Hiro SDIEEE PACE 20140429_SAE_Detroit_wireless_evaluation.pdf
DSRC Performance Assessment DSRC Performance Assessment for Crash Warning Applications Fumio Watanabe (Alps Electric North America, Inc.) Carlos Velasquez (Alps Electric North America, Inc.) Hiro Onishi (Alpine Electronics Research of America, Inc.) Fanny Mlinarsky (octoScope, Inc.)
© 2014 Alps Electronics, Inc. Not for commercial distribution.
1
INDEX
1. 2. 3. 4. 5. 6.
Background Objectives Approach‐1: Approach 1: Develop Antenna Develop Antenna Approach‐2: Identify Critical Use Cases Preliminary Assessment Results Preliminary Assessment Results Suggestions: ‐ Industrial performance requirements Industrial performance requirements ‐ Performance assessment Comparison of assessment methodologies ‐ Comparison of assessment methodologies
7. Summary
2
1. Background Crash warning with DSRC is a hot research topic worldwide.
Possibility to detect objectives in NLOS ((Non Line‐of‐Sight) g )
Utilize communication module for other applications. e.g. ‐ Hazard information delivery ‐ Traffic signal timing delivery Traffic signal timing delivery ‐ Electric toll booth, etc
In general, affordable integration, compared to Radar or camera based systems. b d Required: DSRC, Antenna, (GPS*)
*: Can be shared with other applications
Not required: Radar, Camera, Image processor, etc Building
~ SafetySign.com ~ Public announcement by Fujitsu Ten
3
1. Background However wireless communication is NOT almighty. Radio channel impairment ‐1:
Radio channel impairment‐2:
Range, Obstacles in a communication path*
Multipath, Doppler, etc.
*: Trucks/buses, Buildings, etc
e.g. Urban canyon, Tunnel
Building
Dependency of antenna performance /installation:
Radio interference: From other vehicles, from other radio resources from other radio resources
Source US‐DOT
4
2. Objectives Prerequisite for crash warning with DSRC Prerequisite for crash warning with DSRC
I) DSRC performance assessment for crash warning: a) Performance, considering vehicle integration, including antenna performance/installation. b) Performance, considering critical use cases,
II) Unified DSRC performance rules for crash warning: c) Unified performance requirements U ifi d f i , in the entire of the industry d) Repeatable and unified performance assessment methodology
5
3. Approach‐1: Develop Antenna Communication performance strongly depends on antenna performance and profile. Sample‐1:
Sample‐2:
(ordinarily vehicles communicate Antenna (output & sensing) on horizontal planes, but) h i t l l b t) / profile impacts design/usability Vertical antenna profile impacts of crash warning applications. communication performance in mountain/hilly areas. / y
Source US‐DOT
Sample‐3: Too strong peak gain interferes ith other interferes with other communications between neighboring vehicles.
To be determined: Diversity antenna, MIMO, Antenna location etc. ÆAnalyze the effectiveness & unify the rules in the industry
6
3. Approach‐1: Develop Antenna Antenna performance and profiles Antenna performance and profiles
＜
Plane
ZX plane
ZY plane
XY plane
Peak Gain P kG i Average Gain
5.8 5 8 [dBi] ‐0.2 [dBi]
4.4 4 4 [dBi] ‐1.7 [dBi]
5.1 5 1 [dBi] 0.8 [dBi]
＜
＞
ZX Plane
ZY Plane
＜
＞
Diversity
＞
XY Plane 0 10
-30
30 0
Unit : dBi
-10
-60
60
-20 -30 -90
-40
90
-120
120
-150
150 180
0deg
0deg
0deg
X
Z Z X ‐90deg
‐90deg
Y
90deg
Y
90deg
* Condition: 5900MHz. Antenna is installed at ordinary sharkfin position of GM Buick.
