aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorPiotr Krysik <ptrkrysik@gmail.com>2017-08-21 09:26:05 +0200
committerPiotr Krysik <ptrkrysik@gmail.com>2017-08-21 09:26:05 +0200
commit938681128b47a90c274e48b6e36bea6cfa295d70 (patch)
treecea5d333416bcc9f022b8a1c2f4789b71f5250eb
parent8c5e5520a562d8b985f387df54f3a0bf4115d9a6 (diff)
parentcc82cf0300cb005f424bf6f5ebc02674a5537cbf (diff)
Merge refactoring of the receiver done in branch 'fixeria/receiver' of https://github.com/axilirator/gr-gsm into axilirator-fixeria/receiver
-rw-r--r--lib/receiver/receiver_impl.cc1657
-rw-r--r--lib/receiver/receiver_impl.h9
2 files changed, 930 insertions, 736 deletions
diff --git a/lib/receiver/receiver_impl.cc b/lib/receiver/receiver_impl.cc
index e69183f..76ed08d 100644
--- a/lib/receiver/receiver_impl.cc
+++ b/lib/receiver/receiver_impl.cc
@@ -26,884 +26,1073 @@
#include <gnuradio/io_signature.h>
#include <gnuradio/math.h>
-#include <math.h>
-#include <boost/circular_buffer.hpp>
+
#include <algorithm>
-#include <numeric>
-#include <vector>
-#include <viterbi_detector.h>
#include <string.h>
#include <iostream>
-//#include <iomanip>
-#include <boost/scoped_ptr.hpp>
+#include <numeric>
+#include <math.h>
+#include <vector>
-#include <sch.h>
-#include "receiver_impl.h"
+#include <boost/circular_buffer.hpp>
+#include <boost/scoped_ptr.hpp>
#include <grgsm/endian.h>
-//files included for debuging
-//#include "plotting/plotting.hpp"
-//#include <pthread.h>
+#include "receiver_impl.h"
+#include "viterbi_detector.h"
+#include "sch.h"
+
+#if 0
+/* Files included for debuging */
+#include "plotting/plotting.hpp"
+#include <pthread.h>
+#include <iomanip>
+#endif
#define SYNC_SEARCH_RANGE 30
namespace gr
{
-namespace gsm
-{
-receiver::sptr
-receiver::make(int osr, const std::vector<int> &cell_allocation, const std::vector<int> &tseq_nums, bool process_uplink)
-{
- return gnuradio::get_initial_sptr
- (new receiver_impl(osr, cell_allocation, tseq_nums, process_uplink));
-}
+ namespace gsm
+ {
+
+ /* The public constructor */
+ receiver::sptr
+ receiver::make(
+ int osr, const std::vector<int> &cell_allocation,
+ const std::vector<int> &tseq_nums, bool process_uplink)
+ {
+ return gnuradio::get_initial_sptr
+ (new receiver_impl(osr, cell_allocation,
+ tseq_nums, process_uplink));
+ }
-/*
- * The private constructor
- */
-receiver_impl::receiver_impl(int osr, const std::vector<int> &cell_allocation, const std::vector<int> &tseq_nums, bool process_uplink)
- : gr::sync_block("receiver",
- gr::io_signature::make(1, -1, sizeof(gr_complex)),
- gr::io_signature::make(0, 0, 0)),
- d_OSR(osr),
- d_process_uplink(process_uplink),
- d_chan_imp_length(CHAN_IMP_RESP_LENGTH),
- d_counter(0),
- d_fcch_start_pos(0),
- d_freq_offset_setting(0),
- d_state(fcch_search),
- d_burst_nr(osr),
- d_failed_sch(0),
- d_signal_dbm(-120),
- d_tseq_nums(tseq_nums),
- d_cell_allocation(cell_allocation),
- d_last_time(0.0)
-{
- int i;
- //don't send samples to the receiver until there are at least samples for one
- set_output_multiple(floor((TS_BITS + 2 * GUARD_PERIOD) * d_OSR)); // burst and two gurad periods (one gurard period is an arbitrary overlap)
- gmsk_mapper(SYNC_BITS, N_SYNC_BITS, d_sch_training_seq, gr_complex(0.0, -1.0));
- for (i = 0; i < TRAIN_SEQ_NUM; i++)
+ /* The private constructor */
+ receiver_impl::receiver_impl(
+ int osr, const std::vector<int> &cell_allocation,
+ const std::vector<int> &tseq_nums, bool process_uplink
+ ) : gr::sync_block("receiver",
+ gr::io_signature::make(1, -1, sizeof(gr_complex)),
+ gr::io_signature::make(0, 0, 0)),
+ d_OSR(osr),
+ d_process_uplink(process_uplink),
+ d_chan_imp_length(CHAN_IMP_RESP_LENGTH),
+ d_counter(0),
+ d_fcch_start_pos(0),
+ d_freq_offset_setting(0),
+ d_state(fcch_search),
+ d_burst_nr(osr),
+ d_failed_sch(0),
+ d_signal_dbm(-120),
+ d_tseq_nums(tseq_nums),
+ d_cell_allocation(cell_allocation),
+ d_last_time(0.0)
{
- gr_complex startpoint = (train_seq[i][0]==0) ? gr_complex(1.0, 0.0) : gr_complex(-1.0, 0.0); //if first bit of the seqeunce ==0 first symbol ==1
- //if first bit of the seqeunce ==1 first symbol ==-1
- gmsk_mapper(train_seq[i], N_TRAIN_BITS, d_norm_training_seq[i], startpoint);
+ /**
+ * Don't send samples to the receiver
+ * until there are at least samples for one
+ */
+ set_output_multiple(floor((TS_BITS + 2 * GUARD_PERIOD) * d_OSR));
+
+ /**
+ * Prepare SCH sequence bits
+ *
+ * (TS_BITS + 2 * GUARD_PERIOD)
+ * Burst and two guard periods
+ * (one guard period is an arbitrary overlap)
+ */
+ gmsk_mapper(SYNC_BITS, N_SYNC_BITS,
+ d_sch_training_seq, gr_complex(0.0, -1.0));
+
+ /* Prepare bits of training sequences */
+ for (int i = 0; i < TRAIN_SEQ_NUM; i++) {
+ /**
+ * If first bit of the sequence is 0
+ * => first symbol is 1, else -1
+ */
+ gr_complex startpoint = train_seq[i][0] == 0 ?
+ gr_complex(1.0, 0.0) : gr_complex(-1.0, 0.0);
+ gmsk_mapper(train_seq[i], N_TRAIN_BITS,
+ d_norm_training_seq[i], startpoint);
+ }
+
+ /* Register output ports */
+ message_port_register_out(pmt::mp("C0"));
+ message_port_register_out(pmt::mp("CX"));
+ message_port_register_out(pmt::mp("measurements"));
+
+ /**
+ * Configure the receiver,
+ * i.e. tell it where to find which burst type
+ */
+ configure_receiver();
}
- message_port_register_out(pmt::mp("C0"));
- message_port_register_out(pmt::mp("CX"));
- message_port_register_out(pmt::mp("measurements"));
- configure_receiver(); //configure the receiver - tell it where to find which burst type
-}
-/*
- * Our virtual destructor.
