BP-based supervised learning algorithm for multilayer photonic spiking neural network and hardware implementation

We introduce a supervised learning algorithm for photonic spiking neural network (SNN) based on back propagation. For the supervised learning algorithm, the information is encoded into spike trains with different strength, and the SNN is trained according to different patterns composed of different...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Optics express Ročník 31; číslo 10; s. 16549
Hlavní autoři: Zhang, Yahui, Xiang, Shuiying, Han, Yanan, Guo, Xingxing, Zhang, Wu, Tan, Qinggui, Han, Genquan, Hao, Yue
Médium: Journal Article
Jazyk:angličtina
Vydáno: United States 08.05.2023
ISSN:1094-4087, 1094-4087
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract We introduce a supervised learning algorithm for photonic spiking neural network (SNN) based on back propagation. For the supervised learning algorithm, the information is encoded into spike trains with different strength, and the SNN is trained according to different patterns composed of different spike numbers of the output neurons. Furthermore, the classification task is performed numerically and experimentally based on the supervised learning algorithm in the SNN. The SNN is composed of photonic spiking neuron based on vertical-cavity surface-emitting laser which is functionally similar to leaky-integrate and fire neuron. The results prove the demonstration of the algorithm implementation on hardware. To seek ultra-low power consumption and ultra-low delay, it is great significance to design and implement a hardware-friendly learning algorithm of photonic neural networks and realize hardware-algorithm collaborative computing.
AbstractList We introduce a supervised learning algorithm for photonic spiking neural network (SNN) based on back propagation. For the supervised learning algorithm, the information is encoded into spike trains with different strength, and the SNN is trained according to different patterns composed of different spike numbers of the output neurons. Furthermore, the classification task is performed numerically and experimentally based on the supervised learning algorithm in the SNN. The SNN is composed of photonic spiking neuron based on vertical-cavity surface-emitting laser which is functionally similar to leaky-integrate and fire neuron. The results prove the demonstration of the algorithm implementation on hardware. To seek ultra-low power consumption and ultra-low delay, it is great significance to design and implement a hardware-friendly learning algorithm of photonic neural networks and realize hardware-algorithm collaborative computing.
We introduce a supervised learning algorithm for photonic spiking neural network (SNN) based on back propagation. For the supervised learning algorithm, the information is encoded into spike trains with different strength, and the SNN is trained according to different patterns composed of different spike numbers of the output neurons. Furthermore, the classification task is performed numerically and experimentally based on the supervised learning algorithm in the SNN. The SNN is composed of photonic spiking neuron based on vertical-cavity surface-emitting laser which is functionally similar to leaky-integrate and fire neuron. The results prove the demonstration of the algorithm implementation on hardware. To seek ultra-low power consumption and ultra-low delay, it is great significance to design and implement a hardware-friendly learning algorithm of photonic neural networks and realize hardware-algorithm collaborative computing.We introduce a supervised learning algorithm for photonic spiking neural network (SNN) based on back propagation. For the supervised learning algorithm, the information is encoded into spike trains with different strength, and the SNN is trained according to different patterns composed of different spike numbers of the output neurons. Furthermore, the classification task is performed numerically and experimentally based on the supervised learning algorithm in the SNN. The SNN is composed of photonic spiking neuron based on vertical-cavity surface-emitting laser which is functionally similar to leaky-integrate and fire neuron. The results prove the demonstration of the algorithm implementation on hardware. To seek ultra-low power consumption and ultra-low delay, it is great significance to design and implement a hardware-friendly learning algorithm of photonic neural networks and realize hardware-algorithm collaborative computing.