180deg
7
3. Approach‐1: Develop Antenna Antenna installation Antenna installation LTE/3G + DSRC‐1 GPS + XM
LTE/3G + DSRC LTE/3G DSRC‐2 2
Di Dimension/shape could be upgraded, per customers’ requests. i /h ld b d d t ’ t 8
4. Approach‐2: Identify Critical Use Cases Identify critical use cases from 6 applications in the demo of ITS World Congress 2011 (by US‐DOT & CAMP VSC‐3) Source 18th ITS World Congress
Reference: US DOT & CAMP VSC‐3, Connected Vehicle Technology Demo, in ITS World Congress 2011 www.itsworldcongress.org/techshowcase_usdot.html
9
4. Approach‐2: Identify Critical Use Cases 6 crash warning candidates 6 crash warning candidates I) FCW (Forward Collision Warning)
II) EEBL (Emergency Electronic Brake Light)
III) BSW (Blind Spot Warning)
IV) DNPW (Do Not Pass Warning)
V) LTA (Left Turn Assist)
VI) IMA
(Intersection Movement Assist)
*HV: Host Vehicle RV: Remote Vehicle Reference: US DOT , V2V‐SP Light Vehicle Driver Acceptance Clinics and Model Deployment Support in ITS World Congress 2011. www.its.dot.gov/presentations/pdf/V2V_SP_WC2011.pdf
10
4. Approach‐2: Identify Critical Use Cases Identified critical use cases Identified critical use cases a) LTA beyond a large vehicle b) EEBL be ond a large ehicle b) EEBL beyond a large vehicle
c) IMA beyond a building without reflection signal ith t fl ti i l
Building
d) IMA beyond a building with reflection signal ith fl ti i l Building
Building
Building
Building
11
5. Preliminary Assessment Results Measurement condition / base‐line test Measurement condition / base‐line test Co d t o s Conditions /Parameters Frequency
Values / Results Values / Results 5890 MHz(=Channel 178), Channel Width = 10 MHz
Output Power Antenna Architecture PER(Packet Error Rate) Count
20 dBm Rx‐diversity (Not MIMO) ‐ Count on physical layer (without multiple continuous transmission) ‐ Used a counter provided by DSRC radio supplier TTx Vehicle: V hi l Mercedes Benz – sedan Rx Vehicle: Toyota Camry – sedan
Antenna Installation
Base‐line Base line test test
Antenna
LOS(Line of Sight) (Line of Sight) ¼ miles ¼ miles((~ 400m) 400m): PER = 0.8 % : PER = 0 8 %
Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
12
5. Preliminary Assessment Results Beyond buses/trucks Beyond buses/trucks PER: 0 2 % PER: 0.2 %
Test car 1
Test car 2
PER: 0 %
Test car 1
Test car 2 Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
13
5. Preliminary Assessment Results Beyond building without reflection signal Beyond building without reflection signal
Corner Rx
Tx
Distance((from the Corner)) Tx Rx 0 m 15 m 15 m 15 m 30 m 15 m 30 m 30 m 30 m 30 m
PER
(Packet Error Rate)
0 % 47.1 % 32.6 % 87 0 % 87.0 %
Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
14
5. Preliminary Assessment Results Beyond building with reflection signal Beyond building with reflection signal
W ll Wall Tx
Rx
Distance(from the Corner) Tx Rx 30 m 30 m 30 m 60 m 43 m 43 m 60 m 60 m
Corner
PER
(Packet Error Rate)
1.4 % 43.9 % 56 8 % 56.8 %
Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
15
6. Suggestions a) Industry Performance Requirement a) Industry Performance Requirement Wide deployment of crash warning with DSRC requires Wide deployment of crash warning with DSRC requires unified DSRC performance rules in the entire industry. (as well as unified standards and protocols)
Unified and repeatable performance measurements should be available in laboratories around the world. 16
6. Suggestions b) Major Assessment Methodologies b) Major Assessment Methodologies
3 major wireless assessment methodologies I) Computer simulation I) Computer simulation II) Field assessment III) Channel emulation
17
6. Suggestions c) Performance Assessment I – Computer Simulation c) Performance Assessment I Computer Simulation
Simulation with Geographical Data
Computer simulation – Concept
Computer simulation – Sample output
Reference: Japan Automotive Research Institute, Prototype integration of the ITS simulator, (Jun. ’10) www.ieee‐jp.org/japancouncil/chapter/VT‐06/vt.files/VTS‐ITS20100623‐5.pdf
18