- */
-receiver_impl::~receiver_impl()
-{
-}
+ /* Our virtual destructor */
+ receiver_impl::~receiver_impl() {}
-int
-receiver_impl::work(int noutput_items,
- gr_vector_const_void_star &input_items,
- gr_vector_void_star &output_items)
-{
-// std::vector<const gr_complex *> iii = (std::vector<const gr_complex *>) input_items; // jak zrobić to rzutowanie poprawnie
- gr_complex * input = (gr_complex *) input_items[0];
- std::vector<tag_t> freq_offset_tags;
- uint64_t start = nitems_read(0);
- uint64_t stop = start + noutput_items;
-
- float current_time = static_cast<float>(start)/(GSM_SYMBOL_RATE*d_OSR);
- if((current_time - d_last_time) > 0.1)
+ int
+ receiver_impl::work(
+ int noutput_items,
+ gr_vector_const_void_star &input_items,
+ gr_vector_void_star &output_items)
{
- pmt::pmt_t msg = pmt::make_tuple(pmt::mp("current_time"),pmt::from_double(current_time));
+ gr_complex *input = (gr_complex *) input_items[0];
+ uint64_t start = nitems_read(0);
+ uint64_t stop = start + noutput_items;
+ d_freq_offset_tag_in_fcch = false;
+
+#if 0
+ /* FIXME: jak zrobić to rzutowanie poprawnie */
+ std::vector<const gr_complex *> iii =
+ (std::vector<const gr_complex *>) input_items;
+#endif
+
+ /* Time synchronization loop */
+ float current_time =
+ static_cast<float>(start / (GSM_SYMBOL_RATE * d_OSR));
+ if ((current_time - d_last_time) > 0.1) {
+ pmt::pmt_t msg = pmt::make_tuple(pmt::mp("current_time"),
+ pmt::from_double(current_time));
message_port_pub(pmt::mp("measurements"), msg);
d_last_time = current_time;
- }
+ }
+
+ /* Frequency correction loop */
+ std::vector<tag_t> freq_offset_tags;
+ pmt::pmt_t key = pmt::string_to_symbol("setting_freq_offset");
+ get_tags_in_range(freq_offset_tags, 0, start, stop, key);
- pmt::pmt_t key = pmt::string_to_symbol("setting_freq_offset");
- get_tags_in_range(freq_offset_tags, 0, start, stop, key);
- bool freq_offset_tag_in_fcch = false;
-
- if(!freq_offset_tags.empty()){
+ if (!freq_offset_tags.empty()) {
tag_t freq_offset_tag = freq_offset_tags[0];
uint64_t tag_offset = freq_offset_tag.offset - start;
-
+ d_freq_offset_setting = pmt::to_double(freq_offset_tag.value);
+
burst_type b_type = d_channel_conf.get_burst_type(d_burst_nr);
- if(d_state == synchronized && b_type == fcch_burst){
- uint64_t last_sample_nr = ceil((GUARD_PERIOD + 2.0 * TAIL_BITS + 156.25) * d_OSR) + 1;
- if(tag_offset < last_sample_nr){
- freq_offset_tag_in_fcch = true;
- }
+ if (d_state == synchronized && b_type == fcch_burst){
+ uint64_t last_sample_nr =
+ ceil((GUARD_PERIOD + 2.0 * TAIL_BITS + 156.25) * d_OSR) + 1;
+ d_freq_offset_tag_in_fcch = tag_offset < last_sample_nr;
}
- d_freq_offset_setting = pmt::to_double(freq_offset_tag.value);
+ }
+
+ /* Main state machine */
+ switch (d_state) {
+ case fcch_search:
+ fcch_search_handler(input, noutput_items);
+ break;
+ case sch_search:
+ sch_search_handler(input, noutput_items);
+ break;
+ case synchronized:
+ synchronized_handler(input, input_items, noutput_items);
+ break;
+ }
+
+ return 0;
}
-
- switch (d_state)
- {
- //bootstrapping
- case fcch_search:
+
+ void
+ receiver_impl::fcch_search_handler(gr_complex *input, int noutput_items)
{
- double freq_offset_tmp;
- if (find_fcch_burst(input, noutput_items,freq_offset_tmp))
- {
- pmt::pmt_t msg = pmt::make_tuple(pmt::mp("freq_offset"),pmt::from_double(freq_offset_tmp-d_freq_offset_setting),pmt::mp("fcch_search"));
- message_port_pub(pmt::mp("measurements"), msg);
+ double freq_offset_tmp;
- d_state = sch_search;
- }
- else
- {
- d_state = fcch_search;
- }
- break;
+ /* Check if received samples is a FCCN burst */
+ if (!find_fcch_burst(input, noutput_items, freq_offset_tmp))
+ return;
+
+ /* We found it, compose a message */
+ pmt::pmt_t msg = pmt::make_tuple(
+ pmt::mp("freq_offset"),
+ pmt::from_double(freq_offset_tmp - d_freq_offset_setting),
+ pmt::mp("fcch_search")
+ );
+
+ /* Notify FCCH loop */
+ message_port_pub(pmt::mp("measurements"), msg);
+
+ /* Update current state */
+ d_state = sch_search;
}
- case sch_search:
+ void
+ receiver_impl::sch_search_handler(gr_complex *input, int noutput_items)
{
- std::vector<gr_complex> channel_imp_resp(CHAN_IMP_RESP_LENGTH*d_OSR);
- int t1, t2, t3;
- int burst_start = 0;
- unsigned char output_binary[BURST_SIZE];
+ std::vector<gr_complex> channel_imp_resp(CHAN_IMP_RESP_LENGTH * d_OSR);
+ unsigned char burst_buf[BURST_SIZE];
+ int rc, t1, t2, t3;
+ int burst_start;
+
+ /* Wait until we get a SCH burst */
+ if (!reach_sch_burst(noutput_items))
+ return;
+
+ /* Get channel impulse response from it */
+ burst_start = get_sch_chan_imp_resp(input, &channel_imp_resp[0]);
+
+ /* Detect bits using MLSE detection */
+ detect_burst(input, &channel_imp_resp[0], burst_start, burst_buf);
+
+ /* Attempt to decode BSIC and frame number */
+ rc = decode_sch(&burst_buf[3], &t1, &t2, &t3, &d_ncc, &d_bcc);
+ if (rc) {
+ /**
+ * There is error in the SCH burst,
+ * go back to the FCCH search state
+ */
+ d_state = fcch_search;
+ return;
+ }
+
+ /* Set counter of bursts value */
+ d_burst_nr.set(t1, t2, t3, 0);
+ d_burst_nr++;
+
+ /* Consume samples up to the next guard period */
+ consume_each(burst_start + BURST_SIZE * d_OSR + 4 * d_OSR);
+
+ /* Update current state */
+ d_state = synchronized;
+ }
- if (reach_sch_burst(noutput_items)) //wait for a SCH burst
+ void
+ receiver_impl::synchronized_handler(gr_complex *input,
+ gr_vector_const_void_star &input_items, int noutput_items)
+ {
+ /**
+ * In this state receiver is synchronized and it processes
+ * bursts according to burst type for given burst number
+ */
+ std::vector<gr_complex> channel_imp_resp(CHAN_IMP_RESP_LENGTH * d_OSR);
+ unsigned int inputs_to_process = d_cell_allocation.size();
+ unsigned char output_binary[BURST_SIZE];
+ burst_type b_type;
+ int to_consume = 0;
+ int offset = 0;
+
+ if (d_process_uplink)
+ inputs_to_process *= 2;
+
+ /* Process all connected inputs */
+ for (int input_nr = 0; input_nr < inputs_to_process; input_nr++) {
+ input = (gr_complex *) input_items[input_nr];
+ double signal_pwr = 0;
+
+ for (int ii = GUARD_PERIOD; ii < TS_BITS; ii++)
+ signal_pwr += abs(input[ii]) * abs(input[ii]);
+
+ signal_pwr = signal_pwr / (TS_BITS);
+ d_signal_dbm = round(10 * log10(signal_pwr / 50));
+
+ if (input_nr == 0)
+ d_c0_signal_dbm = d_signal_dbm;
+
+ /* Get burst type for given burst number */
+ b_type = input_nr == 0 ?
+ d_channel_conf.get_burst_type(d_burst_nr) : normal_or_noise;
+
+ /* Process burst according to its type */
+ switch (b_type) {
+ case fcch_burst:
{
- burst_start = get_sch_chan_imp_resp(input, &channel_imp_resp[0]); //get channel impulse response from it
- detect_burst(input, &channel_imp_resp[0], burst_start, output_binary); //detect bits using MLSE detection
- if (decode_sch(&output_binary[3], &t1, &t2, &t3, &d_ncc, &d_bcc) == 0) //decode SCH burst
- {
- d_burst_nr.set(t1, t2, t3, 0); //set counter of bursts value
- d_burst_nr++;
-
- consume_each(burst_start + BURST_SIZE * d_OSR + 4*d_OSR); //consume samples up to next guard period
- d_state = synchronized;
- }
- else
- {
- d_state = fcch_search; //if there is error in the sch burst go back to fcch search phase
+ if (d_freq_offset_tag_in_fcch)
+ break;
+
+ /* Send all-zero sequence message */
+ send_burst(d_burst_nr, fc_fb, GSMTAP_BURST_FCCH, input_nr);
+
+ /* Extract frequency offset */
+ const unsigned first_sample =
+ ceil((GUARD_PERIOD + 2 * TAIL_BITS) * d_OSR) + 1;
+ const unsigned last_sample =
+ first_sample + USEFUL_BITS * d_OSR - TAIL_BITS * d_OSR;
+ double freq_offset_tmp =
+ compute_freq_offset(input, first_sample, last_sample);
+
+ /* Frequency correction loop */
+ pmt::pmt_t msg = pmt::make_tuple(
+ pmt::mp("freq_offset"),
+ pmt::from_double(freq_offset_tmp - d_freq_offset_setting),
+ pmt::mp("synchronized"));
+ message_port_pub(pmt::mp("measurements"), msg);
+
+ break;
+ }
+
+ case sch_burst:
+ {
+ int d_ncc, d_bcc;
+ int t1, t2, t3;
+ int rc;
+
+ /* Get channel impulse response */
+ d_c0_burst_start = get_sch_chan_imp_resp(input,
+ &channel_imp_resp[0]);
+
+ /* Perform MLSE detection */
+ detect_burst(input, &channel_imp_resp[0],
+ d_c0_burst_start, output_binary);
+
+ /* Compose a message with GSMTAP header and bits */
+ send_burst(d_burst_nr, output_binary,
+ GSMTAP_BURST_SCH, input_nr);
+
+ /* Attempt to decode SCH burst */
+ rc = decode_sch(&output_binary[3], &t1, &t2, &t3, &d_ncc, &d_bcc);
+ if (rc) {
+ if (++d_failed_sch >= MAX_SCH_ERRORS) {
+ /* We have to resynchronize, change state */
+ d_state = fcch_search;
+
+ /* Frequency correction loop */
+ pmt::pmt_t msg = pmt::make_tuple(pmt::mp("freq_offset"),
+ pmt::from_double(0.0),pmt::mp("sync_loss"));
+ message_port_pub(pmt::mp("measurements"), msg);
}
+
+ break;
+ }
+
+ /**
+ * Decoding was successful, now
+ * compute offset from burst_start,
+ * burst should start after a guard period.