Author Han, Yanan
Zhang, Wu
Han, Genquan
Guo, Xingxing
Zhang, Yahui
Tan, Qinggui
Hao, Yue
Xiang, Shuiying
Author_xml – sequence: 1
  givenname: Yahui
  surname: Zhang
  fullname: Zhang, Yahui
– sequence: 2
  givenname: Shuiying
  orcidid: 0000-0002-1698-2083
  surname: Xiang
  fullname: Xiang, Shuiying
– sequence: 3
  givenname: Yanan
  surname: Han
  fullname: Han, Yanan
– sequence: 4
  givenname: Xingxing
  surname: Guo
  fullname: Guo, Xingxing
– sequence: 5
  givenname: Wu
  surname: Zhang
  fullname: Zhang, Wu
– sequence: 6
  givenname: Qinggui
  surname: Tan
  fullname: Tan, Qinggui
– sequence: 7
  givenname: Genquan
  surname: Han
  fullname: Han, Genquan
– sequence: 8
  givenname: Yue
  surname: Hao
  fullname: Hao, Yue
BackLink https://www.ncbi.nlm.nih.gov/pubmed/37157731$$D View this record in MEDLINE/PubMed
BookMark eNpt0T1PwzAQBmALgfge-APIIwwBO3ZiZ4SqfEhIZYA5cpxLa3DsYDsg_j2tWhBCTPcOz510dwdo23kHCJ1QckFZyS9n0wsuBeFiC-1TUvGMEym2f-U9dBDjCyGUi0rsoj0maCEEo_vo7foxa1SEFsdxgPBuVtGCCs64OVZ27oNJix53PuB-tMlY9QkBDwufvDMax8G8rqSDMSi7LOnDh1esXIsXKrQfKgA2_WChB5dUMt4doZ1O2QjHm3qInm-mT5O77GF2ez-5esg0y0nKmqaSHW8gV0IWmrGmzHOQhIIuqlK0vGKs06QjuYSKqKbkrdBKakHbVpQFEHaIztZzh-DfRoip7k3UYK1y4MdY55LSopQsr5b0dEPHpoe2HoLpVfisv8-0BOdroIOPMUD3QyipVy-oZ9N6_YKlvfxjtVlvnoIy9p-OL1OoigE
CitedBy_id crossref_primary_10_35848_1347_4065_ad53b2
Cites_doi 10.1007/s11432-020-3040-1
10.1109/JSTQE.2017.2678170
10.1038/s41586-019-1157-8
10.1109/IPC53466.2022.9975676
10.1103/PhysRevA.72.033808
10.1126/science.1254642
10.1016/S0030-4018(98)00568-9
10.1007/s10489-022-04258-w
10.1038/323533a0
10.1038/nn1643
10.1364/PRJ.413742
10.1364/OE.476110
10.1109/LED.2022.3218626
10.1109/JLT.2022.3146157
10.1109/ACCESS.2018.2878940
10.1109/TASLP.2022.3221011
10.1063/1.1330217
10.1515/nanoph-2022-0441
10.1109/TNNLS.2020.3006263
10.1109/JSTQE.2019.2929187
10.1109/MSP.2019.2931595
10.1364/ACPC.2021.T4A.244
10.1109/JSTQE.2017.2685140
10.1109/MM.2018.112130359
10.1109/JSTQE.2022.3226138
10.3390/electronics11132097
10.34133/icomputing.0031
10.1364/OPTICA.468347
10.1109/TBCAS.2017.2759700
10.1364/OPTICA.475493
10.1109/LES.2020.3025873
10.1162/neco.2009.11-08-901
10.1364/OPTCON.461448
10.1109/JSTQE.2004.837012
10.1109/TCDS.2018.2833071
10.1109/JSTQE.2022.3217819
10.1109/JSTQE.2013.2257700
ContentType Journal Article
DBID AAYXX
CITATION
NPM
7X8
DOI 10.1364/OE.487047
DatabaseName CrossRef
PubMed
MEDLINE - Academic
DatabaseTitle CrossRef
PubMed
MEDLINE - Academic
DatabaseTitleList PubMed
CrossRef
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Physics
EISSN 1094-4087
ExternalDocumentID 37157731
10_1364_OE_487047
Genre Journal Article
GroupedDBID ---
123
29N
2WC
8SL
AAFWJ
AAWJZ
AAYXX
ABGOQ
ACGFO
ADBBV
AEDJG
AENEX
AFPKN
AKGWG
ALMA_UNASSIGNED_HOLDINGS
ATHME
AYPRP
AZSQR
AZYMN
BAWUL
BCNDV
CITATION
CS3
DIK
DSZJF
DU5
E3Z
EBS
F5P
GROUPED_DOAJ
GX1
KQ8
M~E
OFLFD
OK1
OPJBK
OPLUZ
OVT
P2P
RNS
ROL
ROS
TR2
TR6
XSB
NPM
ROP
7X8
ID FETCH-LOGICAL-c320t-bb98f4be2a785c33b622e801ec5967d4933fc0f028e90ab64d7ca8c71dd765e03
ISICitedReferencesCount 4
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000996319400008&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1094-4087
IngestDate Thu Jul 10 18:46:52 EDT 2025
Wed Feb 19 02:23:48 EST 2025
