Project

General

Profile

OsmoTRX » History » Version 32

ttsou, 02/19/2016 10:47 PM

1 1 ttsou
= OsmoTRX =
2
3
OsmoTRX is a software-defined radio transceiver that implements the Layer 1 physical layer of a BTS comprising the following 3GPP specifications:
4
 * TS 05.01 "Physical layer on the radio path"
5
 * TS 05.02 "Multiplexing and Multiple Access on the Radio Path"
6
 * TS 05.04 "Modulation"
7
 * TS 05.10 "Radio subsystem synchronization"
8
9 29 ttsou
OsmoTRX is based on the OpenBTS transceiver, but setup to operate independently with the purpose of using with non-OpenBTS software and projects. Currently there are numerous features contained in OsmoTRX that extend the functionality of the OpenBTS transceiver. These features include enhanced support for various embedded platforms - notably ARM - and dual channel diversity support for the Fairwaves UmTRX. Most of these features will eventually be merged into mainline OpenBTS, but development will occur primarily on OsmoTRX.
10 6 ttsou
11
== Features ==
12
13 16 ttsou
'''Intel SSE Support'''
14 6 ttsou
* SSE3
15
* SSE4.1
16
17 20 ttsou
On Intel processors, OsmoTRX makes heavy use of the Streaming SIMD Extensions (SSE) instruction set. Accelerated operations include pulse shape filtering, resampling, sequence correlation, and many other signal processing operations. SSE3 is the minimum requirement for accelerated use.
18 1 ttsou
19 26 ttsou
SSE3 is present in the majority of Intel processors since later versions of the Pentium 4 architecture and is also present on low power Atom processors. Support is automatically detected at build time. For additional performance information, please see the performance and benchmarks section.
20 20 ttsou
21 29 ttsou
'''ARM Support'''
22 6 ttsou
* NEON
23 1 ttsou
* NEON-VFPv4
24 6 ttsou
25 20 ttsou
OsmoTRX runs on a variety of ARM processors with and without NEON coprocessors. Like SSE on Intel processors, NEON provides acceleration with SIMD vectorized instructions.
26 1 ttsou
27 20 ttsou
Tested popular architectures include ARM11 (Raspberry Pi), Cortex-A8 (!BeagleBoard), and Cortex-A15 (!ArndaleBoard). Loosely speaking, these platforms are representative of low cost embedded devices, mid-level handsets, and high-end smartphones respectively. Similarly, in order, these platforms include no NEON coprocessor, standard NEON, and NEON-VFPv4. The latter NEON variation, VFPv4, provides additional fused-multiply-accumulate (FMA) instructions useful for many DSP operations.
28
29 26 ttsou
NEON support must be enabled by the user at build time. For additional information, please see the configuration and performance and benchmarks sections.
30 20 ttsou
31 6 ttsou
'''Dual Channel (UmTRX only)'''
32 7 ttsou
33
Two dual channel modes are available: standard dual channel mode and diversity. In standard dual channel mode, each RF
34
path of the dual channel device - currently only UmTRX - supports a different ARFCN. Each path operates independently a
35 28 ttsou
nd operates similarly to two separate devices. GSM channel capacity in this mode is doubled. This option can be configured at run time from the command line.
36 1 ttsou
37 8 ttsou
'''Dual Channel Diversity (UmTRX only)'''
38 1 ttsou
39 28 ttsou
Diversity mode is similar to the standard dual channel mode except each antenna supports both ARFCN channels. In this case, the receiver sample bandwidth is widened to handle both ARFCN's and subsequently converted and demultiplexed into separate sample streams. Each GSM receive path is fed dual signals, where antenna selection diversity is performed by taking the stronger signal on a burst-by-burst basis. This diversity setup improves uplink reception performance in multipath fading environments.
40 16 ttsou
41 28 ttsou
Limitations are increased CPU utilization and that ARFCN spacing is restricted (currently at 400 kHz) by the receiver sampling bandwidth. Setting the ARFCN spacing beyond the sampling limit will disable the diversity path and operate in standard dual channel mode. This options can be configured at run time from the command line.
42 20 ttsou
43 30 ttsou
'''High Performance Receiver'''
44
45
OsmoTRX utilizes a recently updated receive burst detection algorithm that provides greater sensitivity and reliability than the previous approach, which relied on energy detection for the initial stage of burst acquisition.