+ */
+ offset = d_c0_burst_start - floor((GUARD_PERIOD) * d_OSR);
+ to_consume += offset;
+ d_failed_sch = 0;
+
+ break;
}
- else
+
+ case normal_burst:
{
- d_state = sch_search;
+ float normal_corr_max;
+ /**
+ * Get channel impulse response for given
+ * training sequence number - d_bcc
+ */
+ d_c0_burst_start = get_norm_chan_imp_resp(input,
+ &channel_imp_resp[0], &normal_corr_max, d_bcc);
+
+ /* Perform MLSE detection */
+ detect_burst(input, &channel_imp_resp[0],
+ d_c0_burst_start, output_binary);
+
+ /* Compose a message with GSMTAP header and bits */
+ send_burst(d_burst_nr, output_binary,
+ GSMTAP_BURST_NORMAL, input_nr);
+
+ break;
}
- break;
- }
- //in this state receiver is synchronized and it processes bursts according to burst type for given burst number
- case synchronized:
- {
- std::vector<gr_complex> channel_imp_resp(CHAN_IMP_RESP_LENGTH*d_OSR);
- int offset = 0;
- int to_consume = 0;
- unsigned char output_binary[BURST_SIZE];
- burst_type b_type;
- unsigned int inputs_to_process=d_cell_allocation.size();
-
- if(d_process_uplink)
+
+ case dummy_or_normal:
{
- inputs_to_process = 2*inputs_to_process;
+ unsigned int normal_burst_start, dummy_burst_start;
+ float dummy_corr_max, normal_corr_max;
+
+ dummy_burst_start = get_norm_chan_imp_resp(input,
+ &channel_imp_resp[0], &dummy_corr_max, TS_DUMMY);
+ normal_burst_start = get_norm_chan_imp_resp(input,
+ &channel_imp_resp[0], &normal_corr_max, d_bcc);
+
+ if (normal_corr_max > dummy_corr_max) {
+ d_c0_burst_start = normal_burst_start;
+
+ /* Perform MLSE detection */
+ detect_burst(input, &channel_imp_resp[0],
+ normal_burst_start, output_binary);
+
+ /* Compose a message with GSMTAP header and bits */
+ send_burst(d_burst_nr, output_binary,
+ GSMTAP_BURST_NORMAL, input_nr);
+ } else {
+ d_c0_burst_start = dummy_burst_start;
+
+ /* Compose a message with GSMTAP header and bits */
+ send_burst(d_burst_nr, dummy_burst,
+ GSMTAP_BURST_DUMMY, input_nr);
+ }
+
+ break;
}
-
- for(int input_nr=0; input_nr<inputs_to_process; input_nr++)
+
+ case normal_or_noise:
{
- double signal_pwr = 0;
- input = (gr_complex *)input_items[input_nr];
-
- for(int ii=GUARD_PERIOD;ii<TS_BITS;ii++)
- {
- signal_pwr += abs(input[ii])*abs(input[ii]);
- }
- signal_pwr = signal_pwr/(TS_BITS);
- d_signal_dbm = round(10*log10(signal_pwr/50));
- if(input_nr==0){
- d_c0_signal_dbm = d_signal_dbm;
+ std::vector<gr_complex> v(input, input + noutput_items);
+ float normal_corr_max = -1e6;
+ float normal_corr_max_tmp;
+ unsigned int burst_start;
+ int max_tn, tseq_num;
+
+ if (d_tseq_nums.size() == 0) {
+ /**
+ * There is no information about training sequence,
+ * however the receiver can detect it with use of a
+ * very simple algorithm based on finding
+ */
+ get_norm_chan_imp_resp(input, &channel_imp_resp[0],
+ &normal_corr_max, 0);
+
+ float ts_max = normal_corr_max;
+ int ts_max_num = 0;
+
+ for (int ss = 1; ss <= 7; ss++) {
+ get_norm_chan_imp_resp(input, &channel_imp_resp[0],
+ &normal_corr_max, ss);
+
+ if (ts_max < normal_corr_max) {
+ ts_max = normal_corr_max;
+ ts_max_num = ss;
+ }
}
-
- if(input_nr==0) //for c0 channel burst type is controlled by channel configuration
- {
- b_type = d_channel_conf.get_burst_type(d_burst_nr); //get burst type for given burst number
- }
- else
- {
- b_type = normal_or_noise; //for the rest it can be only normal burst or noise (at least at this moment of development)
- }
-
- switch (b_type)
- {
- case fcch_burst: //if it's FCCH burst
- {
- if(freq_offset_tag_in_fcch==false)
- {
- const unsigned first_sample = ceil((GUARD_PERIOD + 2 * TAIL_BITS) * d_OSR) + 1;
- const unsigned last_sample = first_sample + USEFUL_BITS * d_OSR - TAIL_BITS * d_OSR;
- double freq_offset_tmp = compute_freq_offset(input, first_sample, last_sample); //extract frequency offset from it
- send_burst(d_burst_nr, fc_fb, GSMTAP_BURST_FCCH, input_nr);
-
- pmt::pmt_t msg = pmt::make_tuple(pmt::mp("freq_offset"),pmt::from_double(freq_offset_tmp-d_freq_offset_setting),pmt::mp("synchronized"));
- message_port_pub(pmt::mp("measurements"), msg);
- }
- break;
- }
- case sch_burst: //if it's SCH burst
- {
- int t1, t2, t3, d_ncc, d_bcc;
- d_c0_burst_start = get_sch_chan_imp_resp(input, &channel_imp_resp[0]); //get channel impulse response
-
- detect_burst(input, &channel_imp_resp[0], d_c0_burst_start, output_binary); //MLSE detection of bits
- send_burst(d_burst_nr, output_binary, GSMTAP_BURST_SCH, input_nr);
- if (decode_sch(&output_binary[3], &t1, &t2, &t3, &d_ncc, &d_bcc) == 0) //and decode SCH data
- {
- // d_burst_nr.set(t1, t2, t3, 0); //but only to check if burst_start value is correct
- d_failed_sch = 0;
- offset = d_c0_burst_start - floor((GUARD_PERIOD) * d_OSR); //compute offset from burst_start - burst should start after a guard period
- to_consume += offset; //adjust with offset number of samples to be consumed
- }
- else
- {
- d_failed_sch++;
- if (d_failed_sch >= MAX_SCH_ERRORS)
- {
- d_state = fcch_search;
- pmt::pmt_t msg = pmt::make_tuple(pmt::mp("freq_offset"),pmt::from_double(0.0),pmt::mp("sync_loss"));
- message_port_pub(pmt::mp("measurements"), msg);
- //DCOUT("Re-Synchronization!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!");
- }
- }
- break;
- }
- case normal_burst:
- {
- float normal_corr_max; //if it's normal burst
- d_c0_burst_start = get_norm_chan_imp_resp(input, &channel_imp_resp[0], &normal_corr_max, d_bcc); //get channel impulse response for given training sequence number - d_bcc
- detect_burst(input, &channel_imp_resp[0], d_c0_burst_start, output_binary); //MLSE detection of bits
- send_burst(d_burst_nr, output_binary, GSMTAP_BURST_NORMAL, input_nr);
- break;
- }
- case dummy_or_normal:
- {
- unsigned int normal_burst_start, dummy_burst_start;
- float dummy_corr_max, normal_corr_max;
-
- dummy_burst_start = get_norm_chan_imp_resp(input, &channel_imp_resp[0], &dummy_corr_max, TS_DUMMY);
- normal_burst_start = get_norm_chan_imp_resp(input, &channel_imp_resp[0], &normal_corr_max, d_bcc);
-
- if (normal_corr_max > dummy_corr_max)
- {
- d_c0_burst_start = normal_burst_start;
- detect_burst(input, &channel_imp_resp[0], normal_burst_start, output_binary);
- send_burst(d_burst_nr, output_binary, GSMTAP_BURST_NORMAL, input_nr);
- }
- else
- {
- d_c0_burst_start = dummy_burst_start;
- send_burst(d_burst_nr, dummy_burst, GSMTAP_BURST_DUMMY, input_nr);
- }
- break;
- }
- case rach_burst:
- break;
- case dummy:
- send_burst(d_burst_nr, dummy_burst, GSMTAP_BURST_DUMMY, input_nr);
- break;
- case normal_or_noise:
- {
- unsigned int burst_start;
- float normal_corr_max_tmp;
- float normal_corr_max=-1e6;
- int max_tn;
- std::vector<gr_complex> v(input, input + noutput_items);
- //if(d_signal_dbm>=d_c0_signal_dbm-13)
- {
- if(d_tseq_nums.size()==0) //there is no information about training sequence