Sat Nov 29 06:06:57 EST 2025
Tue Nov 18 21:25:27 EST 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 10
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c320t-bb98f4be2a785c33b622e801ec5967d4933fc0f028e90ab64d7ca8c71dd765e03
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0002-1698-2083
OpenAccessLink https://doi.org/10.1364/oe.487047
PMID 37157731
PQID 2811568329
PQPubID 23479
ParticipantIDs proquest_miscellaneous_2811568329
pubmed_primary_37157731
crossref_primary_10_1364_OE_487047
crossref_citationtrail_10_1364_OE_487047
PublicationCentury 2000
PublicationDate 2023-05-08
2023-May-08
20230508
PublicationDateYYYYMMDD 2023-05-08
PublicationDate_xml – month: 05
  year: 2023
  text: 2023-05-08
  day: 08
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle Optics express
PublicationTitleAlternate Opt Express
PublicationYear 2023
References Zhang (oe-31-10-16549-R9) 2021; 64
Vladimirov (oe-31-10-16549-R33) 2005; 72
Merolla (oe-31-10-16549-R16) 2014; 345
Wu (oe-31-10-16549-R25) 2023; 29
Nahmias (oe-31-10-16549-R38) 2013; 19
Chlouverakis (oe-31-10-16549-R32) 2004; 10
oe-31-10-16549-R19
Fu (oe-31-10-16549-R30) 2022; 30
Moradi (oe-31-10-16549-R18) 2018; 12
Jha (oe-31-10-16549-R31) 2022; 40
Deng (oe-31-10-16549-R34) 2017; 23
Xiang (oe-31-10-16549-R40) 2023; 10
Arsalan (oe-31-10-16549-R5) 2022; 1
Lian (oe-31-10-16549-R26) 2022; 43
Ponulak (oe-31-10-16549-R6) 2010; 22
Chen (oe-31-10-16549-R43) 2022; 1
Xiang (oe-31-10-16549-R46) 2021; 32
Deng (oe-31-10-16549-R37) 2018; 6
Xiao (oe-31-10-16549-R1) 2023; 31
Lee (oe-31-10-16549-R8) 2019; 11
Xiang (oe-31-10-16549-R35) 2017; 23
Balaji (oe-31-10-16549-R4) 2020; 13
Filipovich (oe-31-10-16549-R22) 2022; 9
Dubbeldam (oe-31-10-16549-R39) 1999; 159
Han (oe-31-10-16549-R45) 2021; 9
Xiang (oe-31-10-16549-R13) 2022; 11
Gütig (oe-31-10-16549-R7) 2006; 9
Pammi (oe-31-10-16549-R42) 2020; 26
Shi (oe-31-10-16549-R24) 2023; 29
Davies (oe-31-10-16549-R17) 2018; 38
Liu (oe-31-10-16549-R27) 2021
oe-31-10-16549-R23
Willemsen (oe-31-10-16549-R36) 2000; 77
Rumelhart (oe-31-10-16549-R47) 1986; 323
Neftci (oe-31-10-16549-R15) 2019; 36
Brückerhoff-Plückelmann (oe-31-10-16549-R20) 2023; 12
Feldmann (oe-31-10-16549-R44) 2019; 569
References_xml – volume: 64
  start-page: 122403
  year: 2021
  ident: oe-31-10-16549-R9
  publication-title: Sci. China Inf. Sci.
  doi: 10.1007/s11432-020-3040-1
– volume: 23
  start-page: 1
  year: 2017
  ident: oe-31-10-16549-R35
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2017.2678170
– volume: 569
  start-page: 208
  year: 2019
  ident: oe-31-10-16549-R44
  publication-title: Nature
  doi: 10.1038/s41586-019-1157-8
– ident: oe-31-10-16549-R19
  doi: 10.1109/IPC53466.2022.9975676
– volume: 72
  start-page: 033808
  year: 2005
  ident: oe-31-10-16549-R33
  publication-title: Phys. Rev. A
  doi: 10.1103/PhysRevA.72.033808
– volume: 345
  start-page: 668
  year: 2014
  ident: oe-31-10-16549-R16
  publication-title: Science
  doi: 10.1126/science.1254642
– volume: 159
  start-page: 325
  year: 1999
  ident: oe-31-10-16549-R39
  publication-title: Opt. Commun.
  doi: 10.1016/S0030-4018(98)00568-9
– volume: 1
  start-page: 1
  year: 2022
  ident: oe-31-10-16549-R5
  publication-title: Appl. Intell.