46
47
The limitation of the previous approach was that it was slow to adapt to highly transient power levels and false burst detection in challenging situations such as receiver saturation, which may occur in close range lab testing. The other issue was that a high degree of level tuning was often necessary to operate reliably.
48
49
The current receiver code removes those limitations. Noise and signal level measurements are also now handled in a more responsive manner.
50
51 16 ttsou
'''Low Phase Error Modulator'''
52
53 1 ttsou
The default GSM downlink signal is configured for low distortion using a linearized GMSK modulator. The implementation is based on a two pulse Laurent approximation of continuous phase modulated (CPM) signals. The baseband output signal measures with very low phase error and is capable of passing industry spectrum mask requirements. Please note that actual performance will depend strongly on the particular device in use.
54
55
Theoretical details can be found in the report on [http://tsou.cc/gsm/report_gmsk.pdf GMSK]. Octave / Matlab code for [http://tsou.cc/gsm/laurent.m pulse generation] is also available.
56 28 ttsou
57
This option can be enabled or disabled at run time from the command line.
58 16 ttsou
59 20 ttsou
Very Low Phase Error (Ettus Research N200)
60 1 ttsou
61
[[Image(http://tsou.cc/gsm/osmo-trx-phase75.gif)]]
62
63 21 ttsou
Spectrum Mask (Ettus Research N200)
64 1 ttsou
65
[[Image(http://tsou.cc/gsm/osmo-trx-spectrum75.gif)]]
66
67 20 ttsou
== RF Hardware support ==
68 1 ttsou
69 20 ttsou
Multiple RF devices are currently supported. These include USRP family products from Ettus Research, and the UmTRX from Fairwaves.
70 1 ttsou
71 20 ttsou
||'''Fairwaves'''||'''Notes'''||
72
||UmTRX||Dual channel||
73
74
All Ettus Research devices are supported.
75
76
||'''Ettus Research'''||'''Notes'''||
77
||USRP1||Requires legacy libusrp driver and clocking modification||
78
||USRP2||10 MHz external reference required||
79 1 ttsou
||B100||
80
||B110||
81 20 ttsou
||B200||10 MHz external reference recommended||
82
||B210||* Dual channel, 10 MHz external reference recommended||
83 1 ttsou
||N200||
84
||N210||
85
||E100||
86
||E110||
87
88 20 ttsou
* Ettus B210 dual channel support with OsmoTRX is currently unavailable, but is expected to be added at a later time.
89
90 1 ttsou
== Embedded Platform Support ==
91
92 20 ttsou
OsmoTRX has been tested on the multiple embedded platforms representing a wide range of device types. Low cost ARM devices are generally limited by memory and I/O as much CPU utilization.
93 1 ttsou
94 20 ttsou
Running a full or near full ARFCN configuration (7 simultaneous TCH channels with Combination V) may require running the GSM stack remotely, which can be configured at runtime on the command line. This limitation appears to be scheduling related more so than lack of CPU resources, and may be resolved at a later time.
95
96
||'''Platform'''||'''SoC'''||'''Processor'''||'''SIMD/FPU'''||'''Testing Notes'''
97
||!ArndaleBoard||Samsung Exynos 5250||ARM Cortex-A15||NEON-VFPv4||7 TCH||
98 21 ttsou
||!BeagleBoard-xM||Texas Instruments OMAP3||ARM Cortex-A8||NEON||7 TCH, remote OsmoBTS stack||
99
||Ettus E100||Texas Instruments OMAP3||ARM Cortex-A8||NEON||7 TCH, remote OsmoBTS stack||
100
||Raspberry Pi||Broadcom BCM2835||ARM11||VFP||2 TCH, remote OsmoBTS stack||
101 1 ttsou
||Shuttle PC||NA||Intel Atom D2550||SSE3||Dual channel, 15 TCH||
102 20 ttsou
103 25 ttsou
All embedded plaforms were tested with low-phase error modulator disabled. Use of the more accurate modulator on embedded platforms has not been extensively tested.
104 19 ttsou
105
== Mailing List ==
106 18 ttsou
107 1 ttsou
For development purposes, OsmoTRX is discussed on both OpenBTS and OpenBSC mailing lists at openbts-discuss@lists.sourceforge.net and openbsc@lists.osmocom.org respectively.
108
109 22 ttsou
Please direct questions and bug reports to the list appropriate for the GSM stack being used.
110 19 ttsou
111 16 ttsou
Subscription information is available at [https://lists.sourceforge.net/lists/listinfo/openbts-discuss] and [http://lists.osmocom.org/mailman/listinfo/openbsc/].