- { //however the receiver can detect it
- get_norm_chan_imp_resp(input, &channel_imp_resp[0], &normal_corr_max, 0);
- float ts_max=normal_corr_max; //with use of a very simple algorithm based on finding
- int ts_max_num=0; //maximum correlation
- for(int ss=1; ss<=7; ss++)
- {
- get_norm_chan_imp_resp(input, &channel_imp_resp[0], &normal_corr_max, ss);
- if(ts_max<normal_corr_max)
- {
- ts_max = normal_corr_max;
- ts_max_num = ss;
- }
- }
- d_tseq_nums.push_back(ts_max_num);
- }
- int tseq_num;
- if(input_nr<=d_tseq_nums.size())
- {
- tseq_num = d_tseq_nums[input_nr-1];
- } else
- {
- tseq_num = d_tseq_nums.back();
- }
- burst_start = get_norm_chan_imp_resp(input, &channel_imp_resp[0], &normal_corr_max, tseq_num);
-// if(abs(d_c0_burst_start-burst_start)<=2){ //unused check/filter based on timing
- // if((normal_corr_max/sqrt(signal_pwr))>=0.9)
- {
- detect_burst(input, &channel_imp_resp[0], burst_start, output_binary);
- send_burst(d_burst_nr, output_binary, GSMTAP_BURST_NORMAL, input_nr);
- }
- }
- break;
- }
- case empty: //if it's empty burst
- break; //do nothing
- }
-
- if(input_nr==input_items.size()-1)
- {
- d_burst_nr++; //go to next burst
- to_consume += TS_BITS * d_OSR + d_burst_nr.get_offset(); //consume samples of the burst up to next guard period
- consume_each(to_consume);
- }
- //and add offset which is introduced by
- //0.25 fractional part of a guard period
+
+ d_tseq_nums.push_back(ts_max_num);
+ }
+
+ /* Choose proper training sequence number */
+ tseq_num = input_nr <= d_tseq_nums.size() ?
+ d_tseq_nums[input_nr - 1] : d_tseq_nums.back();
+
+ /* Get channel impulse response */
+ burst_start = get_norm_chan_imp_resp(input, &channel_imp_resp[0],
+ &normal_corr_max, tseq_num);
+
+ /* Perform MLSE detection */
+ detect_burst(input, &channel_imp_resp[0],
+ burst_start, output_binary);
+
+ /* Compose a message with GSMTAP header and bits */
+ send_burst(d_burst_nr, output_binary, GSMTAP_BURST_NORMAL, input_nr);
+
+ break;
}
- }
- break;
- }
- return 0;
-}
-bool receiver_impl::find_fcch_burst(const gr_complex *input, const int nitems, double & computed_freq_offset)
-{
- boost::circular_buffer<float> phase_diff_buffer(FCCH_HITS_NEEDED * d_OSR); //circular buffer used to scan throug signal to find
- //best match for FCCH burst
- float phase_diff = 0;
- gr_complex conjprod;
- int start_pos = -1;
- int hit_count = 0;
- int miss_count = 0;
- float min_phase_diff;
- float max_phase_diff;
- double best_sum = 0;
- float lowest_max_min_diff = 99999;
-
- int to_consume = 0;
- int sample_number = 0;
- bool end = false;
- bool result = false;
- boost::circular_buffer<float>::iterator buffer_iter;
-
- /**@name Possible states of FCCH search algorithm*/
- //@{
- enum states
- {
- init, ///< initialize variables
- search, ///< search for positive samples
- found_something, ///< search for FCCH and the best position of it
- fcch_found, ///< when FCCH was found
- search_fail ///< when there is no FCCH in the input vector
- } fcch_search_state;
- //@}
+ case dummy:
+ send_burst(d_burst_nr, dummy_burst, GSMTAP_BURST_DUMMY, input_nr);
+ break;
- fcch_search_state = init;
+ case rach_burst:
+ case empty:
+ /* Do nothing */
+ break;
+ }
- while (!end)
- {
- switch (fcch_search_state)
- {
+ if (input_nr == input_items.size() - 1) {
+ /* Go to the next burst */
+ d_burst_nr++;
- case init: //initialize variables
- hit_count = 0;
- miss_count = 0;
- start_pos = -1;
- lowest_max_min_diff = 99999;
- phase_diff_buffer.clear();
- fcch_search_state = search;
+ /* Consume samples of the burst up to next guard period */
+ to_consume += TS_BITS * d_OSR + d_burst_nr.get_offset();
+ consume_each(to_consume);
+ }
+ }
+ }
- break;
+ bool
+ receiver_impl::find_fcch_burst(const gr_complex *input,
+ const int nitems, double &computed_freq_offset)
+ {
+ /* Circular buffer used to scan through signal to find */
+ boost::circular_buffer<float>
+ phase_diff_buffer(FCCH_HITS_NEEDED * d_OSR);
+ boost::circular_buffer<float>::iterator buffer_iter;
+
+ float lowest_max_min_diff;
+ float phase_diff; /* Best match for FCCH burst */
+ float min_phase_diff;
+ float max_phase_diff;
+ double best_sum = 0;
+ gr_complex conjprod;
+ int start_pos;
+ int hit_count;
+ int miss_count;
+
+ int sample_number = 0;
+ int to_consume = 0;
+ bool result = false;
+ bool end = false;
+
+ /* Possible states of FCCH search algorithm */
+ enum states
+ {
+ init, /* initialize variables */
+ search, /* search for positive samples */
+ found_something, /* search for FCCH and the best position of it */
+ fcch_found, /* when FCCH was found */
+ search_fail /* when there is no FCCH in the input vector */
+ } fcch_search_state;
+
+ /* Set initial state */
+ fcch_search_state = init;
+
+ while (!end)
+ {
+ switch (fcch_search_state) {
+ case init:
+ {
+ hit_count = 0;
+ miss_count = 0;
+ start_pos = -1;
+ lowest_max_min_diff = 99999;
+ phase_diff_buffer.clear();
- case search: // search for positive samples
- sample_number++;
+ /* Change current state */
+ fcch_search_state = search;
- if (sample_number > nitems - FCCH_HITS_NEEDED * d_OSR) //if it isn't possible to find FCCH because
- {
- //there's too few samples left to look into,
- to_consume = sample_number; //don't do anything with those samples which are left
- //and consume only those which were checked
- fcch_search_state = search_fail;
- }
- else
- {
- phase_diff = compute_phase_diff(input[sample_number], input[sample_number-1]);
-
- if (phase_diff > 0) //if a positive phase difference was found
- {
- to_consume = sample_number;
- fcch_search_state = found_something; //switch to state in which searches for FCCH
- }
- else
- {
- fcch_search_state = search;
- }
- }
+ break;
+ }
+ case search:
+ {
+ sample_number++;
+
+ if (sample_number > nitems - FCCH_HITS_NEEDED * d_OSR) {
+ /**
+ * If it isn't possible to find FCCH, because
+ * there is too few samples left to look into,
+ * don't do anything with those samples which are left
+ * and consume only those which were checked
+ */
+ to_consume = sample_number;
+ fcch_search_state = search_fail;
break;
+ }
+
+ phase_diff = compute_phase_diff(input[sample_number],
+ input[sample_number - 1]);
+
+ /**
+ * If a positive phase difference was found
+ * switch to state in which searches for FCCH
+ */
+ if (phase_diff > 0) {
+ to_consume = sample_number;
+ fcch_search_state = found_something;
+ } else {
+ fcch_search_state = search;
+ }
+
+ break;
+ }
- case found_something: // search for FCCH and the best position of it
+ case found_something:
{
- if (phase_diff > 0)
- {
- hit_count++; //positive phase differencies increases hits_count
- }
- else
- {
- miss_count++; //negative increases miss_count
- }
-
- if ((miss_count >= FCCH_MAX_MISSES * d_OSR) && (hit_count <= FCCH_HITS_NEEDED * d_OSR))
