Re: [Discuss-gnuradio] [USRP-users] Questions about benchmark_tx/benchmark_rx
On Fri, Jun 1, 2012 at 1:01 PM, Weixian Zhou idea...@gmail.com wrote: I used the benchmark_tx and benchmark_rx to transmit data and I have some questions: 1. I set the frequency of both benchmark_tx and benchmark_rx to 300M, the transmission succeed; but it failed when I set the frequency to 2.4G or other frequencies. Why is this? What daughter board u are using, maybe your daughter board doesn't support 2.4GHz, Target frequency: 300.00 MHz Actual frequency: 4500.00 MHz means u daughter board runs on 4.5GHz 2. Both the benchmark_tx and benchmark_rx gives the warning as I marked red: the actual frequency is not same but the transmission succeed. Why is it? Target frequency: 300.00 MHz Actual frequency: 4500.00 MHz means u daughter board runs on 4.5GHz, maybe it doesn't support 300MHz 3. In the receiving end, the pktno is disordered some pktnos are much larger than averages. Why is it? They are disrupted, so maybe some random numbers 4. What does the n_right=0 means? Does it mean the packet is corrupted in the receiving end? n_right means how many packets are correctly received, which means their crc check should be true (ok=True) *The benchmark_tx messages as following:* belltestbed@belltestbed-OptiPlex-790 :~/Desktop/gnuradio/gr-digital/les/narrowband$ ./benchmark_tx.py -f 300M --from-file=pic.jpg linux; GNU C++ version 4.6.3; Boost_104601; UHD_003.004.002-128-g12f7a5c9 gr_fir_ccf: using SSE -- Opening a USRP2/N-Series device... -- Current recv frame size: 1472 bytes -- Current send frame size: 1472 bytes UHD Warning: The recv buffer could not be resized sufficiently. Target sock buff size: 5000 bytes. Actual sock buff size: 100 bytes. See the transport application notes on buffer resizing. Please run: sudo sysctl -w net.core.rmem_max=5000 UHD Warning: The recv buffer could not be resized sufficiently. Target sock buff size: 5000 bytes. Actual sock buff size: 100 bytes. See the transport application notes on buffer resizing. Please run: sudo sysctl -w net.core.rmem_max=5000 UHD Warning: The send buffer could not be resized sufficiently. Target sock buff size: 1048576 bytes. Actual sock buff size: 100 bytes. See the transport application notes on buffer resizing. Please run: sudo sysctl -w net.core.wmem_max=1048576 UHD Warning: Unable to set the thread priority. Performance may be negatively affected. Please see the general application notes in the manual for instructions. EnvironmentError: OSError: error in pthread_setschedparam No gain specified. Setting gain to 17.50 (from [0.00, 35.00]) UHD Warning: The hardware does not support the requested TX frequency: Target frequency: 300.00 MHz Actual frequency: 4500.00 MHz UHD Warning: The hardware does not support the requested TX sample rate: Target sample rate: 0.05 MSps Actual sample rate: 0.195312 MSps Symbol Rate: 25000.00 Requested sps: 2.00 Given sample rate: 195312.50 Actual sps for rate: 7.812500 Requested sample rate: 5.00 Actual sample rate: 195312.50 Using Volk machine: avx_32 Warning: failed to enable realtime scheduling . *And the benchmark_rx messages as follwing:* bell@bell-HP-Compaq-4000-Pro-SFF-PC :~/Desktop/USRP/gnuradio/gr-digital/examples/narrowband$./benchmark_rx.py -f 300M linux; GNU C++ version 4.6.3; Boost_104601; UHD_003.004.001-129-g23344268 gr_fir_ccf: using SSE -- Opening a USRP2/N-Series device... -- Current recv frame size: 1472 bytes -- Current send frame size: 1472 bytes UHD Warning: The recv buffer could not be resized sufficiently. Target sock buff size: 5000 bytes. Actual sock buff size: 100 bytes. See the transport application notes on buffer resizing. Please run: sudo sysctl -w net.core.rmem_max=5000 UHD Warning: The recv buffer could not be resized sufficiently. Target sock buff size: 5000 bytes. Actual sock buff size: 100 bytes. See the transport application notes on buffer resizing. Please run: sudo sysctl -w net.core.rmem_max=5000 UHD Warning: The send buffer could not be resized sufficiently. Target sock buff size: 1048576 bytes. Actual sock buff size: 100 bytes. See the transport application notes on buffer resizing. Please run: sudo sysctl -w net.core.wmem_max=1048576 UHD Warning: Unable to set the thread priority. Performance may be negatively affected. Please see the general application notes in the manual for instructions. EnvironmentError: OSError: error in pthread_setschedparam No gain specified. Setting gain to 49.50 (from [0.00, 99.00]) UHD Warning: The hardware does not
[Discuss-gnuradio] Restore USRP2 images to support old USRP2 (Not UHD) code
