r_gomes
Newbie level 6
Hello everyone,
I have been working on OFDM for some time and since OFDM is characterized to have out-of-band emissions due to the sinc frequency response of each subcarrier. In order to overcome this issue, I found in the literature the FS-FBMC. Basically, this waveform uses the same principle introducing a filter to shape the response of the OFDM symbol. However, I noticed that after following the references the model is not working properly.
See the references:
https://meocloud.pt/link/1e7aacda-9764-499c-8957-3a1bc716bdde/FBMC_ref/
To transmit FBMC symbols and to achieve the same throughput of an OFDM system, prior to the transmission the FBMC symbols must be overlapped and summed in order to transmit the same number of samples in the time domain, as in OFDM. The counter part of this block is known as time window. I tried to implement both blocks, however the performance of the system is not the expected, i.e high values of BER in a back to back connection.
Can anyone share with him his thoughts about what may be wrong in the system? or if can provide some more informarion regarding to the overlap and sum/time window? (I do not found a lot of info about these blocks)
PS: I will post the model later on, if someone would like to analyse it.
I am looking forward to hearing from you,
Rodolfo
I have been working on OFDM for some time and since OFDM is characterized to have out-of-band emissions due to the sinc frequency response of each subcarrier. In order to overcome this issue, I found in the literature the FS-FBMC. Basically, this waveform uses the same principle introducing a filter to shape the response of the OFDM symbol. However, I noticed that after following the references the model is not working properly.
See the references:
https://meocloud.pt/link/1e7aacda-9764-499c-8957-3a1bc716bdde/FBMC_ref/
To transmit FBMC symbols and to achieve the same throughput of an OFDM system, prior to the transmission the FBMC symbols must be overlapped and summed in order to transmit the same number of samples in the time domain, as in OFDM. The counter part of this block is known as time window. I tried to implement both blocks, however the performance of the system is not the expected, i.e high values of BER in a back to back connection.
Can anyone share with him his thoughts about what may be wrong in the system? or if can provide some more informarion regarding to the overlap and sum/time window? (I do not found a lot of info about these blocks)
PS: I will post the model later on, if someone would like to analyse it.
I am looking forward to hearing from you,
Rodolfo
