1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
|
"""
Convert data length to radio TX/RX energy.
Contains classes for some embedded CPUs/MCUs. Given a configuration, each
class can convert a cycle count to an energy consumption.
"""
import numpy as np
def get_class(radio_name: str):
"""Return model class for radio_name."""
if radio_name == 'CC1200tx':
return CC1200tx
if radio_name == 'NRF24L01tx':
return NRF24L01tx
if radio_name == 'NRF24L01dtx':
return NRF24L01dtx
if radio_name == 'esp8266dtx':
return ESP8266dtx
def _param_list_to_dict(device, param_list):
param_dict = dict()
for i, parameter in enumerate(sorted(device.parameters.keys())):
param_dict[parameter] = param_list[i]
return param_dict
class CC1200tx:
"""CC1200 TX energy based on aemr measurements."""
name = 'CC1200tx'
parameters = {
'symbolrate' : [6, 12, 25, 50, 100, 200, 250], # ksps
'txbytes' : [],
'txpower' : [10, 20, 30, 40, 47], # dBm = f(txpower)
}
default_params = {
'symbolrate' : 100,
'txpower' : 47,
}
def get_energy(params):
if type(params) != dict:
return CC1200tx.get_energy(_param_list_to_dict(CC1200tx, params))
# Mittlere TX-Leistung, gefitted von AEMR
# Messdaten erhoben bei 3.6V
power = 8.18053941e+04
power -= 1.24208376e+03 * np.sqrt(params['symbolrate'])
power -= 5.73742779e+02 * np.log(params['txbytes'])
power += 1.76945886e+01 * (params['txpower'])**2
power += 2.33469617e+02 * np.sqrt(params['symbolrate']) * np.log(params['txbytes'])
power -= 6.99137635e-01 * np.sqrt(params['symbolrate']) * (params['txpower'])**2
power -= 3.31365158e-01 * np.log(params['txbytes']) * (params['txpower'])**2
power += 1.32784945e-01 * np.sqrt(params['symbolrate']) * np.log(params['txbytes']) * (params['txpower'])**2
# txDone-Timeout, gefitted von AEMR
duration = 3.65513500e+02
duration += 8.01016526e+04 * 1/(params['symbolrate'])
duration -= 7.06364515e-03 * params['txbytes']
duration += 8.00029860e+03 * 1/(params['symbolrate']) * params['txbytes']
# TX-Energie, gefitted von AEMR
# Achtung: Energy ist in µJ, nicht (wie in AEMR-Transitionsmodellen üblich) in pJ
# Messdaten erhoben bei 3.6V
energy = 1.74383259e+01
energy += 6.29922138e+03 * 1/(params['symbolrate'])
energy += 1.13307135e-02 * params['txbytes']
energy -= 1.28121377e-04 * (params['txpower'])**2
energy += 6.29080184e+02 * 1/(params['symbolrate']) * params['txbytes']
energy += 1.25647926e+00 * 1/(params['symbolrate']) * (params['txpower'])**2
energy += 1.31996202e-05 * params['txbytes'] * (params['txpower'])**2
energy += 1.25676966e-01 * 1/(params['symbolrate']) * params['txbytes'] * (params['txpower'])**2
return energy * 1e-6
def get_energy_per_byte(params):
A = 8.18053941e+04
A -= 1.24208376e+03 * np.sqrt(params['symbolrate'])
A += 1.76945886e+01 * (params['txpower'])**2
A -= 6.99137635e-01 * np.sqrt(params['symbolrate']) * (params['txpower'])**2
B = -5.73742779e+02
B += 2.33469617e+02 * np.sqrt(params['symbolrate'])
B -= 3.31365158e-01 * (params['txpower'])**2
B += 1.32784945e-01 * np.sqrt(params['symbolrate']) * (params['txpower'])**2
C = 3.65513500e+02
C += 8.01016526e+04 * 1/(params['symbolrate'])
D = -7.06364515e-03
D += 8.00029860e+03 * 1/(params['symbolrate'])
x = params['txbytes']
# in pJ
de_dx = A * D + B * C * 1/x + B * D * (np.log(x) + 1)
# in µJ
de_dx = 1.13307135e-02
de_dx += 6.29080184e+02 * 1/(params['symbolrate'])
de_dx += 1.31996202e-05 * (params['txpower'])**2
de_dx += 1.25676966e-01 * 1/(params['symbolrate']) * (params['txpower'])**2
#de_dx = (B * 1/x) * (C + D * x) + (A + B * np.log(x)) * D
return de_dx * 1e-6
# PYTHONPATH=lib bin/analyze-archive.py --show-model=all --show-quality=table ../data/*_RF24_no_retries.tar
class NRF24L01tx:
"""NRF24L01+ TX(*) energy based on aemr measurements (32B fixed packet size, (*)ack-await, no retries)."""