112 19 ttsou
113
== GPRS support ==
114 16 ttsou
115 1 ttsou
OsmoTRX supports GPRS through OsmoBTS.
116 16 ttsou
117 1 ttsou
For GPRS support with OpenBTS, please use the transceiver supplied with OpenBTS.
118
119
== Source code ==
120 16 ttsou
121 1 ttsou
The source code is available from git.osmocom.org (module osmo-trx).
122
123
Public read-only access is available via
124 18 ttsou
 git clone git://git.osmocom.org/osmo-trx
125 1 ttsou
You can browse it via cgit: http://cgit.osmocom.org/cgit/osmo-trx/
126 18 ttsou
127
== Configuration and Build ==
128 19 ttsou
129 1 ttsou
The only package dependency is the Universal Hardware Driver (UHD), which is available from Ettus Research or Fairwaves depending on the device. Please note that the UHD implementation must match hardware (i.e. Ettus Research UHD for USRP devices and Fairwaves UHD with UmTRX). The one device that does not use the UHD driver is the USRP1, which is supported through the legacy libusrp driver provided in GNU Radio 3.4.2.
130
131 18 ttsou
'''Intel Platforms (All)'''
132
133
Intel SSE support is automatically detected on Intel x86 platforms. No user intervention is necessary. The general configuration defaults to the low phase error modulator. Atom users may wish to use the low-CPU utilization modulator, which can be later enabled from the command line at runtime.
134
{{{
135
$ ./configure
136
...
137 1 ttsou
checking whether mmx is supported... yes
138 18 ttsou
checking whether sse is supported... yes
139
checking whether sse2 is supported... yes
140
checking whether sse3 is supported... yes
141
checking whether ssse3 is supported... yes
142
checking whether sse4.1 is supported... yes
143 1 ttsou
checking whether sse4.2 is supported... yes
144 19 ttsou
...
145 18 ttsou
}}}
146
147
'''ARM Platforms with NEON'''
148
149
Many popular ARM development boards fall under this category including !BeagleBoard, !PandaBoard, and Ettus E100 USRP. This option will disable the low phase error modulator, which can be re-enabled at runtime. NEON support must be manually enabled.
150
{{{
151 19 ttsou
$ ./configure --with-neon
152 18 ttsou
}}}
153
154
'''ARM Platforms with NEON-VFPv4'''
155
156
Currently very few development platforms support this instruction set, which is seen mainly in high end smartphones and tablets. Available development boards are !ArndaleBoard and ODROID-XU. This option will disable the low phase error modulator, which can be re-enabled at runtime. NEON-VFPv4 support must be manually enabled.
157
{{{
158
$ ./configure --with-neon-vfpv4
159
}}}
160
161
'''ARM Platforms without NEON'''
162
163
This configuration mainly targets the Raspberry Pi. ARM platforms without NEON vector units are almost always very slow processors, and generally not very suitable for running OsmoTRX. Running OsmoTRX on a Raspberry Pi, however, is possible along with limited TCH (voice) channel support. Currently this configuration requires minor code changes.
164 1 ttsou
165 24 ttsou
Coming soon...
166 18 ttsou
167 16 ttsou
'''Build and Install'''
168
169
After configuration, installation is simple.
170
171
{{{
172 18 ttsou
$ make
173 19 ttsou
$ sudo make install
174 16 ttsou
}}}
175
176
== Running ==
177
178
OsmoTRX can be configured with a variety of options on the command line. In most cases, the default settings will suffice. Notable options include UHD device argument passing, which is often useful for using network based devices with firewalls, and external 10 MHz reference support.
179
180
{{{
181
$ osmo-trx -h
182
linux; GNU C++ version 4.8.1 20130603 (Red Hat 4.8.1-1); Boost_105300; UHD_003.005.004-140-gfb32ed16
183
184
Options:
185
  -h    This text
186
  -a    UHD device args
187
  -l    Logging level (EMERG, ALERT, CRT, ERR, WARNING, NOTICE, INFO, DEBUG)
188
  -i    IP address of GSM core
189
  -p    Base port number
190
  -d    Enable dual channel diversity receiver
191
  -x    Enable external 10 MHz reference
192
  -s    Samples-per-symbol (1 or 4)
193
  -c    Number of ARFCN channels (default=1)
194
}}}
195
196
{{{
197
$ osmo-trx -a "addr=192.168.10.2"
198
linux; GNU C++ version 4.8.1 20130603 (Red Hat 4.8.1-1); Boost_105300; UHD_003.004.000-b14cde5
199
200
Config Settings
201
   Log Level............... INFO
202
   Device args............. addr=192.168.10.2
203
   TRX Base Port........... 5700
204 1 ttsou
   TRX Address............. 127.0.0.1
205 16 ttsou
   Channels................ 1
206 1 ttsou
   Samples-per-Symbol...... 4
207
   External Reference...... Disabled
208 16 ttsou
   Diversity............... Disabled
209
210
-- Opening a UmTRX device...