6. Suggestions c) Performance Assessment II – Field Assessment c) Performance Assessment II Field Assessment Evaluation for “VI) IMA (Intersection Movement Assist)” beyond a building y g with/without reflection signal g at Urban street ◆Test Condition : 4building corner
• System : UMTRI safety pilot test bed • 2‐Vehicles (CAR‐L, CAR‐R) : Honda Accord 2 V hi l (CAR L CAR R) H d A d • DSRC box : DSRC radio supplier • Antenna : ALPS Twin‐Sharkfin • Test location :
・1 building corner : “S.Main ・1 building corner : “S Main St & W William St” in Ann Arbor, Michigan St & W William St” in Ann Arbor Michigan ・4 building corner : “S.Main St & W Liberty St” in Ann Arbor, Michigan
CAR-R
CAR-L
1building corner
Field assessment – Location Map
Field assessment – Test Condition
Measured by Alps on Jun.12 & 13/2013 (in Ann Arbor, MI)
19
6. Suggestions c) Performance Assessment II – Field Assessment ( c) Performance Assessment II Field Assessment (Continued) Comparizon : "1-building corner" vs "4-building corner"
CAR-R CAR R Point-N
Distancce from Car-L to o Corner [m]
80
CAR-R Point-1
CAR-L Point-1
CAR-L Point-N
Field assessment – “1 building corner”
70
1b-test1-1 1b-test2-1 1b-test3-1 1b-test4-1 1b-test5-1
60 50 40
4b-test1-2 4b-test2-2 4b-test3-2 4b-test4-2 4b-test5-2 4b test5 2
30 20 10 0 0
10 20 30 40 50 60 70 Distance from Car-R Car R to Corner [m]
80
Field assessment – Result
Expectation from this field test i) If the distance of each building is more than 20m, it would be less impact of multi‐path from building, but if it is getting more narrow, it should give more influence. ii) 5.9GHz radiation have a quite strong characteristic to go straight, so it covers almost all of 5 9GHz radiation have a quite strong characteristic to go straight so it covers almost all of driver's field of vision. So, it might be better to install antenna at the front of vehicles. Measured by Alps on Jun 12 & 13 ’13 (in Ann Arbor, MI)
20
6. Suggestions c) Performance Assessment III – Channel Emulation c) Performance Assessment III Channel Emulation
Channel emulation Channel emulation – Equipment
Channel emulation Channel emulation – Concept
Reference: F. Mlinarsky, DSRC Evaluation under Controlled Environment, in ITS‐America web‐seminar (Feb. ’13) http://itsa.org/images/Francis/its‐america‐webinar‐30208.pdf
21
6. Suggestions d) Comparison of Assessment Methodologies d) Comparison of Assessment Methodologies I Item
Computer Simulation
II Field Assessment
Actual Module Assessment
Not Available
OK
OK
Easy Radio Parameter Easy Radio Parameter Control
OK
Not Available il bl
OK
Avoidance of Vehicles Drivers, Courses, etc Drivers, Courses, etc Assessment without FCC’s Certification Avoidance of Unexpected Interference
OK
Not Available
OK
OK
Not Available
OK
OK
Not Available Not Available
OK
III Channel Emulation
22
7. Summary DSRC performance is a key for crash warning applications ‐ Develop antenna, considering vehicle integration Develop antenna considering vehicle integration ‐ Identified critical use cases in crash warning
Wider deployment with crash warning with DSRC requires Wider deployment with crash warning with DSRC requires ‐ Industrial DSRC performance requirements ‐ Unified/repeatable methodology to assess performance U ifi d/ t bl th d l t f Æ Auto industry can learn from wireless industry, y y, which utilizes ‘Channel emulation’ (refer to our paper)
23
Acknowledgement
Acknowledgement
To: Ms. Debra Bezzina and UMTRI(University of Michigan Transportation Research Institute) staffs For: + Offering the test vehicles g + Support on DSRC evaluation in Ann Arbor testbed
24
Th k you for your attention!! Thank f tt ti !! Fumio Watanabe ‐ Contact to DSRC antenna and field assessment Alps Electric North America, Inc. fumio watanabe@alps com Tel: +1 248 391 5284 [email protected], Tel: +1‐248‐391‐5284
Fanny Mlinarsky ‐ Contact to DSRC measurement octoScope, Inc. octoScope Inc [email protected], Tel: +1‐978‐376‐5841
Hiro Onishi Alpine Electronics Research of America, Inc. honishi@alpine‐la.com, Tel: +1‐310‐783‐7281 Slide design: Slid d i Mari Hatazawa mhatazawa@alpine‐la.com
25
© 2014 Alps Electronics, Inc. Not for commercial distribution.