- {
- //if miss_count exceeds limit before hit_count
- fcch_search_state = init; //go to init
- continue;
- }
- else if (((miss_count >= FCCH_MAX_MISSES * d_OSR) && (hit_count > FCCH_HITS_NEEDED * d_OSR)) || (hit_count > 2 * FCCH_HITS_NEEDED * d_OSR))
- {
- //if hit_count and miss_count exceeds limit then FCCH was found
- fcch_search_state = fcch_found;
- continue;
- }
- else if ((miss_count < FCCH_MAX_MISSES * d_OSR) && (hit_count > FCCH_HITS_NEEDED * d_OSR))
- {
- //find difference between minimal and maximal element in the buffer
- //for FCCH this value should be low
- //this part is searching for a region where this value is lowest
- min_phase_diff = * (min_element(phase_diff_buffer.begin(), phase_diff_buffer.end()));
- max_phase_diff = * (max_element(phase_diff_buffer.begin(), phase_diff_buffer.end()));
-
- if (lowest_max_min_diff > max_phase_diff - min_phase_diff)
- {
- lowest_max_min_diff = max_phase_diff - min_phase_diff;
- start_pos = sample_number - FCCH_HITS_NEEDED * d_OSR - FCCH_MAX_MISSES * d_OSR; //store start pos
- best_sum = 0;
-
- for (buffer_iter = phase_diff_buffer.begin();
- buffer_iter != (phase_diff_buffer.end());
- buffer_iter++)
- {
- best_sum += *buffer_iter - (M_PI / 2) / d_OSR; //store best value of phase offset sum
- }
- }
+ if (phase_diff > 0)
+ hit_count++;
+ else
+ miss_count++;
+
+ if ((miss_count >= FCCH_MAX_MISSES * d_OSR)
+ && (hit_count <= FCCH_HITS_NEEDED * d_OSR))
+ {
+ /* If miss_count exceeds limit before hit_count */
+ fcch_search_state = init;
+ continue;
+ }
+
+ if (((miss_count >= FCCH_MAX_MISSES * d_OSR)
+ && (hit_count > FCCH_HITS_NEEDED * d_OSR))
+ || (hit_count > 2 * FCCH_HITS_NEEDED * d_OSR))
+ {
+ /**
+ * If hit_count and miss_count exceeds
+ * limit then FCCH was found
+ */
+ fcch_search_state = fcch_found;
+ continue;
+ }
+
+ if ((miss_count < FCCH_MAX_MISSES * d_OSR)
+ && (hit_count > FCCH_HITS_NEEDED * d_OSR))
+ {
+ /**
+ * Find difference between minimal and maximal
+ * element in the buffer. For FCCH this value
+ * should be low. This part is searching for
+ * a region where this value is lowest.
+ */
+ min_phase_diff = *(min_element(phase_diff_buffer.begin(),
+ phase_diff_buffer.end()));
+ max_phase_diff = *(max_element(phase_diff_buffer.begin(),
+ phase_diff_buffer.end()));
+
+ if (lowest_max_min_diff > max_phase_diff - min_phase_diff) {
+ lowest_max_min_diff = max_phase_diff - min_phase_diff;
+ start_pos = sample_number - FCCH_HITS_NEEDED
+ * d_OSR - FCCH_MAX_MISSES * d_OSR;
+ best_sum = 0;
+
+ for (buffer_iter = phase_diff_buffer.begin();
+ buffer_iter != (phase_diff_buffer.end());
+ buffer_iter++) {
+ /* Store best value of phase offset sum */
+ best_sum += *buffer_iter - (M_PI / 2) / d_OSR;
+ }
}
+ }
- sample_number++;
+ /* If there is no single sample left to check */
+ if (++sample_number >= nitems) {
+ fcch_search_state = search_fail;
+ continue;
+ }
- if (sample_number >= nitems) //if there's no single sample left to check
- {
- fcch_search_state = search_fail;//FCCH search failed
- continue;
- }
+ phase_diff = compute_phase_diff(input[sample_number],
+ input[sample_number-1]);
+ phase_diff_buffer.push_back(phase_diff);
+ fcch_search_state = found_something;
- phase_diff = compute_phase_diff(input[sample_number], input[sample_number-1]);
- phase_diff_buffer.push_back(phase_diff);
- fcch_search_state = found_something;
+ break;
}
- break;
case fcch_found:
{
- to_consume = start_pos + FCCH_HITS_NEEDED * d_OSR + 1; //consume one FCCH burst
-
- d_fcch_start_pos = d_counter + start_pos;
-
- //compute frequency offset
- double phase_offset = best_sum / FCCH_HITS_NEEDED;
- double freq_offset = phase_offset * 1625000.0/6 / (2 * M_PI); //1625000.0/6 - GMSK symbol rate in GSM
- computed_freq_offset = freq_offset;
-
- end = true;
- result = true;
- break;
+ /* Consume one FCCH burst */
+ to_consume = start_pos + FCCH_HITS_NEEDED * d_OSR + 1;
+ d_fcch_start_pos = d_counter + start_pos;
+
+ /**
+ * Compute frequency offset
+ *
+ * 1625000.0 / 6 - GMSK symbol rate in GSM
+ */
+ double phase_offset = best_sum / FCCH_HITS_NEEDED;
+ double freq_offset = phase_offset * 1625000.0 / 6 / (2 * M_PI);
+ computed_freq_offset = freq_offset;
+
+ end = true;
+ result = true;
+
+ break;
}
case search_fail:
- end = true;
- result = false;
- break;
+ end = true;
+ result = false;
+ break;
}
- }
-
- d_counter += to_consume;
- consume_each(to_consume);
+ }
- return result;
-}
+ d_counter += to_consume;
+ consume_each(to_consume);
-double receiver_impl::compute_freq_offset(const gr_complex * input, unsigned first_sample, unsigned last_sample)
-{
- double phase_sum = 0;
- unsigned ii;
+ return result;
+ }
- for (ii = first_sample; ii < last_sample; ii++)
+ double
+ receiver_impl::compute_freq_offset(const gr_complex * input,
+ unsigned first_sample, unsigned last_sample)
{
- double phase_diff = compute_phase_diff(input[ii], input[ii-1]) - (M_PI / 2) / d_OSR;
- phase_sum += phase_diff;
- }
+ double phase_sum = 0;
+ unsigned ii;
- double phase_offset = phase_sum / (last_sample - first_sample);
- double freq_offset = phase_offset * 1625000.0 / (12.0 * M_PI);
- return freq_offset;
-}
+ for (ii = first_sample; ii < last_sample; ii++)
+ {
+ double phase_diff = compute_phase_diff(input[ii],
+ input[ii-1]) - (M_PI / 2) / d_OSR;
+ phase_sum += phase_diff;
+ }
-inline float receiver_impl::compute_phase_diff(gr_complex val1, gr_complex val2)
-{
- gr_complex conjprod = val1 * conj(val2);
- return fast_atan2f(imag(conjprod), real(conjprod));
-}
+ double phase_offset = phase_sum / (last_sample - first_sample);
+ double freq_offset = phase_offset * 1625000.0 / (12.0 * M_PI);
-bool receiver_impl::reach_sch_burst(const int nitems)
-{
- //it just consumes samples to get near to a SCH burst
- int to_consume = 0;
- bool result = false;
- unsigned sample_nr_near_sch_start = d_fcch_start_pos + (FRAME_BITS - SAFETY_MARGIN) * d_OSR;
+ return freq_offset;
+ }
- //consume samples until d_counter will be equal to sample_nr_near_sch_start
- if (d_counter < sample_nr_near_sch_start)
+ inline float
+ receiver_impl::compute_phase_diff(gr_complex val1, gr_complex val2)
{
- if (d_counter + nitems >= sample_nr_near_sch_start)
- {
- to_consume = sample_nr_near_sch_start - d_counter;
- }
- else
- {
- to_consume = nitems;
- }
- result = false;
+ gr_complex conjprod = val1 * conj(val2);
+ return fast_atan2f(imag(conjprod), real(conjprod));
}
- else
+
+ bool
+ receiver_impl::reach_sch_burst(const int nitems)
{
+ /* It just consumes samples to get near to a SCH burst */
+ int to_consume = 0;
+ bool result = false;
+ unsigned sample_nr = d_fcch_start_pos
+ + (FRAME_BITS - SAFETY_MARGIN) * d_OSR;
+
+ /* Consume samples until d_counter will be equal to sample_nr */
+ if (d_counter < sample_nr) {
+ to_consume = d_counter + nitems >= sample_nr ?