  doi: 10.1007/s10489-022-04258-w
– volume: 323
  start-page: 533
  year: 1986
  ident: oe-31-10-16549-R47
  publication-title: Nature
  doi: 10.1038/323533a0
– volume: 9
  start-page: 420
  year: 2006
  ident: oe-31-10-16549-R7
  publication-title: Nat. Neurosci.
  doi: 10.1038/nn1643
– volume: 9
  start-page: B119
  year: 2021
  ident: oe-31-10-16549-R45
  publication-title: Photonics Res.
  doi: 10.1364/PRJ.413742
– volume: 30
  start-page: 44943
  year: 2022
  ident: oe-31-10-16549-R30
  publication-title: Opt. Express
  doi: 10.1364/OE.476110
– volume: 43
  start-page: 2192
  year: 2022
  ident: oe-31-10-16549-R26
  publication-title: IEEE Electron Device Lett.
  doi: 10.1109/LED.2022.3218626
– volume: 40
  start-page: 2901
  year: 2022
  ident: oe-31-10-16549-R31
  publication-title: J. Lightwave Technol.
  doi: 10.1109/JLT.2022.3146157
– volume: 6
  start-page: 67951
  year: 2018
  ident: oe-31-10-16549-R37
  publication-title: IEEE Access
  doi: 10.1109/ACCESS.2018.2878940
– volume: 31
  start-page: 439
  year: 2023
  ident: oe-31-10-16549-R1
  publication-title: IEEE/ACM Trans. Audio Speech Lang. Process.
  doi: 10.1109/TASLP.2022.3221011
– volume: 77
  start-page: 3514
  year: 2000
  ident: oe-31-10-16549-R36
  publication-title: Appl. Phys. Lett.
  doi: 10.1063/1.1330217
– volume: 12
  start-page: 819
  year: 2023
  ident: oe-31-10-16549-R20
  publication-title: Nanophotonics
  doi: 10.1515/nanoph-2022-0441
– volume: 32
  start-page: 2494
  year: 2021
  ident: oe-31-10-16549-R46
  publication-title: IEEE Trans. Neural Netw. Learning Syst.
  doi: 10.1109/TNNLS.2020.3006263
– volume: 26
  start-page: 1
  year: 2020
  ident: oe-31-10-16549-R42
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2019.2929187
– volume: 36
  start-page: 51
  year: 2019
  ident: oe-31-10-16549-R15
  publication-title: IEEE Signal Process. Mag.
  doi: 10.1109/MSP.2019.2931595
– year: 2021
  ident: oe-31-10-16549-R27
  article-title: An energy-efficient non-volatile silicon photonic accelerator for convolutional neural networks (NVSP-CNNs)
  doi: 10.1364/ACPC.2021.T4A.244
– volume: 23
  start-page: 1
  year: 2017
  ident: oe-31-10-16549-R34
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2017.2685140
– volume: 38
  start-page: 82
  year: 2018
  ident: oe-31-10-16549-R17
  publication-title: IEEE Micro
  doi: 10.1109/MM.2018.112130359
– volume: 29
  start-page: 1
  year: 2023
  ident: oe-31-10-16549-R24
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2022.3226138
– volume: 11
  start-page: 2097
  year: 2022
  ident: oe-31-10-16549-R13
  publication-title: Electronics
  doi: 10.3390/electronics11132097
– ident: oe-31-10-16549-R23
  doi: 10.34133/icomputing.0031
– volume: 10
  start-page: 162
  year: 2023
  ident: oe-31-10-16549-R40
  publication-title: Optica
  doi: 10.1364/OPTICA.468347
– volume: 12
  start-page: 106
  year: 2018
  ident: oe-31-10-16549-R18
  publication-title: IEEE Trans. Biomed. Circuits Syst.
  doi: 10.1109/TBCAS.2017.2759700
– volume: 9
  start-page: 1323
  year: 2022
  ident: oe-31-10-16549-R22
  publication-title: Optica
  doi: 10.1364/OPTICA.475493
– volume: 13
  start-page: 142
  year: 2020
  ident: oe-31-10-16549-R4
  publication-title: IEEE Comput. Archit. Lett.
  doi: 10.1109/LES.2020.3025873
– volume: 22
  start-page: 467
  year: 2010
  ident: oe-31-10-16549-R6
  publication-title: Neural Comput.
  doi: 10.1162/neco.2009.11-08-901
– volume: 1
  start-page: 1859
  year: 2022
  ident: oe-31-10-16549-R43
  publication-title: Opt. Commun.
