On 03/28/2012 07:01 PM, Zhonghua wrote:
Hi all,
I am trying to use cc1k of ucla_zigbee_phy. The testbench are
cc1k_txtest.py and cc1k_rxtest.py in which the signal source and sink
are amended to UHD since my platform is USRP E100. There is nothing
outputs from the RX TB. I checked the ucla_sos_packet_sink.cc, it can
receive data when TX sends although there is no data matches the
access code. I want to know why there is no access code be demodulated
correctly so I studied the relevant codes of cc1k.py and
cc1k_sos_pkt.py. Many problems emerge in front of me. I don't know if
anybody has succeeded in using it. If so, very thankful for your
explain and suggestions.
1 The spb in TX and sps in RX are set to 8, what is the reason?
2 In TX there is preamble but in ucla_sos_packet_sink.cc, no preamble
needs to be found before the sync data. Actually it is just a simple
comparison between input data and the access code. So why we need
transmit preamble in TX?
3 In cc1k_txtest.py the access code is set as 0x33 plus 0xcc, but in
cc1k_sos_pkt the access code of cc1k_demod_pkts module is
0x999999995a5aa5a5. The ucla_sos_packet_sink.cc also requires the
access code to be 8 bytes. How can the access code in TX be discerned
in RX? BTW, even after I changed the access code of TX to
0x999999995a5aa5a5, the packet_sink still can not detect out the sync
data.
4 In cc1k.py, the cc1k_mod module sets 'sensitivity' to 3.5*pi / spb
but the cc1k_demod module set 'sensitivity' to (pi/2)/sps. Actually I
am not sure what is the functionality of this variable but I think it
may should be set as identical in both mod and demod module.
5 In cc1k_demod module, the alpha variable, which is parameter of
gr.single_pole_iir_filter_ff, is set to 0.0512/sps. I don't know why
it should be this value. In ieee802_15_4.py, the demodulation module
sets it to 0.0008/self.sps. What is foundation to set this variable?
At last, what configuration in TX and RX can make the communication
runs well?
Appreciate in advance for your reply.
Sincerely,
Zhonghua
I composed a tidy version by directly connect TX and RX together to
thoroughly exclude the influence of radio front end. TX is nothing but
packet making and modulation. RX is demodulation and packet sink(to do
Manchester decoding and sync detect). I also experimented to set sps and
spb to be 1, 2 or 8. It is still has no correct data in packet sink.
I suspect is it right that the adopted bsfk mechanism which uses
'gr.frequency_modulator_fc(sensitivity)' as modulator and
'gr.quadrature_demod_cf(1 / sensitivity)' as demodulator. In
ieee802_15_4.py we can see the modulator is ucla.qpsk_modulator_cc() and
the demodulator is also gr.quadrature_demod_cf(1). That means the
quadrature demodulation can demodulate both qpsk and fm?
Could somebody give me any help? Thanks!
Br,
Zhonghua
_______________________________________________
Discuss-gnuradio mailing list
[email protected]
https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