1
INDEX
1. 2. 3. 4. 5. 6.
Background Objectives Approach‐1: Approach 1: Develop Antenna Develop Antenna Approach‐2: Identify Critical Use Cases Preliminary Assessment Results Preliminary Assessment Results Suggestions: ‐ Industrial performance requirements Industrial performance requirements ‐ Performance assessment Comparison of assessment methodologies ‐ Comparison of assessment methodologies
7. Summary
2
1. Background Crash warning with DSRC is a hot research topic worldwide.
Possibility to detect objectives in NLOS ((Non Line‐of‐Sight) g )
Utilize communication module for other applications. e.g. ‐ Hazard information delivery ‐ Traffic signal timing delivery Traffic signal timing delivery ‐ Electric toll booth, etc
In general, affordable integration, compared to Radar or camera based systems. b d Required: DSRC, Antenna, (GPS*)
*: Can be shared with other applications
Not required: Radar, Camera, Image processor, etc Building
~ SafetySign.com ~ Public announcement by Fujitsu Ten
3
1. Background However wireless communication is NOT almighty. Radio channel impairment ‐1:
Radio channel impairment‐2:
Range, Obstacles in a communication path*
Multipath, Doppler, etc.
*: Trucks/buses, Buildings, etc
e.g. Urban canyon, Tunnel
Building
Dependency of antenna performance /installation:
Radio interference: From other vehicles, from other radio resources from other radio resources
Source US‐DOT
4
2. Objectives Prerequisite for crash warning with DSRC Prerequisite for crash warning with DSRC
I) DSRC performance assessment for crash warning: a) Performance, considering vehicle integration, including antenna performance/installation. b) Performance, considering critical use cases,
II) Unified DSRC performance rules for crash warning: c) Unified performance requirements U ifi d f i , in the entire of the industry d) Repeatable and unified performance assessment methodology
5
3. Approach‐1: Develop Antenna Communication performance strongly depends on antenna performance and profile. Sample‐1:
Sample‐2:
(ordinarily vehicles communicate Antenna (output & sensing) on horizontal planes, but) h i t l l b t) / profile impacts design/usability Vertical antenna profile impacts of crash warning applications. communication performance in mountain/hilly areas. / y
Source US‐DOT
Sample‐3: Too strong peak gain interferes ith other interferes with other communications between neighboring vehicles.
To be determined: Diversity antenna, MIMO, Antenna location etc. ÆAnalyze the effectiveness & unify the rules in the industry
6
3. Approach‐1: Develop Antenna Antenna performance and profiles Antenna performance and profiles
＜
Plane
ZX plane
ZY plane
XY plane
Peak Gain P kG i Average Gain
5.8 5 8 [dBi] ‐0.2 [dBi]
4.4 4 4 [dBi] ‐1.7 [dBi]
5.1 5 1 [dBi] 0.8 [dBi]
＜
＞
ZX Plane
ZY Plane
＜
＞
Diversity
＞
XY Plane 0 10
-30
30 0
Unit : dBi
-10
-60
60
-20 -30 -90
-40
90
-120
120
-150
150 180
0deg
0deg
0deg
X
Z Z X ‐90deg
‐90deg
Y
90deg
Y
90deg
* Condition: 5900MHz. Antenna is installed at ordinary sharkfin position of GM Buick.