name = 'NRF24L01'
parameters = {
'datarate' : [250, 1000, 2000], # kbps
'txbytes' : [],
'txpower' : [-18, -12, -6, 0], # dBm
'voltage' : [1.9, 3.6],
}
default_params = {
'datarate' : 1000,
'txpower' : -6,
'voltage' : 3,
}
# AEMR:
# TX power / energy:
#TX : 0 + regression_arg(0) + regression_arg(1) * 1/(parameter(datarate)) + regression_arg(2) * (19.47+parameter(txpower))**2 + regression_arg(3) * 1/(parameter(datarate)) * (19.47+parameter(txpower))**2
# [6.30323056e+03 2.59889924e+06 7.82186268e+00 8.69746093e+03]
#TX : 0 + regression_arg(0) + regression_arg(1) * 1/(parameter(datarate)) + regression_arg(2) * (19.47+parameter(txpower))**2 + regression_arg(3) * 1/(parameter(datarate)) * (19.47+parameter(txpower))**2
# [7.67932887e+00 1.02969455e+04 4.55116475e-03 2.99786534e+01]
#epilogue : timeout : 0 + regression_arg(0) + regression_arg(1) * 1/(parameter(datarate))
# [ 1624.06589147 332251.93798766]
def get_energy(params):
if type(params) != dict:
return NRF24L01tx.get_energy(_param_list_to_dict(NRF24L01tx, params))
# TX-Leistung, gefitted von AEMR
# Messdaten erhoben bei 3.6V
power = 6.30323056e+03
power += 2.59889924e+06 * 1/params['datarate']
power += 7.82186268e+00 * (19.47+params['txpower'])**2
power += 8.69746093e+03 * 1/params['datarate'] * (19.47+params['txpower'])**2
# TX-Dauer, gefitted von AEMR
duration = 1624.06589147
duration += 332251.93798766 * 1/params['datarate']
# TX-Energie, gefitted von AEMR
# Achtung: Energy ist in µJ, nicht (wie in AEMR-Transitionsmodellen üblich) in pJ
# Messdaten erhoben bei 3.6V
energy = 7.67932887e+00
energy += 1.02969455e+04 * 1/params['datarate']
energy += 4.55116475e-03 * (19.47+params['txpower'])**2
energy += 2.99786534e+01 * 1/params['datarate'] * (19.47+params['txpower'])**2
energy = power * 1e-6 * duration * 1e-6 * np.ceil(params['txbytes'] / 32)
return energy
def get_energy_per_byte(params):
if type(params) != dict:
return NRF24L01tx.get_energy_per_byte(_param_list_to_dict(NRF24L01tx, params))
# in µJ
de_dx = 0
class NRF24L01dtx:
"""nRF24L01+ TX energy based on datasheet values (probably unerestimated)"""
name = 'NRF24L01'
parameters = {
'datarate' : [250, 1000, 2000], # kbps
'txbytes' : [],
'txpower' : [-18, -12, -6, 0], # dBm
'voltage' : [1.9, 3.6],
}
default_params = {
'datarate' : 1000,
'txpower' : -6,
'voltage' : 3,
}
# 130 us RX settling: 8.9 mE
# 130 us TX settling: 8 mA
def get_energy(params):
if type(params) != dict:
return NRF24L01dtx.get_energy(_param_list_to_dict(NRF24L01dtx, params))
header_bytes = 7
# TX settling: 130 us @ 8 mA
energy = 8e-3 * params['voltage'] * 130e-6
if params['txpower'] == -18:
current = 7e-3
elif params['txpower'] == -12:
current = 7.5e-3
elif params['txpower'] == -6:
current = 9e-3
elif params['txpower'] == 0:
current = 11.3e-3
energy += current * params['voltage'] * ((header_bytes + params['txbytes']) * 8 / (params['datarate'] * 1e3))
return energy
class ESP8266dtx:
"""esp8266 TX energy based on (hardly documented) datasheet values)"""
name = 'esp8266'
parameters = {
'voltage' : [2.5, 3.0, 3.3, 3.6],
'txbytes' : [],
'bitrate' : [65e6],
'tx_current' : [120e-3],
}
default_params = {
'voltage' : 3,
'bitrate' : 65e6,
'tx_current' : 120e-3
}
def get_energy(params):
# TODO
return 0
def get_energy_per_byte(params):
if type(params) != dict:
return ESP8266dtx.get_energy_per_byte(_param_list_to_dict(ESP8266dtx, params))
# TX in 802.11n MCS7 -> 64QAM, 65/72.2 Mbit/s @ 20MHz channel, 135/150 Mbit/s @ 40MHz
# -> Value for 65 Mbit/s @ 20MHz channel
return params['tx_current'] * params['voltage'] / params['bitrate']
|