+ sample_nr - d_counter : nitems;
+ } else {
to_consume = 0;
result = true;
- }
-
- d_counter += to_consume;
- consume_each(to_consume);
- return result;
-}
+ }
-int receiver_impl::get_sch_chan_imp_resp(const gr_complex *input, gr_complex * chan_imp_resp)
-{
- std::vector<gr_complex> correlation_buffer;
- std::vector<float> power_buffer;
- std::vector<float> window_energy_buffer;
+ d_counter += to_consume;
+ consume_each(to_consume);
- int strongest_window_nr;
- int burst_start = 0;
- int chan_imp_resp_center = 0;
- float max_correlation = 0;
- float energy = 0;
+ return result;
+ }
- for (int ii = SYNC_POS * d_OSR; ii < (SYNC_POS + SYNC_SEARCH_RANGE) *d_OSR; ii++)
+ int
+ receiver_impl::get_sch_chan_imp_resp(const gr_complex *input,
+ gr_complex * chan_imp_resp)
{
- gr_complex correlation = correlate_sequence(&d_sch_training_seq[5], N_SYNC_BITS - 10, &input[ii]);
+ std::vector<gr_complex> correlation_buffer;
+ std::vector<float> window_energy_buffer;
+ std::vector<float> power_buffer;
+
+ int chan_imp_resp_center = 0;
+ int strongest_window_nr;
+ int burst_start;
+ float energy = 0;
+
+ int len = (SYNC_POS + SYNC_SEARCH_RANGE) * d_OSR;
+ for (int ii = SYNC_POS * d_OSR; ii < len; ii++) {
+ gr_complex correlation = correlate_sequence(&d_sch_training_seq[5],
+ N_SYNC_BITS - 10, &input[ii]);
correlation_buffer.push_back(correlation);
power_buffer.push_back(std::pow(abs(correlation), 2));
- }
- //compute window energies
- std::vector<float>::iterator iter = power_buffer.begin();
- bool loop_end = false;
- while (iter != power_buffer.end())
- {
+ }
+
+ /* Compute window energies */
+ std::vector<float>::iterator iter = power_buffer.begin();
+ while (iter != power_buffer.end()) {
std::vector<float>::iterator iter_ii = iter;
+ bool loop_end = false;
energy = 0;
- for (int ii = 0; ii < (d_chan_imp_length) *d_OSR; ii++, iter_ii++)
- {
- if (iter_ii == power_buffer.end())
- {
- loop_end = true;
- break;
- }
- energy += (*iter_ii);
- }
- if (loop_end)
- {
+ for (int ii = 0; ii < (d_chan_imp_length) * d_OSR; ii++, iter_ii++) {
+ if (iter_ii == power_buffer.end()) {
+ loop_end = true;
break;
+ }
+
+ energy += (*iter_ii);
}
- iter++;
+
+ if (loop_end)
+ break;
+
window_energy_buffer.push_back(energy);
- }
+ iter++;
+ }
- strongest_window_nr = max_element(window_energy_buffer.begin(), window_energy_buffer.end()) - window_energy_buffer.begin();
- // d_channel_imp_resp.clear();
+ strongest_window_nr = max_element(window_energy_buffer.begin(),
+ window_energy_buffer.end()) - window_energy_buffer.begin();
- max_correlation = 0;
- for (int ii = 0; ii < (d_chan_imp_length) *d_OSR; ii++)
- {
+#if 0
+ d_channel_imp_resp.clear();
+#endif
+
+ float max_correlation = 0;
+ for (int ii = 0; ii < (d_chan_imp_length) * d_OSR; ii++) {
gr_complex correlation = correlation_buffer[strongest_window_nr + ii];
- if (abs(correlation) > max_correlation)
- {
- chan_imp_resp_center = ii;
- max_correlation = abs(correlation);
+ if (abs(correlation) > max_correlation) {
+ chan_imp_resp_center = ii;
+ max_correlation = abs(correlation);
}
- // d_channel_imp_resp.push_back(correlation);
+
+#if 0
+ d_channel_imp_resp.push_back(correlation);
+#endif
+
chan_imp_resp[ii] = correlation;
- }
+ }
- burst_start = strongest_window_nr + chan_imp_resp_center - 48 * d_OSR - 2 * d_OSR + 2 + SYNC_POS * d_OSR;
- return burst_start;
-}
+ burst_start = strongest_window_nr + chan_imp_resp_center
+ - 48 * d_OSR - 2 * d_OSR + 2 + SYNC_POS * d_OSR;
+ return burst_start;
+ }
-void receiver_impl::detect_burst(const gr_complex * input, gr_complex * chan_imp_resp, int burst_start, unsigned char * output_binary)
-{
- float output[BURST_SIZE];
- std::vector<gr_complex> rhh_temp(CHAN_IMP_RESP_LENGTH*d_OSR);
- gr_complex rhh[CHAN_IMP_RESP_LENGTH];
- gr_complex filtered_burst[BURST_SIZE];
- int start_state = 3;
- unsigned int stop_states[2] = {4, 12};
-
- autocorrelation(chan_imp_resp, &rhh_temp[0], d_chan_imp_length*d_OSR);
- for (int ii = 0; ii < (d_chan_imp_length); ii++)
+ void
+ receiver_impl::detect_burst(const gr_complex * input,
+ gr_complex * chan_imp_resp, int burst_start,
+ unsigned char * output_binary)
{
+ std::vector<gr_complex> rhh_temp(CHAN_IMP_RESP_LENGTH * d_OSR);
+ unsigned int stop_states[2] = {4, 12};
+ gr_complex filtered_burst[BURST_SIZE];
+ gr_complex rhh[CHAN_IMP_RESP_LENGTH];
+ float output[BURST_SIZE];
+ int start_state = 3;
+
+ autocorrelation(chan_imp_resp, &rhh_temp[0], d_chan_imp_length*d_OSR);
+ for (int ii = 0; ii < d_chan_imp_length; ii++)
rhh[ii] = conj(rhh_temp[ii*d_OSR]);
- }
- mafi(&input[burst_start], BURST_SIZE, chan_imp_resp, d_chan_imp_length*d_OSR, filtered_burst);
+ mafi(&input[burst_start], BURST_SIZE, chan_imp_resp,
+ d_chan_imp_length * d_OSR, filtered_burst);
- viterbi_detector(filtered_burst, BURST_SIZE, rhh, start_state, stop_states, 2, output);
+ viterbi_detector(filtered_burst, BURST_SIZE, rhh,
+ start_state, stop_states, 2, output);
- for (int i = 0; i < BURST_SIZE ; i++)
- {
- output_binary[i] = (output[i] > 0);
+ for (int i = 0; i < BURST_SIZE; i++)
+ output_binary[i] = output[i] > 0;
}
-}
-void receiver_impl::gmsk_mapper(const unsigned char * input, int nitems, gr_complex * gmsk_output, gr_complex start_point)
-{
- gr_complex j = gr_complex(0.0, 1.0);
+ void
+ receiver_impl::gmsk_mapper(const unsigned char * input,
+ int nitems, gr_complex * gmsk_output, gr_complex start_point)
+ {
+ gr_complex j = gr_complex(0.0, 1.0);
+ gmsk_output[0] = start_point;
- int current_symbol;
- int encoded_symbol;
- int previous_symbol = 2 * input[0] - 1;
- gmsk_output[0] = start_point;
+ int previous_symbol = 2 * input[0] - 1;
+ int current_symbol;
+ int encoded_symbol;
- for (int i = 1; i < nitems; i++)
- {
- //change bits representation to NRZ
+ for (int i = 1; i < nitems; i++) {
+ /* Change bits representation to NRZ */
current_symbol = 2 * input[i] - 1;
- //differentially encode
+
+ /* Differentially encode */
encoded_symbol = current_symbol * previous_symbol;
- //and do gmsk mapping
- gmsk_output[i] = j * gr_complex(encoded_symbol, 0.0) * gmsk_output[i-1];
+
+ /* And do GMSK mapping */
+ gmsk_output[i] = j * gr_complex(encoded_symbol, 0.0)
+ * gmsk_output[i-1];
+
previous_symbol = current_symbol;
+ }
}
-}
-
-gr_complex receiver_impl::correlate_sequence(const gr_complex * sequence, int length, const gr_complex * input)
-{
- gr_complex result(0.0, 0.0);
- int sample_number = 0;
- for (int ii = 0; ii < length; ii++)
+ gr_complex
+ receiver_impl::correlate_sequence(const gr_complex * sequence,
+ int length, const gr_complex * input)
{
- sample_number = (ii * d_OSR) ;
- result += sequence[ii] * conj(input[sample_number]);
- }
+ gr_complex result(0.0, 0.0);
- result = result / gr_complex(length, 0);
- return result;
-}
+ for (int ii = 0; ii < length; ii++)
+ result += sequence[ii] * conj(input[ii * d_OSR]);
-//computes autocorrelation for positive arguments