180deg
7
3. Approach‐1: Develop Antenna Antenna installation Antenna installation LTE/3G + DSRC‐1 GPS + XM
LTE/3G + DSRC LTE/3G DSRC‐2 2
Di Dimension/shape could be upgraded, per customers’ requests. i /h ld b d d t ’ t 8
4. Approach‐2: Identify Critical Use Cases Identify critical use cases from 6 applications in the demo of ITS World Congress 2011 (by US‐DOT & CAMP VSC‐3) Source 18th ITS World Congress
Reference: US DOT & CAMP VSC‐3, Connected Vehicle Technology Demo, in ITS World Congress 2011 www.itsworldcongress.org/techshowcase_usdot.html
9
4. Approach‐2: Identify Critical Use Cases 6 crash warning candidates 6 crash warning candidates I) FCW (Forward Collision Warning)
II) EEBL (Emergency Electronic Brake Light)
III) BSW (Blind Spot Warning)
IV) DNPW (Do Not Pass Warning)
V) LTA (Left Turn Assist)
VI) IMA
(Intersection Movement Assist)
*HV: Host Vehicle RV: Remote Vehicle Reference: US DOT , V2V‐SP Light Vehicle Driver Acceptance Clinics and Model Deployment Support in ITS World Congress 2011. www.its.dot.gov/presentations/pdf/V2V_SP_WC2011.pdf
10
4. Approach‐2: Identify Critical Use Cases Identified critical use cases Identified critical use cases a) LTA beyond a large vehicle b) EEBL be ond a large ehicle b) EEBL beyond a large vehicle
c) IMA beyond a building without reflection signal ith t fl ti i l
Building
d) IMA beyond a building with reflection signal ith fl ti i l Building
Building
Building
Building
11
5. Preliminary Assessment Results Measurement condition / base‐line test Measurement condition / base‐line test Co d t o s Conditions /Parameters Frequency
Values / Results Values / Results 5890 MHz(=Channel 178), Channel Width = 10 MHz
Output Power Antenna Architecture PER(Packet Error Rate) Count
20 dBm Rx‐diversity (Not MIMO) ‐ Count on physical layer (without multiple continuous transmission) ‐ Used a counter provided by DSRC radio supplier TTx Vehicle: V hi l Mercedes Benz – sedan Rx Vehicle: Toyota Camry – sedan
Antenna Installation
Base‐line Base line test test
Antenna
LOS(Line of Sight) (Line of Sight) ¼ miles ¼ miles((~ 400m) 400m): PER = 0.8 % : PER = 0 8 %
Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
12
5. Preliminary Assessment Results Beyond buses/trucks Beyond buses/trucks PER: 0 2 % PER: 0.2 %
Test car 1
Test car 2
PER: 0 %
Test car 1
Test car 2 Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
13
5. Preliminary Assessment Results Beyond building without reflection signal Beyond building without reflection signal
Corner Rx
Tx
Distance((from the Corner)) Tx Rx 0 m 15 m 15 m 15 m 30 m 15 m 30 m 30 m 30 m 30 m
PER
(Packet Error Rate)
0 % 47.1 % 32.6 % 87 0 % 87.0 %
Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
14
5. Preliminary Assessment Results Beyond building with reflection signal Beyond building with reflection signal
W ll Wall Tx
Rx
Distance(from the Corner) Tx Rx 30 m 30 m 30 m 60 m 43 m 43 m 60 m 60 m
Corner
PER
(Packet Error Rate)
1.4 % 43.9 % 56 8 % 56.8 %
Measured by octoScope/Alpine on Jun 20 ’13 (in Marlboro, MA)
15
6. Suggestions a) Industry Performance Requirement a) Industry Performance Requirement Wide deployment of crash warning with DSRC requires Wide deployment of crash warning with DSRC requires unified DSRC performance rules in the entire industry. (as well as unified standards and protocols)
Unified and repeatable performance measurements should be available in laboratories around the world. 16
6. Suggestions b) Major Assessment Methodologies b) Major Assessment Methodologies
3 major wireless assessment methodologies I) Computer simulation I) Computer simulation II) Field assessment III) Channel emulation
17
6. Suggestions c) Performance Assessment I – Computer Simulation c) Performance Assessment I Computer Simulation
Simulation with Geographical Data
Computer simulation – Concept
Computer simulation – Sample output
Reference: Japan Automotive Research Institute, Prototype integration of the ITS simulator, (Jun. ’10) www.ieee‐jp.org/japancouncil/chapter/VT‐06/vt.files/VTS‐ITS20100623‐5.pdf
18