  doi: 10.1364/OPTCON.461448
– volume: 10
  start-page: 982
  year: 2004
  ident: oe-31-10-16549-R32
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2004.837012
– volume: 11
  start-page: 384
  year: 2019
  ident: oe-31-10-16549-R8
  publication-title: IEEE Trans. Cogn. Dev. Syst.
  doi: 10.1109/TCDS.2018.2833071
– volume: 29
  start-page: 1
  year: 2023
  ident: oe-31-10-16549-R25
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2022.3217819
– volume: 19
  start-page: 1
  year: 2013
  ident: oe-31-10-16549-R38
  publication-title: IEEE J. Select. Topics Quantum Electron.
  doi: 10.1109/JSTQE.2013.2257700
SSID ssj0014797
Score 2.4424286
Snippet We introduce a supervised learning algorithm for photonic spiking neural network (SNN) based on back propagation. For the supervised learning algorithm, the...
SourceID proquest
pubmed
crossref
SourceType Aggregation Database
Index Database
Enrichment Source
StartPage 16549
Title BP-based supervised learning algorithm for multilayer photonic spiking neural network and hardware implementation
URI https://www.ncbi.nlm.nih.gov/pubmed/37157731
https://www.proquest.com/docview/2811568329
Volume 31
WOSCitedRecordID wos000996319400008&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVAON
  databaseName: DOAJ Directory of Open Access Journals
  customDbUrl:
  eissn: 1094-4087
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014797
  issn: 1094-4087
  databaseCode: DOA
  dateStart: 19980101
  isFulltext: true
  titleUrlDefault: https://www.doaj.org/
  providerName: Directory of Open Access Journals
– providerCode: PRVHPJ
  databaseName: ROAD: Directory of Open Access Scholarly Resources
  customDbUrl:
  eissn: 1094-4087
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0014797
  issn: 1094-4087
  databaseCode: M~E
  dateStart: 19970101
  isFulltext: true
  titleUrlDefault: https://road.issn.org
  providerName: ISSN International Centre
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Jj9MwFLY6w6K5IHbKUhnEAakKpHYS20dABS5M5zBI4RRlcWikNA1NO3Qu_Haet7SVZiQ4cLEi23GkfJ-e3-ZnhF4zUlAOI14J6q4XSDLxsiJlXlb6NONFKEQR6Msm2Okpj2NxNhhcurMwFzVrGr7diva_Qg19ALY6OvsPcPeLQgc8A-jQAuzQ_hXwH848tTUV427TKkGgHmvnAEnrH8tVtZ4vdHqhziasU9C6x-18udaX4XRtpbznY1XoEuBrTJq4jjGoA1q_VKZYtXBZ5z2sVr-dtbrss9y2fWrHvlf6ezrfVK43rpyvGjov3RaqpWFjJjc76n7eaJ9uDNO2bqp1VhCTGmjkqzQCFsxJsFntJmslsN0HLNP8PXmqzlqJKyU9jQJAYjZ9CxaXb4p27oHbLjS6lE1Cxuzyh2W13dARukFYKFQ24Nff0z4AFTDBbCEq-NK7_jsn6LZ781CTucY80WrK-V10x9oX-L3hxT00kM19dEvn-ebdA_TTsQPv2IEdO3DPDgzswDt2YMcObNmBDTuwZQcGdmDHDnzIjofo26fp-ccvnr11w8sp8ddelgleBpkkKeNhTmkWESJBj5F5KCJWBILSMvdL0Eul8NMsCgqWpzxnk6JgUSh9-ggdN8tGPlH1AECbZETSgJRBwVWEHhpQWbkPVjQth-iN-4NJbkvSq5tR6kTHWaMgmU0T89-H6FU_tTV1WK6a9NLBkICUVKGvtJHLTZcQDpZPBLuXGKLHBp9-GYfn02tHnqGTHZmfo-P1aiNfoJv5xbrqViN0xGI-0s6ckSbRH-Euk5I
linkProvider ISSN International Centre
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=BP-based+supervised+learning+algorithm+for+multilayer+photonic+spiking+neural+network+and+hardware+implementation&rft.jtitle=Optics+express&rft.au=Zhang%2C+Yahui&rft.au=Xiang%2C+Shuiying&rft.au=Han%2C+Yanan&rft.au=Guo%2C+Xingxing&rft.date=2023-05-08&rft.eissn=1094-4087&rft.volume=31&rft.issue=10&rft.spage=16549&rft_id=info:doi/10.1364%2FOE.487047&rft_id=info%3Apmid%2F37157731&rft.externalDocID=37157731
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1094-4087&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1094-4087&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1094-4087&client=summon