211
-- Current recv frame size: 1472 bytes
212
-- Current send frame size: 1472 bytes
213
-- Setting UmTRX 4 SPS
214 19 ttsou
-- Transceiver active with 1 channel(s)
215 13 ttsou
}}}
216 19 ttsou
217
== Benchmarks ==
218 17 ttsou
219 13 ttsou
A variety of performance benchmarks are available for various code optimizations. These include floating point - integer conversions, convolution, and convolutional decoding. Note that convolutional decoding does not take place in OsmoTRX, but one stop higher in the Layer 1 stack - either in OsmoBTS or OpenBTS core.
220
221 1 ttsou
Selected benchmark results are provided below. All tests are run on a single core only.
222 24 ttsou
223 13 ttsou
224
'''Intel Haswell (i7 4770K 3.5 GHz)'''
225
226
{{{
227 10 ttsou
--- Floating point to integer conversions
228
-- Testing 40000 iterations of 3120 values
229 1 ttsou
- Measuring conversion time
230
- Elapsed time base...                  0.065508 secs
231
- Validating SIMD conversion results... PASS
232
- Measuring conversion time
233
- Elapsed time SIMD ...                 0.011424 secs
234
- Speedup...                            5.734244
235
}}}
236
237
{{{
238
[+] Testing: GSM TCH/AFS 7.95 (recursive, flushed, punctured)
239
[.] Input length  : ret = 165  exp = 165 -> OK
240
[.] Output length : ret = 448  exp = 448 -> OK
241 3 ttsou
[.] Pre computed vector checks:
242
[..] Encoding: OK
243
[..] Decoding base: 
244
[..] Decoding SIMD: 
245
[..] Code N 3
246
[..] Code K 7
247
OK
248
[.] Random vector checks:
249
[.] Testing baseline:
250
[..] Encoding / Decoding 10000 cycles:
251
[.] Elapsed time........................ 1.435066 secs
252
[.] Rate................................ 3.121808 Mbps
253 1 ttsou
[.] Testing SIMD:
254
[..] Encoding / Decoding 10000 cycles:
255 17 ttsou
[.] Elapsed time........................ 0.073524 secs
256 1 ttsou
[.] Rate................................ 60.932485 Mbps
257
[.] Speedup............................. 19.518334
258
}}}
259
260 17 ttsou
'''Intel Atom (D2500 1.86 GHz)'''
261
{{{
262
--- Floating point to integer conversions
263
-- Testing 40000 iterations of 3120 values
264
- Measuring conversion time
265
- Elapsed time base...                 1.147449 secs
266
- Validating SSE conversion results... PASS
267
- Measuring conversion time
268
- Elapsed time SSE ...                 0.347838 secs
269
- Quotient...                          3.298803
270
}}}
271
272
{{{
273
[+] Testing: GSM TCH/AFS 7.95 (recursive, flushed, punctured)
274
[.] Input length  : ret = 165  exp = 165 -> OK
275
[.] Output length : ret = 448  exp = 448 -> OK
276
[.] Pre computed vector checks:
277
[..] Encoding: OK
278
[..] Decoding base: 
279
[..] Decoding SIMD: 
280
[..] Code N 3
281 1 ttsou
[..] Code K 7
282
OK
283
[.] Random vector checks:
284 17 ttsou
[.] Testing baseline:
285
[..] Encoding / Decoding 10000 cycles:
286
[.] Elapsed time........................ 11.822688 secs
287
[.] Rate................................ 0.378932 Mbps
288
[.] Testing SIMD:
289
[..] Encoding / Decoding 10000 cycles:
290
[.] Elapsed time........................ 0.550423 secs
291 19 ttsou
[.] Rate................................ 8.139195 Mbps
292
[.] Speedup............................. 21.479277
293
}}}
294 17 ttsou
295
'''!ArndaleBoard (ARM Cortex-A15 1.7 GHz)'''
296
297
Please note that the Viterbi implementations on ARM is largely C based with speedup generated primarily through algorithm changes. In comparison, vector optimization on Intel platforms with SSE is currently much more aggressive, which explains the disparity on decoding performance.
