pipeline period.

13. The method of claim 10 , involving at least three execution units, characterized in that it further comprises the steps of mapping, through the memory management unit, the output memory range of the first execution unit to the input memory range of the second execution unit during a first pipeline period and mapping the output memory range of the first execution unit to the input memory range of the third execution unit during a second pipeline period.

14. The method of claim 10 , characterized in that, it further comprises the step of deactivating the execution units not currently involved in the processing to save power supply.

15. The method of claim 14 , characterized in that, the deactivation of an execution unit is achieved by freezing the clock supplied.

16. The method of claim 14 , characterized in that, the deactivation of an execution unit is achieved by switching off the power supply.

17. The method of claim 10 , characterized in that, it further comprises the step of loading one set of parameters among several set of parameters, each set defining the program to be loaded in the execution units, the execution units involved to process the data flow between the base band input and the base band output data, and the mapping between the working memory and the input and output memory range for each execution unit relative to the pipeline period.

18. The method of claim 17 , characterized in that, a different set of parameters is loaded according to different protocols such as DAB, DVB, WLAN or WiMAX.

19. A device to process OFDM-based symbols comprising a base band input data and a base band output data, and comprising at least two programmable execution units connected to at least one working memory, this device being characterized in that, the programmable execution units are connected to the memory through a memory management unit, said programmable execution units having an input memory range and an output memory range, the device further comprising a state scheduler connected to the execution units and to the memory management unit, said scheduler having means to

load a related program into the execution units thus defining the task

load into the memory management unit the input and output memory ranges of said programmable execution units allowing access by the execution units to the baseband input and base band output data of said execution units,

assign the input and output memory range of said working memory relative to the programmable execution units with the programmable execution units input and output memory ranges through the memory management unit at each pipeline period,

define a processing data flow of the symbols by defining the number of programmable execution units and the order in which the programmable execution units are involved.

20. A method of processing OFDM-based symbols in a device comprising base band input data and base band output data, and comprising at least two programmable execution units connected to at least one working memory, wherein the method comprises the connection of the programmable execution units to the working memory through a memory management unit, said programmable execution units having an input memory range and an output memory range, and further comprising the connection of a state scheduler to the execution units and to the memory management unit, allowing the execution units to be mapped to the working memory, this method comprising the steps of:

loading a related program into the execution units thus defining an execution unit function,

loading into the memory management unit the input and output memory ranges of said programmable execution units allowing access by the execution units to the base band input and base band output data of said execution unit,

assigning the input and output memory range of said working memory relative to the programmable execution unit with the programmable execution unit input and output memory ranges through the memory management unit at each pipeline period,

defining a processing data flow of the symbols by defining the number of programmable execution units and the order in which the programmable execution units are involved.