Hi All, I have burned the USRP2 to support the UHD, but now I need to run some old USRP2 code. The new burned USRP2 images don't support the old code. So, is there any ways I can restore USRP2 images to old version to support the old code? Any help is appreciated, thanks in advance. -- Wenbo ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] Apply Trellis in benchmark_tx and rx
Hi, I applied trellis in benchmark_tx2 and rx2 using dbpsk2. I have modified the dbpsk2.py file according to the test_tcm_bit.py. On the TX side, I add: # turn bytes into k-bit vectors self.bytes2chunks = \ gr.packed_to_unpacked_bb(self.bits_per_symbol(), gr.GR_MSB_FIRST) #add the error correlation code fsm = trellis.fsm(/home/symbol/gnuradio/gr-trellis/src/examples/fsm_files/awgn1o2_128.fsm) self.trellis_enc = trellis.encoder_bb(fsm, 0); #gray code self.symbol_mapper = gr.map_bb(psk.binary_to_gray[arity]) self.diffenc = gr.diff_encoder_bb(arity) self.chunks2symbols = gr.chunks_to_symbols_bc(psk.constellation[arity]) # pulse shaping filter nfilts = 32 ntaps = nfilts * 11 * int(self._samples_per_symbol)# make nfilts filters of ntaps each self.rrc_taps = gr.firdes.root_raised_cosine( nfilts, # gain nfilts, # sampling rate based on 32 filters in resampler 1.0, # symbol rate self._excess_bw, # excess bandwidth (roll-off factor) ntaps) self.rrc_filter = gr.pfb_arb_resampler_ccf(self._samples_per_symbol, self.rrc_taps) # Connect self.connect(self, self.bytes2chunks, self.symbol_mapper, self.trellis_enc, self.diffenc, self.chunks2symbols, self.rrc_filter, self) On the RX side: self.phase_recov = gr.costas_loop_cc(self._phase_alpha, self._phase_beta, fmax, fmin, arity) # Do differential decoding based on phase change of symbols self.diffdec = gr.diff_phasor_cc() #the error correlation code fsm = trellis.fsm(/home/symbol/gnuradio/gr-trellis/src/examples/fsm_files/awgn1o2_128.fsm) metrics = trellis.metrics_c(fsm.O(), arity, psk.constellation[arity], trellis.TRELLIS_EUCLIDEAN) va = trellis.viterbi_b(fsm, 1520, 0, -1) self.symbol_mapper = gr.map_bb(psk.gray_to_binary[arity]) # unpack the k bit vector into a stream of bits self.unpack = gr.unpack_k_bits_bb(self.bits_per_symbol()) # Connect self.connect(self, self.agc, self.freq_recov, self.time_recov, self.phase_recov, self.diffdec, metrics, va, self.symbol_mapper, self.unpack, self) I can not get any packets on the receiver side. If I remove the trellis encode and decode block, everything works fine. There must be some problems with the use of trellis blocks. Could some one give me any clue? Any help is appreciated! Best, -- Wenbo ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
Re: [Discuss-gnuradio] error correcting codes block
OK, thanks Colby On Sat, Jun 11, 2011 at 11:46 AM, Colby Boyer colby.bo...@gmail.com wrote: There is a Viterbi encoder/decoder block called 'trellis' and it handles hard and soft decoding. What you need to do is find the generating polynomial or convolutional code that matches your needs. --Colby On Fri, Jun 10, 2011 at 1:27 PM, Shen Wenbo shenwenbosm...@gmail.comwrote: Hi All, Does GNURadio the error correcting code (ecc) blocks ? I have googled, and found some info about directory gr-error-correcting-codes, but this folder is not in GNURadio 3.3. So, can any block in GNURadio3.3 do error correcting? -- Wenbo ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio -- Wenbo ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] error correcting codes block
Hi All, Does GNURadio the error correcting code (ecc) blocks ? I have googled, and found some info about directory gr-error-correcting-codes, but this folder is not in GNURadio 3.3. So, can any block in GNURadio3.3 do error correcting? -- Wenbo ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
[Discuss-gnuradio] how to use gr_clock_recovery_mm_cc
Hi All, In the gmsk.py, the gr.clock_recovery_mm_ff is used. I want to use gr.clock_recovery_mm_cc (float omega, float gain_omega, float mu, float gain_mu) (which implements Mueller and Müller (MM) based clock recovery block with complex input, complex output) to replace the gr.clock_recovery_mm_ff I have done these in two steps: 1. use the same parameters (used to init gr.clock_recovery_mm_ff) to init gr.clock_recovery_mm_cc. omege=2.00, gain_omega=0.000625, mu=0.05, gain_mu=0.05, omega_relative_limit=0.005000 2. move the clock recovery block forward before demod block self.connect(self, self.clock_recovery, self.fmdemod, self.slicer, self) Before modification, if I run the benchmark program using gmsk, receive rate is almost 100%, but after my modification, receive rate drops to 0%. Does my method have problem, any help is appreciated! -- Wenbo ___ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org https://lists.gnu.org/mailman/listinfo/discuss-gnuradio