298
299
{{{
300
--- Floating point to integer conversions
301
-- Testing 40000 iterations of 3120 values
302
- Measuring conversion time
303
- Elapsed time base...                 0.384097 secs
304
- Validating SSE conversion results... PASS
305
- Measuring conversion time
306
- Elapsed time SSE ...                 0.100877 secs
307
- Quotient...                          3.807578
308
}}}
309
310
{{{
311
[+] Testing: GSM TCH/AFS 7.95 (recursive, flushed, punctured)
312
[.] Input length  : ret = 165  exp = 165 -> OK
313
[.] Output length : ret = 448  exp = 448 -> OK
314
[.] Pre computed vector checks:
315
[..] Encoding: OK
316
[..] Decoding base: 
317
[..] Decoding SIMD: 
318
[..] Code N 3
319
[..] Code K 7
320
OK
321
[.] Random vector checks:
322
[.] Testing baseline:
323
[..] Encoding / Decoding 10000 cycles:
324
[.] Elapsed time........................ 5.371288 secs
325
[.] Rate................................ 0.834064 Mbps
326
[.] Testing SIMD:
327
[..] Encoding / Decoding 10000 cycles:
328
[.] Elapsed time........................ 1.016621 secs
329
[.] Rate................................ 4.406755 Mbps
330
[.] Speedup............................. 5.283471
331
}}}
332
333 3 ttsou
'''!BeagleBoard-xM (ARM Cortex-A8 800 MHz)'''
334
{{{
335
--- Floating point to integer conversions
336
-- Testing 40000 iterations of 3120 values
337
- Measuring conversion time
338
- Elapsed time base...                  6.292542 secs
339
- Validating SIMD conversion results... PASS
340
- Measuring conversion time
341
- Elapsed time SIMD ...                 0.839081 secs
342 5 ttsou
- Quotient...                           7.499326
343 3 ttsou
}}}
344
345
{{{
346 4 ttsou
[+] Testing: GSM TCH/AFS 7.95 (recursive, flushed, punctured)
347 3 ttsou
[.] Input length  : ret = 165  exp = 165 -> OK
348
[.] Output length : ret = 448  exp = 448 -> OK
349
[.] Pre computed vector checks:
350
[..] Encoding: OK
351
[..] Decoding base: 
352
[..] Decoding SIMD: 
353
[..] Code N 3
354
[..] Code K 7
355
OK
356
[.] Random vector checks:
357 1 ttsou
[.] Testing baseline:
358 3 ttsou
[..] Encoding / Decoding 10000 cycles:
359
[.] Elapsed time........................ 21.963257 secs
360
[.] Rate................................ 0.203977 Mbps
361
[.] Testing SIMD:
362 1 ttsou
[..] Encoding / Decoding 10000 cycles:
363
[.] Elapsed time........................ 3.083282 secs
364 17 ttsou
[.] Rate................................ 1.452997 Mbps
365 1 ttsou
[.] Speedup............................. 7.123337
366
}}}
367
368 31 ttsou
369
'''Full Results'''
370
371
[http://tsou.cc/gsm/haswell.txt]
372
373
[http://tsou.cc/gsm/shuttle.txt]
374
375
[http://tsou.cc/gsm/arndale.txt]
376
377 1 ttsou
[http://tsou.cc/gsm/beagle.txt]
378 32 ttsou
379
'''Repository'''
380
381
Currently the trx-bench repository holds the test files and contains the same NEON and SSE code as OsmoTRX. The test code may be merged into OsmoTRX at a later time, but, for now, it exists as a separate repository. NEON configure options are the same as OsmoTRX.
382
383
{{{
384
$ git clone https://github.com/ttsou/trx-bench.git
385
386
$ cd trx-bench
387
$ autoreconf -i
388
$ ./configure [--with-neon] [--with-neon-vfp4]
389
$ make
390
$ src/conv_test
391
$ src/convert_test
392
$ src/convolve_test
393
}}}
394 31 ttsou
395 1 ttsou
== Authors ==
396
397 30 ttsou
OsmoTRX is currently developed and maintained by Thomas Tsou with generous support from Fairwaves, the Open Technology Institute, and Ettus Research. The code is derived from the OpenBTS project, which was originally developed by David Burgess and Harvind Samra at Range Networks.
Add picture from clipboard (Maximum size: 48.8 MB)