6. Suggestions c) Performance Assessment II – Field Assessment c) Performance Assessment II Field Assessment Evaluation for “VI) IMA (Intersection Movement Assist)” beyond a building y g with/without reflection signal g at Urban street ◆Test Condition : 4building corner
• System : UMTRI safety pilot test bed • 2‐Vehicles (CAR‐L, CAR‐R) : Honda Accord 2 V hi l (CAR L CAR R) H d A d • DSRC box : DSRC radio supplier • Antenna : ALPS Twin‐Sharkfin • Test location :
・1 building corner : “S.Main ・1 building corner : “S Main St & W William St” in Ann Arbor, Michigan St & W William St” in Ann Arbor Michigan ・4 building corner : “S.Main St & W Liberty St” in Ann Arbor, Michigan
CAR-R
CAR-L
1building corner
Field assessment – Location Map
Field assessment – Test Condition
Measured by Alps on Jun.12 & 13/2013 (in Ann Arbor, MI)
19
6. Suggestions c) Performance Assessment II – Field Assessment ( c) Performance Assessment II Field Assessment (Continued) Comparizon : "1-building corner" vs "4-building corner"
CAR-R CAR R Point-N
Distancce from Car-L to o Corner [m]
80
CAR-R Point-1
CAR-L Point-1
CAR-L Point-N
Field assessment – “1 building corner”
70
1b-test1-1 1b-test2-1 1b-test3-1 1b-test4-1 1b-test5-1
60 50 40
4b-test1-2 4b-test2-2 4b-test3-2 4b-test4-2 4b-test5-2 4b test5 2
30 20 10 0 0
10 20 30 40 50 60 70 Distance from Car-R Car R to Corner [m]
80
Field assessment – Result
Expectation from this field test i) If the distance of each building is more than 20m, it would be less impact of multi‐path from building, but if it is getting more narrow, it should give more influence. ii) 5.9GHz radiation have a quite strong characteristic to go straight, so it covers almost all of 5 9GHz radiation have a quite strong characteristic to go straight so it covers almost all of driver's field of vision. So, it might be better to install antenna at the front of vehicles. Measured by Alps on Jun 12 & 13 ’13 (in Ann Arbor, MI)
20
6. Suggestions c) Performance Assessment III – Channel Emulation c) Performance Assessment III Channel Emulation
Channel emulation Channel emulation – Equipment
Channel emulation Channel emulation – Concept
Reference: F. Mlinarsky, DSRC Evaluation under Controlled Environment, in ITS‐America web‐seminar (Feb. ’13) http://itsa.org/images/Francis/its‐america‐webinar‐30208.pdf
21
6. Suggestions d) Comparison of Assessment Methodologies d) Comparison of Assessment Methodologies I Item
Computer Simulation
II Field Assessment
Actual Module Assessment
Not Available
OK
OK
Easy Radio Parameter Easy Radio Parameter Control
OK
Not Available il bl
OK
Avoidance of Vehicles Drivers, Courses, etc Drivers, Courses, etc Assessment without FCC’s Certification Avoidance of Unexpected Interference
OK
Not Available
OK
OK
Not Available
OK
OK
Not Available Not Available
OK
III Channel Emulation
22
7. Summary DSRC performance is a key for crash warning applications ‐ Develop antenna, considering vehicle integration Develop antenna considering vehicle integration ‐ Identified critical use cases in crash warning
Wider deployment with crash warning with DSRC requires Wider deployment with crash warning with DSRC requires ‐ Industrial DSRC performance requirements ‐ Unified/repeatable methodology to assess performance U ifi d/ t bl th d l t f Æ Auto industry can learn from wireless industry, y y, which utilizes ‘Channel emulation’ (refer to our paper)
23
Acknowledgement
Acknowledgement
To: Ms. Debra Bezzina and UMTRI(University of Michigan Transportation Research Institute) staffs For: + Offering the test vehicles g + Support on DSRC evaluation in Ann Arbor testbed
24
Th k you for your attention!! Thank f tt ti !! Fumio Watanabe ‐ Contact to DSRC antenna and field assessment Alps Electric North America, Inc. fumio watanabe@alps com Tel: +1 248 391 5284 [email protected], Tel: +1‐248‐391‐5284
Fanny Mlinarsky ‐ Contact to DSRC measurement octoScope, Inc. octoScope Inc [email protected], Tel: +1‐978‐376‐5841
Hiro Onishi Alpine Electronics Research of America, Inc. honishi@alpine‐la.com, Tel: +1‐310‐783‐7281 Slide design: Slid d i Mari Hatazawa mhatazawa@alpine‐la.com
25