-inline void receiver_impl::autocorrelation(const gr_complex * input, gr_complex * out, int nitems)
-{
- int i, k;
- for (k = nitems - 1; k >= 0; k--)
+ return result / gr_complex(length, 0);
+ }
+
+ /* Computes autocorrelation for positive arguments */
+ inline void
+ receiver_impl::autocorrelation(const gr_complex * input,
+ gr_complex * out, int nitems)
{
+ for (int k = nitems - 1; k >= 0; k--) {
out[k] = gr_complex(0, 0);
- for (i = k; i < nitems; i++)
- {
- out[k] += input[i] * conj(input[i-k]);
- }
+ for (int i = k; i < nitems; i++)
+ out[k] += input[i] * conj(input[i - k]);
+ }
}
-}
-
-inline void receiver_impl::mafi(const gr_complex * input, int nitems, gr_complex * filter, int filter_length, gr_complex * output)
-{
- int ii = 0, n, a;
- for (n = 0; n < nitems; n++)
+ inline void
+ receiver_impl::mafi(const gr_complex * input, int nitems,
+ gr_complex * filter, int filter_length, gr_complex * output)
{
- a = n * d_OSR;
+ for (int n = 0; n < nitems; n++) {
+ int a = n * d_OSR;
output[n] = 0;
- ii = 0;
- while (ii < filter_length)
- {
- if ((a + ii) >= nitems*d_OSR){
- break;
- }
- output[n] += input[a+ii] * filter[ii];
- ii++;
+ for (int ii = 0; ii < filter_length; ii++) {
+ if ((a + ii) >= nitems * d_OSR)
+ break;
+
+ output[n] += input[a + ii] * filter[ii];
}
+ }
}
-}
-//especially computations of strongest_window_nr
-int receiver_impl::get_norm_chan_imp_resp(const gr_complex *input, gr_complex * chan_imp_resp, float *corr_max, int bcc)
-{
- std::vector<gr_complex> correlation_buffer;
- std::vector<float> power_buffer;
- std::vector<float> window_energy_buffer;
-
- int strongest_window_nr;
- int burst_start = 0;
- int chan_imp_resp_center = 0;
- float max_correlation = 0;
- float energy = 0;
-
- int search_center = (int)((TRAIN_POS + GUARD_PERIOD) * d_OSR);
- int search_start_pos = search_center + 1 - 5*d_OSR;
- // int search_start_pos = search_center - d_chan_imp_length * d_OSR;
- int search_stop_pos = search_center + d_chan_imp_length * d_OSR + 5 * d_OSR;
-
- for(int ii = search_start_pos; ii < search_stop_pos; ii++)
- {
- gr_complex correlation = correlate_sequence(&d_norm_training_seq[bcc][TRAIN_BEGINNING], N_TRAIN_BITS - 10, &input[ii]);
- correlation_buffer.push_back(correlation);
- power_buffer.push_back(std::pow(abs(correlation), 2));
- }
-// plot(power_buffer);
- //compute window energies
- std::vector<float>::iterator iter = power_buffer.begin();
- bool loop_end = false;
- while (iter != power_buffer.end())
+ /* Especially computations of strongest_window_nr */
+ int
+ receiver_impl::get_norm_chan_imp_resp(const gr_complex *input,
+ gr_complex *chan_imp_resp, float *corr_max, int bcc)
{
+ std::vector<gr_complex> correlation_buffer;
+ std::vector<float> window_energy_buffer;
+ std::vector<float> power_buffer;
+
+ int search_center = (int) (TRAIN_POS + GUARD_PERIOD) * d_OSR;
+ int search_start_pos = search_center + 1 - 5 * d_OSR;
+ int search_stop_pos = search_center
+ + d_chan_imp_length * d_OSR + 5 * d_OSR;
+
+ for (int ii = search_start_pos; ii < search_stop_pos; ii++) {
+ gr_complex correlation = correlate_sequence(
+ &d_norm_training_seq[bcc][TRAIN_BEGINNING],
+ N_TRAIN_BITS - 10, &input[ii]);
+ correlation_buffer.push_back(correlation);
+ power_buffer.push_back(std::pow(abs(correlation), 2));
+ }
+
+#if 0
+ plot(power_buffer);
+#endif
+
+ /* Compute window energies */
+ std::vector<float>::iterator iter = power_buffer.begin();
+ while (iter != power_buffer.end()) {
std::vector<float>::iterator iter_ii = iter;
- energy = 0;
+ bool loop_end = false;
+ float energy = 0;
- for (int ii = 0; ii < (d_chan_imp_length)*d_OSR; ii++, iter_ii++)
- {
- if (iter_ii == power_buffer.end())
- {
- loop_end = true;
- break;
- }
- energy += (*iter_ii);
- }
- if (loop_end)
- {
+ int len = d_chan_imp_length * d_OSR;
+ for (int ii = 0; ii < len; ii++, iter_ii++) {
+ if (iter_ii == power_buffer.end()) {
+ loop_end = true;
break;
+ }
+
+ energy += (*iter_ii);
}
- iter++;
+
+ if (loop_end)
+ break;
window_energy_buffer.push_back(energy);
- }
+ iter++;
+ }
- strongest_window_nr = max_element(window_energy_buffer.begin(), window_energy_buffer.end()-((d_chan_imp_length)*d_OSR)) - window_energy_buffer.begin();
- //strongest_window_nr = strongest_window_nr-d_OSR;
- if(strongest_window_nr<0){
- strongest_window_nr = 0;
- }
-
- max_correlation = 0;
- for (int ii = 0; ii < (d_chan_imp_length)*d_OSR; ii++)
- {
+ /* Calculate the strongest window number */
+ int strongest_window_nr = max_element(window_energy_buffer.begin(),
+ window_energy_buffer.end() - d_chan_imp_length * d_OSR)
+ - window_energy_buffer.begin();
+
+ if (strongest_window_nr < 0)
+ strongest_window_nr = 0;
+
+ float max_correlation = 0;
+ for (int ii = 0; ii < d_chan_imp_length * d_OSR; ii++) {
gr_complex correlation = correlation_buffer[strongest_window_nr + ii];
if (abs(correlation) > max_correlation)
- {
- chan_imp_resp_center = ii;
- max_correlation = abs(correlation);
- }
- // d_channel_imp_resp.push_back(correlation);
- chan_imp_resp[ii] = correlation;
- }
-
- *corr_max = max_correlation;
+ max_correlation = abs(correlation);
- //DCOUT("strongest_window_nr_new: " << strongest_window_nr);
- burst_start = search_start_pos + strongest_window_nr - TRAIN_POS * d_OSR; //compute first sample posiiton which corresponds to the first sample of the impulse response
-
- //DCOUT("burst_start: " << burst_start);
- return burst_start;
-}
+#if 0
+ d_channel_imp_resp.push_back(correlation);
+#endif
+ chan_imp_resp[ii] = correlation;
+ }
+
+ *corr_max = max_correlation;
-void receiver_impl::send_burst(burst_counter burst_nr, const unsigned char * burst_binary, uint8_t burst_type, unsigned int input_nr)
-{
- boost::scoped_ptr<gsmtap_hdr> tap_header(new gsmtap_hdr());
-
- tap_header->version = GSMTAP_VERSION;
- tap_header->hdr_len = sizeof(gsmtap_hdr)/4;
- tap_header->type = GSMTAP_TYPE_UM_BURST;
- tap_header->sub_type = burst_type;
- bool uplink_burst = (input_nr >= d_cell_allocation.size());
- if(!uplink_burst) // downlink burst
- {
- tap_header->timeslot = static_cast<uint8_t>(d_burst_nr.get_timeslot_nr());
- tap_header->frame_number = htobe32(d_burst_nr.get_frame_nr());
- tap_header->arfcn = htobe16(d_cell_allocation[input_nr]) ;
+ /**
+ * Compute first sample position, which corresponds
+ * to the first sample of the impulse response
+ */
+ return search_start_pos + strongest_window_nr - TRAIN_POS * d_OSR;
}
- else //uplink burst
+
+
+ void
+ receiver_impl::send_burst(burst_counter burst_nr,
+ const unsigned char * burst_binary, uint8_t burst_type,
+ unsigned int input_nr)
{
- tap_header->timeslot = static_cast<uint8_t>(d_burst_nr.subtract_timeslots(3).get_timeslot_nr());
- tap_header->frame_number = htobe32(d_burst_nr.subtract_timeslots(3).get_frame_nr());
- input_nr = input_nr - d_cell_allocation.size();
- tap_header->arfcn = htobe16(d_cell_allocation[input_nr] | 0x4000);
- }
- tap_header->signal_dbm = static_cast<int8_t>(d_signal_dbm);
- tap_header->snr_db = 0;
-
- int8_t header_plus_burst[sizeof(gsmtap_hdr)+BURST_SIZE];
- memcpy(header_plus_burst, tap_header.get(), sizeof(gsmtap_hdr));
- memcpy(header_plus_burst+sizeof(gsmtap_hdr), burst_binary, BURST_SIZE);
-
- pmt::pmt_t blob_header_plus_burst = pmt::make_blob(header_plus_burst,sizeof(gsmtap_hdr)+BURST_SIZE);
- pmt::pmt_t msg = pmt::cons(pmt::PMT_NIL, blob_header_plus_burst);
-
- if(input_nr==0){
+ /* Buffer for GSMTAP header and burst */
+ uint8_t buf[sizeof(gsmtap_hdr) + BURST_SIZE];
+ uint32_t frame_number;
+ uint16_t arfcn;
+ uint8_t tn;
+
+ /* Set pointers to GSMTAP header and burst inside buffer */
+ struct gsmtap_hdr *tap_header = (struct gsmtap_hdr *) buf;
+ uint8_t *burst = buf + sizeof(gsmtap_hdr);
+
+ tap_header->version = GSMTAP_VERSION;
+ tap_header->hdr_len = sizeof(gsmtap_hdr) / 4;
+ tap_header->type = GSMTAP_TYPE_UM_BURST;
+ tap_header->sub_type = burst_type;
+
+ bool dl_burst = !(input_nr >= d_cell_allocation.size());
+ if (dl_burst) {
+ tn = static_cast<uint8_t>(d_burst_nr.get_timeslot_nr());
+ frame_number = htobe32(d_burst_nr.get_frame_nr());
+ arfcn = htobe16(d_cell_allocation[input_nr]);
+ } else {
+ input_nr -= d_cell_allocation.size();
+ tn = static_cast<uint8_t>
+ (d_burst_nr.subtract_timeslots(3).get_timeslot_nr());
+ frame_number = htobe32(
+ d_burst_nr.subtract_timeslots(3).get_frame_nr());
+ arfcn = htobe16(
+ d_cell_allocation[input_nr] | GSMTAP_ARFCN_F_UPLINK);
+ }
+
+ tap_header->frame_number = frame_number;
+ tap_header->timeslot = tn;
+ tap_header->arfcn = arfcn;
+
+ tap_header->signal_dbm = static_cast<int8_t>(d_signal_dbm);
+ tap_header->snr_db = 0; /* FIXME: Can we calculate this? */
+
+ /* Copy burst to the buffer */
+ memcpy(burst, burst_binary, BURST_SIZE);
+
+ /* Allocate a new message */
+ pmt::pmt_t blob = pmt::make_blob(buf, sizeof(gsmtap_hdr) + BURST_SIZE);
+ pmt::pmt_t msg = pmt::cons(pmt::PMT_NIL, blob);
+
+ /* Send message */
+ if (input_nr == 0)
message_port_pub(pmt::mp("C0"), msg);
- } else {
+ else
message_port_pub(pmt::mp("CX"), msg);
}
-}
-void receiver_impl::configure_receiver()
-{
- d_channel_conf.set_multiframe_type(TIMESLOT0, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT0, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
-
- d_channel_conf.set_burst_types(TIMESLOT0, TEST_CCH_FRAMES, sizeof(TEST_CCH_FRAMES) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_burst_types(TIMESLOT0, FCCH_FRAMES, sizeof(FCCH_FRAMES) / sizeof(unsigned), fcch_burst);
- d_channel_conf.set_burst_types(TIMESLOT0, SCH_FRAMES, sizeof(SCH_FRAMES) / sizeof(unsigned), sch_burst);
-
- d_channel_conf.set_multiframe_type(TIMESLOT1, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT1, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_multiframe_type(TIMESLOT2, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT2, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_multiframe_type(TIMESLOT3, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT3, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_multiframe_type(TIMESLOT4, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT4, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_multiframe_type(TIMESLOT5, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT5, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_multiframe_type(TIMESLOT6, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT6, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
- d_channel_conf.set_multiframe_type(TIMESLOT7, multiframe_51);
- d_channel_conf.set_burst_types(TIMESLOT7, TEST51, sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
-}
-
-void receiver_impl::set_cell_allocation(const std::vector<int> &cell_allocation)
-{
- d_cell_allocation = cell_allocation;
-}
+ void
+ receiver_impl::configure_receiver(void)
+ {
+ d_channel_conf.set_multiframe_type(TIMESLOT0, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT0, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+ d_channel_conf.set_burst_types(TIMESLOT0, TEST_CCH_FRAMES,
+ sizeof(TEST_CCH_FRAMES) / sizeof(unsigned), dummy_or_normal);
+ d_channel_conf.set_burst_types(TIMESLOT0, FCCH_FRAMES,
+ sizeof(FCCH_FRAMES) / sizeof(unsigned), fcch_burst);
+ d_channel_conf.set_burst_types(TIMESLOT0, SCH_FRAMES,
+ sizeof(SCH_FRAMES) / sizeof(unsigned), sch_burst);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT1, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT1, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT2, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT2, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT3, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT3, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT4, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT4, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT5, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT5, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT6, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT6, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+
+ d_channel_conf.set_multiframe_type(TIMESLOT7, multiframe_51);
+ d_channel_conf.set_burst_types(TIMESLOT7, TEST51,
+ sizeof(TEST51) / sizeof(unsigned), dummy_or_normal);
+ }
-void receiver_impl::set_tseq_nums(const std::vector<int> & tseq_nums)
-{
- d_tseq_nums = tseq_nums;
-}
+ void
+ receiver_impl::set_cell_allocation(
+ const std::vector<int> &cell_allocation)
+ {
+ d_cell_allocation = cell_allocation;
+ }
-void receiver_impl::reset()
-{
- d_state = fcch_search;
-}
+ void
+ receiver_impl::set_tseq_nums(const std::vector<int> &tseq_nums)
+ {
+ d_tseq_nums = tseq_nums;
+ }
-} /* namespace gsm */
-} /* namespace gr */
+ void
+ receiver_impl::reset(void)
+ {
+ d_state = fcch_search;
+ }
+ } /* namespace gsm */
+} /* namespace gr */
diff --git a/lib/receiver/receiver_impl.h b/lib/receiver/receiver_impl.h
index 6074dd5..23aa085 100644
--- a/lib/receiver/receiver_impl.h
+++ b/lib/receiver/receiver_impl.h
@@ -63,6 +63,7 @@ namespace gr {
/**@name Variables used to store result of the find_fcch_burst fuction */
//@{
+ bool d_freq_offset_tag_in_fcch; ///< frequency offset tag presence
unsigned d_fcch_start_pos; ///< position of the first sample of the fcch burst
float d_freq_offset_setting; ///< frequency offset set in frequency shifter located upstream
//@}
@@ -200,9 +201,13 @@ namespace gr {
* Configures burst types in different channels
*/
void configure_receiver();
-
-
+ /* State machine handlers */
+ void fcch_search_handler(gr_complex *input, int noutput_items);
+ void sch_search_handler(gr_complex *input, int noutput_items);
+ void synchronized_handler(gr_complex *input,
+ gr_vector_const_void_star &input_items, int noutput_items);
+
public:
receiver_impl(int osr, const std::vector<int> &cell_allocation, const std::vector<int> &tseq_nums, bool process_uplink);
~receiver_impl();